THE IMPACT OF DYSLEXIA ON LEARNING SIGN LANGUAGE. A dissertation submitted to. The University of Bolton. Master of Education

Transcription

1 THE IMPACT OF DYSLEXIA ON LEARNING SIGN LANGUAGE A dissertation submitted to The University of Bolton In part-fulfilment of the requirements for the degree of Master of Education (Professional Development) by Minna Moffatt-Feldman School of Arts, Media and Education University of Bolton Aug 2015

2 Acknowledgments This is dedicated to all those with dyslexia who are, or are contemplating learning sign language. It is hoped that this will serve as an encouragement to continue with your learning; remember you have skills to utilise that those without dyslexia may not have. In addition, thank you to all those who took part in the research and all those that pushed, quizzed and aided me during this crazy time: Ann Canton, Dr. David Kitchener, Dr. Ian Gill, Njal Curlett and last, but by no means least, Rachel and Leo Moffatt-Feldman. Minna Moffatt-Feldman

3 Abstract Sign language learners with dyslexia often assert that they struggle with fingerspelling comprehension and there is no previous research into the impact dyslexia poses on sign language learning or use. This descriptive study explored and highlighted the perceptions and experiences of individuals with dyslexia whilst learning sign language. Focus of interests were language anxieties, any disadvantages to learning sign language and fingerspelling use. In addition comparisons were made between users of American and British sign language systems. A mixed relativist approach to data collection and interpretation was taken; meaning that both qualitative and quantitative data was collected. This was achieved through direct assessment in using the WRAT 4 Single Word Reading and Spelling tests both in English and sign language and through questionnaires. The research highlighted that individuals felt dyslexia did not pose a disadvantage to learning sign language but created a clear disadvantage in the comprehension of fingerspelling. The Wide Range Achievement Test 4 (WRAT4) standardised Single Word Reading and Spelling tests tracked differences between sign language and English use. These identified vast differences between English and sign language word comprehension, yet only marginal differences in spelling ability. Respondents and participants who took part in the research felt the issue was that when reading English words, letters could be viewed simultaneously aiding recall of spelling patterns. However, the comprehension of sign language fingerspelling is difficult for individuals with dyslexia seemingly due to the letters being consecutively produced. Minna Moffatt-Feldman

4 Contents 1.0 Introduction Sign Language Dyslexia Background to Study Phonology Learning Signed Languages Sociolinguistics in Sign Language The Visual Pathway The Affective Filter Hypothesis Methodology and Method Aims and Objectives Study Design Research Questions Access to the Sample Selection of the Sample Data Collection Questionnaires WRAT4 Assessments Informal Interviews Memorandums Ethical Issues Piloting the Research Data Analysis Method Minna Moffatt-Feldman

5 3.10 Credibility and Validity Results and Findings Figure 1: Age of Respondents Figure 2: Time Periods Figure 3: Writing Hands Figure 4: Distinguishing Left from Right Figure 5: Learning to Fingerspell Right Handers Figure 6: Learning to Fingerspell Left Handers Figure 7: Fingerspelling in Left-handed Respondents Figure 8: Hand Dominance Figure 9: Visual Processing Figure 10: WRAT4 Single Word Reading English test (Green Form) Figure 11: WRAT4 English Additional Single Word Reading test (Blue Form) Figure 12: WRAT4 English Spelling test (Green Form) Figure 13: WRAT4 English Averages and Range Figure 14: WRAT4 Additional English Averages and Range Figure 15: Perceptions of Impact on English Figure 16: WRAT4 SL Single Word Reading test (Blue Form) Figure 17: WRAT4 SL Spelling test (Blue Form) Figure 18: WRAT4 SL Averages and Range Figure 19: Perceptions of Impact of Sign Language Figure 20: WRAT4 Single Word Reading Test - Raw Scores Figure 21: WRAT4 Signle Word Spelling Test - Raw Scores Figure 22: WRAT4 Single Word Reading English Only Raw Scores Minna Moffatt-Feldman

6 Figure 23: WRAT4 Single Word Reading Percentile Points Figure 24: WRAT4 Single Word Spelling Percentile Points Figure 25: WRAT4 Single Words Reading English Only Percentile Points Figure 26: Pearson s Correlation Coefficient Calculations Analysis and discussion Hand Dominance Issues Surrounding Phonology American Sign Language and British Sign Language Emotional Responses English or Sign Language Answering the Research Questions Summary and conclusions Implications of the study Limitations of the Study Sample Selection Data Analysis Interviews Dissemination of Findings Recommendations for Future Research Closing Remarks Bibliography Appendices American Sign Language Fingerspelling British Sign Language Fingerspelling Minna Moffatt-Feldman

7 8.3 Research Ethics Form (RH1) Recruiting Participants Form of Consent Participant Information - Questionnaires Participant Information - Assessments Participant Information Post Assessment questionnaire Post Assessment Questions Questionnaire WRAT4 Classifications Assessor Confidentiality Agreement Minna Moffatt-Feldman

8 1.0 Introduction This descriptive study explored and highlighted the perceptions and experiences of people with dyslexia whilst learning sign language (SL). Personal experience informed that learners often disclose their diagnosis of dyslexia during the first sessions of learning a signed language, if not before they commence classes. Blankfield (2001) explains that disclosure of dyslexia causes a great deal of anxiety for some individuals yet on balance disclosure was the correct course of action given reasonable adjustments which could subsequently be made. The oft first taught element of any SL is fingerspelling which has links to the English Language causing anxieties to become elevated. To date, there is very little published research or discourse concerning the combination of dyslexia and SL, yet there is a vast amount of accessible publications describing dyslexia and English. This paper serves to identify the impact, if any, that dyslexia has on learning a signed language as a non-primary language, looking for themes and recommendations to make the learning process smoother for those learners with dyslexia. The research included data collection from the United Kingdom (UK) and the United States of America (USA) seeking to identify any differences between the impact of dyslexia on American Sign Language (ASL) and British Sign Language (BSL). It should be noted that this research focuses on English as the first language and results may vary for those without English as a first language. 1.1 Sign Language The progression of SL linguistic research over the last 50 years has led to the affirmation that SL is a distinct language with a rich grammar system, lexicon and acts as other languages do with changes in reaction to culture, technology and medicine (Sutton-Spence and Woll, 1998). It is not, as individuals unfamiliar with SL may believe, a gestural version 1

9 of spoken language or mime despite the presence of some iconic signs (signs which are easily identifiable due to their visual link with real-life) (ibid, Valli et.al, 2002). BSL is a language of signs, which has developed naturally in Britain and is distinct from English (Sutton-Spence and Woll, 1998). American Sign Language (ASL) is a natural language developed over time and used by the North American Deaf community (Valli et. al, 2002). Despite both America and Britain predominantly speaking English, ASL and BSL are mutually incomprehensible and unrelated to English (Wilcox and Perrin Wilcox, 1997). Both A/BSL have their own vocabulary and grammatical structure, expressed through movement of the hands, body, face and head which, although may have similarities to other countries, are distinct. There are, however SLs that share a common ancestry just as spoken languages do, for example the SL of France, (known as LSF), has a close history with the development of modern ASL to the extent that Lane (1987) estimates that they share approximately 58% of similar signs. The development of SL is a natural part of deaf culture, which is still highlighted today within hearing families into which deaf babies are born (Groce, 1985). Deaf children often display evidence of gestures in order to establish communication, which have become known as home signs (ibid). Although home signs are specific to individual families deaf individuals are not limited to using only them. BSL and ASL both have a vast lexicon of established signs arbitrary (no visual motivation) and iconic signs (visually motivated in nature), which are used to describe concepts (Sutton-Spence and Woll, 1998, Valli et.al, 2002). ASL and BSL (but more so in BSL) have a number of regional variations that mean translations require a skilled language user (A/BSL and English) who can select the appropriate English words to convey accurate meaning (Kyle and Woll, 1985, Sutton-Spence and Woll, 1998, Valli et.al, 2002). This is in contrast to other sign systems that do not have a grammar Minna Moffatt-Feldman

10 system or do they experience regional variations, one such example of this is Makaton (Sutton-Spence and Woll, 1998, The Makaton Charity, 2015). Over the course of history, the education, acceptance and rejection of SL users has caused the development of a sub-culture of deaf individuals who see their differences with the general population as one of language and culture, not disability or disadvantage and thus prefer to be referred to as the Deaf (Capitalised, rather than lowercase deaf) community (Berke, 2010, Ladd, 2010). Deaf people prefer to be distinguished as a distinct community to be respected thus demonstrating a struggle for a positive identity, a feeling shared by common experience and importantly language (Knight and Swanwick, 2002). In contrast, the term 'deaf' is used to represent individuals who are born hearing but later become hearing-impaired, remaining culturally hearing and 'disabled' within this community (ibid). Being referred to as 'hearing-impaired is strongly disliked in the Deaf community due to the insinuation of lacking or being broken as described in the medical model of disability (Ladd, 2010). In turn, this has created some reluctance to be integrated with the culture of hearing peers, creating tension between the Deaf community and non-signers (Ladd, 2010). Sociolinguistic variation is rife within both ASL and BSL yet in both languages the manual dactylology, or fingerspelling, remains constant regardless of other sociolinguistic differences (Valli et. al, 2002, Sutton-Spence and Woll, 1998). Fingerspelling is the only element of SL, which has a direct correspondence with English words using a series of configurations that represent letters to spell out words (Deuchar, 2013). ASL uses one hand to represent all 26 letters (Appendix 8.1) where BSL uses both hands to represent 25 of 26 letters (Appendix 8.2), with the exception of the letter c, which is produced one-handed. It is the only letter that is similar between A/BSL. Fingerspelling is used when there is no sign equivalent for an English word, generally a name or place (ibid). SLs also use fingerspelling in initialised signs, those in which the first letter of the word is manipulated in the phonology Minna Moffatt-Feldman

11 of the sign. For example COURSE is an initialised sign in BSL as it includes the handshape for C within the phonology of the sign. A study carried out by the BSL Corpus Project found that only 2.5% of discourse measured was made up of BSL fingerspelling where in ASL this was 6.4% (Cormier et. al, 2015), which had relatively small sample size (under 700). However, results by Padden and Gunsauls (2003) demonstrate a higher rate of ASL fingerspelling of 12-35% for general discourse. Brennen (2001) explains that BSL has a lower proportion of initialised signs in comparison to ASL, likely because of fingerspelling being produced bi-handed. 1.2 Dyslexia Dyslexia is defined as a language processing disorder affecting the reading and writing of letters, numbers and symbols (generally taken to mean musical) to different degrees (Schneider and Crombie, 2003). It is understood that difficulties occur due to the brain s conflict over processing auditory and visual information (ibid). Individuals with dyslexia can exhibit symptoms of weak short-term and working memory, slow processing speeds, poor phonological processing ability and difficulty with syntax and grammar (Miles, 1993). The British Dyslexia Association (2015) state that dyslexia is a specific learning difficulty which is likely present from birth and affects literacy development and language skills. Dyslexia is characterised through difficulties in phonological processing, processing speed, working memory, rapid naming, and the automatic development of skills, which often do not correspond to the individual s other cognitive abilities. When diagnosing the condition, many practitioners use a discrepancy model, which highlights average or above cognitive ability but suppressed literacy skills compared to their peers (Snowling, 1996). Minna Moffatt-Feldman

12 It is interesting, and important, to note that SLs do not have any auditory input. When English is spoken and written the phonological sequencing ability of the person processes the information in order to assign correct letters to create words. The core phonological deficit hypothesis has provoked a number of studies, which have resulted in the emerging evidence that weak phonological coding is the founding cause of dyslexia (Nijakowska, 2010). Conversely, other scholars favour the Double-deficit Hypothesis (DDH), which argues that an impairment in naming speed is also a prevalent factor in dyslexia (ibid). Moody (2006) explains that those with dyslexia may struggle with reading due to short-term memory problems, which can be aggravated by a slower processing speed. There are also positive aspects of dyslexia that should be mentioned as they may benefit SL use. Individuals with dyslexia are said to be able to create perceptions and are highly aware of their environment as well as naturally thinking and perceiving in a multi-sensory manner (Burm, 2015). Marshall (2012), suggests that dyslexia makes one creative and without dyslexia this creativity may not come to light. Starkiss (2010) describes positive features of dyslexia, which are important in the proper production and reception of sign language use as visual thinking, the ability to read people and above average physical coordination skills. Minna Moffatt-Feldman

13 2.0 Background to Study 2.1 Phonology The term phonology generally relates to the sound systems within language (Crystal, 1980, Chomsky, 2015). Clearly this does not stand within SL, rather this is taken to mean the visual components which make up each sign (Deuchar, 2013, Sutton-Spence and Woll, 1998). Kyle (2015) argues that phonological code is not modality specific and therefore can be derived from the visual aspect of speech (lipreading) as well as audial aspect. The phonology of spoken languages is the pattern of phonics (sounds) and how they form words by their varying positions within words (Coleman, 2015). The phonology of SL was initially debated as a concept, as many believed that signs were holistic unlike words (Wundt, 1921). Seminal research by Stokoe (1960) described SL phonology, listing: handshape, location and movement. Subsequently, researchers began to notate phonology with a total of five phonemes including palm orientation and non-manual features (Sutton-Spence and Woll, 1998). One may infer that as the phonological make up of English and SL differ in sound and silence, then dyslexia cannot present itself when using SL. Studies explained by Sign Sounds (2014) suggest that an alarming number deaf children have severe English language deficits, which if presented in hearing children would encourage urgent specialist intervention. Additionally, a study conducted by Ramus et. al (2002) showed that 75% of participants with dyslexia (hearing) displayed more than one standard deviation below controls in phonological awareness and 50% were similarly impaired in motor skills. This strongly implies a disadvantage for individuals with dyslexia given producing signs require motor skills. Manually producing a signed phoneme incorrectly can result in a total change of meaning as it is with the English Language. Minna Moffatt-Feldman

14 Demonstrating the impact of a change in sign phoneme are the signs for SCARBOROUGH and SLY, which are identical in four phonemes, (handshape, location, movement and orientation). The only difference is the non-manual features produced, namely the mouth movements and facial expression. SCARBOROUGH uses this lip-pattern, but no auditory output, as if speaking the English word Scarborough. The sign SLY, however, is produced with a pout, diagonal eyebrows and a general scowl-type expression. Both of these signs are depicted below: SCARBOROUGH SLY (Pictures: Cath Smith - Let's Sign BSL ). Not only does this highlight the complexity of SL, it also shows the scale of difference that can be produced by a single phoneme change. Similarly, in English when the word kiss can become kill by changing the phoneme /s/ to /l/. There are two models of reading processes for written words, data-driven and concept driven (Reid, 2003). Theorists who favour the data-driven explanation, view that readers process information letter-by-letter whereas those favouring the concept driven view believe that readers absorb meaning from the data holistically (ibid). Minna Moffatt-Feldman

15 2.2 Learning Signed Languages Language learning is the conscious effort to investigate and assimilate the grammar and syntax used in that language with the view of using it correctly (Hamann, 2014). Language acquisition however, is understood to be the process in which language is absorbed naturally akin to a child acquiring language (Krashen, 1987). In general, for hearing people, speech is the primary modality of language with writing being the secondary (Wilcox and Perrin Wilcox, 1997). However, neither of these forms exist naturally in signed languages. For this reason, it is understandable why most individuals who do not know SL mistakenly assume that signed languages have direct correlation with their native spoken language. Wilcox and Perrin Wilcox (1997) assert that this is a reasonable assumption given that within a country people speak and write the same language so it is reasonable to assume people would sign that language. Additional language learners tend to transfer their knowledge of grammar and structure of their first language (L1) to their learning of a second language (L2) (Lado, 1957). This is seemingly uncomplicated for learners of a second oral language but can create errors when L2 is a signed language. Gass and Selinker (2002) state that there are two types of knowledge transfers during additional language learning. Firstly facilitation; where the transfer of knowledge leads to a correct production of L2, and secondly, interference, where the transfer results in an incorrect production of L2. Therefore it is possible that the language deficits of individuals with dyslexia are being transferred to SL. This principal of linguistic knowledge of L1 supporting the understanding and development of L2 is known as the linguistic interdependence model (LIM) (Cummins, 1989). Mayer and Wells (1996) argue it is not accurate to claim that the LIM stands true when L1 is well established BSL and L2 is English (literacy). Moreover, as there is no exposure to speech or English-based signs due to deafness, the conditions do not match the conditions Minna Moffatt-Feldman

16 laid out in the LIM. Using the argument by Mayer and Wells (1996) one can assert that a well-established L1 in English does not result in transferable linguistic knowledge due to the lack of auditory and vocal input within SL, which hearing brains use in processing language. Although the analogy of reading text may be considered, as discussed, SL is not simply letters or words but is concepts and spelling patterns. Pragmatics show that SL conveys meaning through concepts, deixis and presupposition (Sutton-Spence and Woll, 1998). Ellis (2003) explains that L2 learners often experience failures in sociopragmatic and pragmalinguistic language use, which result in difficulties acquiring the target new language. This could possibly be a result of the environment in which the target language is learnt or indeed the learners perception of the environment. Influential social factors would differ according to milieu; classrooms versus a natural setting for instance (Ellis, 2003). It is possible for learners with dyslexia to experience anxiety due to their selfconsciousness of slower processing speeds and weak short term memory, this could be inherent or the product of negative educational experiences. Learners experiencing anxiety due to their environment could perceive situations differently thus resulting in their learning being negatively impacted. Short-term memory is important for SL comprehension, as the on-looker must retain information received from the signer in order for the brain to construct intelligible information. This is achieved by taking into account the phonological form of signs, syntax and semantics (Deuchar, 2013). It is reasonable to assert that a new learner of SL would initially attempt to make sense of the information in L1 before converting it into L2. As those with dyslexia can have slower processing speeds and weaker short-term memory, such a conversion between L1 and L2 can be a lengthy process increasing the possibility of problems with understanding and missing subsequent information. This is predominately different from reading text written in L1, whereas the reader has the opportunity to see all letters in the word simultaneously no conversion between L1 and L2 is needed. Also, with written text one has the ability to re-read. Conversely, if the learner of SL cannot understand Minna Moffatt-Feldman

17 the person signing, the signer can modify their signing to convey the same message in a similar way to speech being re-phrased. This however cannot be done with written text. 2.3 Sociolinguistics in Sign Language SLs, like spoken languages, display both social and regional variations, which may be described through their phonological and lexical use (Bayley et. al, 2015). Unlike spoken languages, SL users will be influenced by when and from whom they acquired their language. Deaf SL users have gathered together for at least 7000 years (Woll and Ladd, 2003) but variances in SLs are often greater than established oral languages due to the breaks in passing signed languages between the generations (Sutton-Spence and Woll, 1998). Bayley et. al (2015) explain that SLs exhibit sociolinguistic variation in a similar pattern to that of spoken languages; and that there has been an increase in the number of studies being undertaking into phonological and morphosyntacic variations. Studies have examined standard variables such as race and gender as well as Deaf specific variables such as being born into a Deaf family and/or attending a Deaf school (Sutton-Spence and Woll, 1998, Bayley et. al, 2015). Some Deaf people claim that they have problems understanding signers from a different generation due to their signing style and lexicon. Older SL users, now over 75 are more likely to predominately use fingerspelling as a means for SL communication, a remnant from their influential educational experiences (Sutton- Spence. and Woll, 1998). Deaf schools have played a major part in the sociolinguistic development of SL. A school s location, language policies and the staffs SL abilities all have contributed to this (Ladd, 2010). Teaching or support staff with a weak SL ability would instantly affect a Deaf child s access to SL, only ever hoping to achieve a SL level akin to those schooling them. Currently, in Greater Manchester personal experience of working with Deaf children in schools show that signs are being modified in all schools by support staff due to their Minna Moffatt-Feldman

18 appearance of similarity to gestures used in the hearing community to symbolise profanity. Some may argue that allowance to amend signs is a natural part of SL evolution whereas others have the view that hearing people do not have the right to change the sacred language of the Deaf community (Ladd, 2012). One such example is the sign HOLIDAY produced by the middle finger on both hands being extended then circled outwards: (Deaf Studies Research Unit, 1992, p. 380) Clearly this could be perceived as inappropriate for a mainstream setting and therefore has seemingly been adapted to use the index fingers rather than the middle. Research suggests that L2 proficiency is influenced by sociolinguistics factors such as age, gender and social class (Ellis, 2003). Interestingly it is also claimed that in the case of ethnicity there is a link between culture similarities and proficiency (Ellis, 2003). Observing SL and its use in the Deaf community, a lay person may assume the Deaf community of USA and UK to have the same cultural traits of hearing people in the USA and UK; this is incorrect. Different generations of Deaf individuals have different cultures depending on their experience during their formative years (Ladd, 2010). In the 1880s, SL was banned from classrooms in favour of an oral approach, but a century later SL re-entered some Deaf pupil s education in Deaf schools (Remark!, 2011). Post Oralism, most Deaf people grew up with an inferiority complex towards hearing people (Ladd, 2010). Minna Moffatt-Feldman

19 Prior to the publication of the Warnock Report (1978) the majority of Deaf children were educated in residential Deaf schools and thus developed sub-cultures specific to the individual schools, or as Ladd (2003, p.297) describes it The Roots of Deaf Culture. Children were sent from all over the country to schools, there was no specific catchment area akin to modern-day Britain (Ladd, 2010). Ladd (2010) explains that it was the norm for Deaf boarders to feel detached from their birth family due to schooling and communication problems at home, and most considered their school their true family. Peggy, (quoted in Ladd, 2010) states that home was all right, but boring. I was excited to get back to school. And as for holiday times, it was like Oh nooo, holiday time is coming (p. 299). This is a stark parallel to mainstream culture and is clearly a factor in the prevalence of sociolinguistics in SL. The Warnock Report (1978) meant less Deaf children were attending residential schools and instead were educated in mainstream, predominately hearing schools due to the process of identifying support needs rather than drawing attention to disability (social model), (Knight and Swanwick, 1999, Remark!, 2011). Not only did this lead to Deaf children being isolated from other Deaf children, it meant the legacy of SL was not passed on to the next generation. Access to SL was minimal and via, in the majority of cases, hearing support staff. This led to significant changes within the language of younger Deaf generations many of whom now use sign supported English (SSE) or Signed English (SE) rather than A/BSL (Ladd, 2012). Mainstream schools encourage the use of SSE or SE as these manual systems can be produced at the same time as English is spoken (Sutton-Spence, and Woll, 1998) thus facilitating learning of the English language and grammar system which differs from A/BSL. Aside from the drawbacks of structure and grammar issues, one must consider that signs selected for use with SSE or SE can be contextually incorrect resulting in misconstrued Minna Moffatt-Feldman

20 information. An example of this are the signs for BOOK: you can book a holiday and you can take a book to read. There are two distinct signs for BOOK and BOOK, rightly so given their meaning is completely different. BOOK (to read a book) is an iconic sign whereas BOOK a holiday is not, as seen here: BOOK (to read a book) BOOK (to book a holiday) (Pictures: Cath Smith - Let's Sign BSL ). Currently oralism is the preferred communication system with 85% of deaf children in the UK being educated this way (National Deaf Children s Society, 2010) meaning that the proportion of SL users is decreasing at a dramatic rate. This coupled with the fact that 90% of deaf children are born to hearing parents (The Hearing Fund UK, 2015), it can be problematic for language transfer between future generations, if SL still exists as a living language. 2.4 The Visual Pathway Studies have shown that the language processes on Deaf SL users occur in the same part of the brain, the Broca s and Wernicke s regions, as spoken languages (MacSweeney et. al, 2008). Moreover, the area used for understanding reading is the same as understanding fingerspelling (Waters et.al, 2007). However, SL only uses one mode for producing the Minna Moffatt-Feldman

21 language (Wilcox and Perrin Wilcox, 1997). Comparison with other languages, which have two modes (spoken and written) can falsely lead people to believe that a signed language is in fact not a language at all. It is true that until March 2003 the UK Government did not recognise BSL as a language in its own right, and it took a 16 year battle for the recognition to finally be passed through parliament (Dominic, and Stiles, 2013). Coughlan (2015) explains that research carried out by Bristol and Newcastle Universities has shown that there are no differences in visual ability between those with dyslexia and those without. Irlen Syndrome is a condition which is often prevalent and associated with dyslexia (Irlen, 2015). To make reading a more comfortable visual process, those with the condition often use coloured overlays or lenses, which work to absorb specific colours of light (ibid). Interestingly, two students with dyslexia currently taught by the researcher use tints to view SL, one with an overlay for computer clips, the other uses specially tinted spectacles. No written English is present during SL production but these students still find the tints beneficial in understanding what is being signed. If what Coughlan (2015) reports is true, it would be of interest to learn why this helps. More interestingly, both dyslexia and Irlen Syndrome are associated with written text, yet this has no place within SL. Technically however, Irlen Syndrome is reportedly a difference in how the brain processes visual information due to the light entering the eyes. This is often heighted during reading (Irlen, 2012) and therefore could have some plausible impact on SL processing, which could contribute evidence that it is the language processing part of the brain that is affected by Irlen Syndrome rather than only visual perception. Further research is definitely warranted in this area and recommendations to support teaching methods for those with Irlen Syndrome would be welcomed. Minna Moffatt-Feldman

22 2.5 The Affective Filter Hypothesis Dulay and Burt (1977) posited the notion of an Affective Filter effectively hindering language acquisition by preventing any input (affective variables presented in the target language) being absorbed fully, thus slowing and hindering the language process. Krashen, (1987) describes how subsequent research has identified three categories of affective variables: motivation, self-confidence and anxiety. He describes how high levels of motivation and self-confidence presented with low levels of anxiety results in a low level of filtering and thus is the optimum situation of second language acquisition assuming that the input is comprehensible, well presented and adequate. It may be argued that dyslexia itself is a type of Affective Filter given that it can cause problems with anxiety, low self-confidence and lack of motivation. Minna Moffatt-Feldman

23 3.0 Methodology and Method 3.1 Aims and Objectives It is important for researchers to distinguish between methodology and method, and be committed to a specific philosophy (Birks and Mills, 2015). Methodology is a set of principles that underpins the philosophy that influences research design, whereas method is the procedure undertaken in order to produce data to analyse (ibid). Exploratory research is a process of discovery for something new (Baker et.al, 2002). An online and library literature search elicited no published evidence of research exploring the impact dyslexia has on learning SL, although as separate topics there is a plethora of published research. It was the aim of this research to formulate a substantive theory concerning the impact, if any, dyslexia has on learning sign language based on a collection of experiences. Results were attained with the view to offering answers to those with dyslexia learning SL as well as presenting a learning opportunity for SL teachers and assessors. 3.2 Study Design Dawson (2009) explains that there has been much debate over quantitative versus qualitative data and which produces the most scientific results. Merriam (1998) explains that research that collects qualitative data uses the philosophy of phenomenology to emphasise experience and interpretation. It is argued that phenomenology is too subjective in nature both to participants and data analysis and as no amount of confirming data can verify a theory, it is treated by many with suspicion (Platt, 2012, Wilson, 2010, Hatch, 2002). Cohen et.al (2007) suggest that scholars who strive to interpret data, begin with individuals and set out to understand their view of the phenomena, in this case learning SL when a person has dyslexia. The term Lebenswelt, a German word which is roughly translated to Minna Moffatt-Feldman

24 mean life-world demonstrates a concept that individuals are embedded in their lived experiences (Brotchie et.al, 2008). Adults with dyslexia have a wealth of both positive and negative lived experiences from within the learning environment. As phenomenologists take the view that individuals try to make sense of their life (Brotchie et.al, 2008) this method of research cannot be easily discounted in favour of pure statistical evidence. Phenomenology carries the assumption that there is an essence to shared experiences (Patton, 1990) and therefore data should speak for itself (Grey, 2006, p.21). Importantly phenomenology itself is not in need of confirming data but rather seeks to describe the Lebenswelt through sifting through the data to identify key areas, themes and classifying data into themes. It is a relatively straight-forward approach for many social science researchers (Corbin and Strauss, 2008); although researchers must ensure that they avoid all influences in order to get to the true nature of the phenomena. Sokolowski (2008) explains that the intentionality of a researcher can have an effect on research and care must be taken to ensure that no bias enters the results from personal perceptions (Baker et.al, 2002). There are two conflicting views surrounding phenomenology, posited by Husserl (1962) and Heidegger (1962), both cited in Brotchie et.al (2008). Husserl (1962) believed that phenomena can be shared by individuals whereas Heidegger (1962) argues that phenomena on individuals can only be in relation to the individual. The researcher follows the view of Husserl, in that although the phenomenon are subjective to individuals and there may be common shared themes, which once deconstructed leaves them open to rigorous scrutiny (ibid). In this case all participants have been diagnosed with dyslexia and have been learning SL. Although both of these elements impact on Lebenswelt, the lives of participants are not limited to these two variants (Brotchie et.al, 2008). Minna Moffatt-Feldman

25 After careful consideration, the researcher believed that a qualitative research method would best support the elucidation of personal experience of the individuals and provide richer data. Thus this would provide a greater understanding of the significant impact dyslexia poses to the individual who are learning SL (Morse, 1991, Davis 2010). However, and most importantly, as quantitative data can support themes and serve to quantify and even support explaining them (Hatch, 2002) a mixed relativist approach to data collection and interpretation was taken; meaning that both qualitative and quantitative data was collected. An anti-positivism phenomenological approach was taken in relation to the qualitative aspect of the research (Dash, 2005). In addition, the coding processes were borrowed from the Grounded Theory Method (GTM) as described by Glaser and Strauss (1967, republished 2012) due to the procedure of impartial coding to identify themes from individuals and their interaction with phenomena (Urquhart, 2013). Codes serve to identify themes in data, which appear interesting to the analyst, and refer to fundamental components of the raw data regarding the phenomenon (Braun, 2006). This provides a basis for exploration (Birks and Mills, 2015) through inductive reasoning. A key feature of phenomenology that fitted with the ideals of the researcher was that there could be many truths, which may coexist and may not be one phenomena that fitted all occasions (Brotchie et.al, 2008). It must be noted that although the coding method was borrowed from GMT as it suited the researcher, GMT methodology was not suitable for the research given that questions were considered prior to embarking on the research in order to obtain ethics approval (Appendix 8.3). Cook (2009) explains that the most useful investigations occur when the researcher is distanced and objectified from the research. Care was taken to keep a distance from literature until after the research was collected and analysed in order to minimise and Minna Moffatt-Feldman

26 researcher bias (Corbin and Strauss, 2008). Furthermore the focus of the research, other than the theme of personal experiences and opinions of if and how dyslexia affects people learning SL, was kept from participants in order not to influence their answers or provide an opportunity for participants to self-adjust due to observation (known as the Hawthorne Effect) (Cherry, 2013). In doing this, one must make some ethical considerations about withholding information from participants, yet in this case it was deemed that they were provided with enough information to make an informed decision about their involvement without causing a Hawthorne Effect (ibid). As previously stated, subjectivity is a concern with qualitative research as different researchers may find different themes and the subjections of the phenomena itself (Wilson, 2010). Therefore, interpretation must be viewed with an air of caution considering the philosophy of the individual researcher(s) and by whom the data was interpreted. To help prevent this and to give themes more validity, it is useful for an external review of the data by a critical reader (without knowledge of the research) to be conducted in order to reduce potential researcher bias. In this case the researcher undertook extensive discourse with the PATOSS assessor making comparisons between the collected qualitative and quantitative data. Grey (2006, p.21) states that: Phenomenology holds that any attempt to understand social reality has to be grounded in people s experience of that social reality. Therefore it seemed appropriate that the people interpreting the data had a good working knowledge of the teaching and use of SL and dyslexia. Qualitative data analysis requires researchers to be pragmatic with an open and personable approach and an enquiring mind given the emphasis on creating a theory (Remenyi, 2014). In essence, this research was both personal and social, with the view to improve access to SL education for those with dyslexia by recognising themes (McNiff et. al., 2003). Minna Moffatt-Feldman

27 The quantitative portion of the research used the Wide Range Achievement Test 4 (WRAT4) standardised single word reading and spelling tests to track differences in results between SL and English use. This method was devised to be used in conjunction with the qualitative data and compare to see if themes were reflected in the quantifiable data. The WRAT4 test, first developed by Bijou and Jastak in 1946, is used to measure academic ability needed for effective learning, communication and thinking, (Dyslexia Action Shop, 2014) and produces raw scores as per the design. Raw scores are converted into standardised scores by using age appropriate tables within the WRAT4 manual. The WRAT4 standardisation is based on the normal distribution bell curve formulated for collating samples from over 3000 people in the sample, ages 5-94 years and 11 months with the mean result being 100 and the standard deviation being 15 points (Wilkinson and Robertson, 2006). For this research only the Single Word Reading and Spelling sub-tests were used, administered in both the intended English form and an amended SL form. Generally this then means that scores can be compared to the mean of the sample and standard deviations can be considered for significance with a predetermined confidence level (Scott, 2015). However, and very importantly, this stands true for the administering of the standard WRAT4 assessment but not for the modified assessment. Converting into standard scores was flawed, as the normal distribution of the test modified to incorporate fingerspelling has never been attempted. The fact that SL raw scores could not be translated into true standardised scores is a stark limitation of using the WRAT4 test in this way but still provided an opportunity to make comparisons. Mrs A. Canton (PATOSS Practicing Certificate number ) agreed to administer the tests (appendix 8.12). According to the manual, the WRAT4 is designed to be administered and interpreted by individuals who have the necessary training and Minna Moffatt-Feldman

28 experience in administering and interpreting individually administered tests (Wilkinson and Robertson, 2006, p.5). Platt (2012) claims that conducting research without a pre-conceived problem statement cannot be done for an academic thesis given the stringent protocols one must pass in order to embark on a research journey. Although in some areas this may be true, in this case the researcher drew theory purely from personal observation of teaching SL to students with dyslexia and wanted to widen the view to others to see if experiences were shared (Husserl, 1967). The qualitative four-stage research model described by Baker et.al (2002) was also used as a guide to during the planning stage supporting the ability to be flexible and reformulate where necessary. Stage 1: forming ideas and information gathering Stage 2: interpreting results Stage 3: comparing with original ideas, Stage 4: being flexible and reformulating when necessary. The analysis style of this research was to look for plausible relationships between the two concepts deriving theory from the qualitative data acquired (Urquhart, 2013). Given the lack of published research focusing on learning signed languages when dyslexia is prevalent, there was little chance of literature influencing coding when deriving theory from the phenomenon. In education, qualitative research holds a key role in shaping policy and practice (O Donoghue and Punch, 2003); currently there is little recognition of the effects of dyslexia on SL apart from examination boards providing a reasonable adjustment of extra time during examinations, akin to that in standard academia on production of a certified dyslexia diagnosis (Signature, 2015). It is this researcher s belief that this may be done in fear of Minna Moffatt-Feldman

29 challenge through the Equality Act (2010), rather than a recognition of the effects of dyslexia given the fact that to date there is no published research on the connection. A timescale was set for the research to be conducted in line with the University of Bolton s submission dates; August 2014 February 2015 along with a plan of action; this was deemed as a reasonable length of time for this scale of research. This, however, was amended in accordance with Baker et.al s (2002) qualitative four-stage research model leading to the research taking longer than expected due to problems scheduling appointments with participants in different time zones between the UK and different parts of the USA. Additionally, the timing of the research was affected by the extensive review of the literature being carried out after the majority of data had been collected. This was done to maintain the approach of suspending ideas and preconceived beliefs during the qualitative data collection (Ray, 1994). It was pivotal for the researcher to have a specific understanding of SL, SL pedagogy and dyslexia itself. This could have also been an issue with the potential for a biased interpretation of results. Action research is practitioner based and always conducted inside a specific situation that inevitably may be influenced due to the individualistic aim of the practitioner (Sokolowski, 2008). However, the blend of qualitative and quantitative data collection helped to prevent this, as well as vigorous restraint and appropriate boundaries for the researcher and an impartial, suitably qualified, individual conducting the testing. Preprepared questions and assessment techniques added the restraint during the assessment and questionnaires. Considerations were also made to possible implications to the participants and respondents; if they became upset or asked to talk further this was granted at no hesitation until distress was allayed, offered by the researcher and the assessor, at the choice of the participant. Minna Moffatt-Feldman

30 The advancement of media in recent decades has provided massive opportunities for action researchers through the ability to record and document data but it is vital that full permission is sought when recording a person s voice or image (McNiff, et. al., 2003). Advances in technology gave the added capability of drawing participants from all over the UK and the USA. Research was conducted through anonymous online data submission and Skype assessments. Filming or recording observed research events is invaluable as data can be reviewed several times to elicit further detail and can improve quality plus add greater validity of claims (McNiff and Whitehead, 2005). Recording Skype transmissions was unnecessary as all data captured throughout the sessions was quantitative and recorded effectively on the standard WRAT4 forms, then translated into results tables, displayed in Chapter Research Questions The initial questions considered for research focused on hearing SL learners with dyslexia and were as follows: 1. To what extent, if any, is dyslexia a disadvantage to learning sign language? 2. Do individuals with dyslexia experience the same 'languageanxieties' over their first language and sign language, and if so are they equal? 3. Reception of fingerspelling - do people with dyslexia feel that understanding fingerspelling presents the same problems to them as understanding written words? The fourth was a consideration from years of practical experience in teaching fingerspelling to students with dyslexia: 4. Does learning fingerspelling improve single word reading and spelling in a student with dyslexia? Minna Moffatt-Feldman

31 After piloting the research with one participant and three respondents, it was decided that the fourth research question needed a larger scale sample in order to appropriately assess the impact and therefore was amended to: 4. Are there any significant differences between single word reading comprehension and spelling in individuals with dyslexia whilst using English or Sign Language? Further reflection led to the addition of a new fifth question: 5. Are there any significant differences between single word reading comprehension and spelling of those with dyslexia using the Two-handed British or One-Handed American fingerspelling systems? It was felt that with these questions, the researcher could help develop novel understanding around those individuals who have dyslexia in the SL classroom, and possibly identify strategies that could be used to aid their inclusion, continued interest and development in SL. 3.4 Access to the Sample Whilst some may feel that access to a broad sample group is pivotal for qualitative data collection, realistically for phenomenological methodology, the key feature is access to specific individuals from whom data surrounding their Lebenswelt may be elicited. In this case social media played a large part in recruiting participants; groups for people learning SL provided a platform for advertising the research and an opportunity for respondents to contact the researcher privately in response when required. Purposeful sampling enables researchers to select samples from which the most can be learned (Merriam, 1998). This is an important part of phenomenology given that themes cannot be elicited for individuals who do not meet the criteria. In addition ensuring that participants want to be involved will ensure they give full and relevant answers, which will provide a good basis to analysis the phenomena, this however can cause problems such as the Hawthorne Effect (Cherry, Minna Moffatt-Feldman

32 2013), attention seeking or skewed results. To do this it was important to set criteria for selection. In this case it was purely two factors: a diagnosis of dyslexia and the learning of SL. Advertisements were placed on social media (Facebook) (Appendix 8.4) for those with dyslexia and those who were learning SL with the view to find participants who fitted both criteria and s were send to contacts such as interpreters and SL students. Qualitative research does not provide a researcher with a guide to sample size (Merriam, 1998). GTM, phenomenology, as well was some other qualitative research models, give researchers the flexibility to adapt as themes form so it is down to the researcher to take note of these and seek more (or indeed less) participants where necessary. Saturation of data is a pivotal point in data collection and the sample size should be adjusted to reach it (Corbin and Strauss, 2008, Merriam, 1998). In addition, qualitative data cannot be used in pursuit of testing hypotheses given that it relates to a lived experience which is not quantifiable to an entire population (Brotchie et.al, 2008); yet some researchers claim it is due to the naturally small number of participants (Scott, 2015). 3.5 Selection of the Sample Involving disabled learners in research requires an understanding of the aim of the research, what participation involves plus any implications for participants and researcher and therefore a form of consent should always be presented to learners and every effort should be made to ensure they understand it in line with the Mental Capacity Act (2005) (Appendix 8.5). One may also consider if it is at all ethical to conduct research with a vulnerable group, especially one which may be regularly targeted for research and therefore institutions ensure that Ethics Committee approval is granted before the research is started (Appendix 8.3). In some cases, the very nature of their disability may make it difficult or Minna Moffatt-Feldman

33 impossible to make a reasoned decision and therefore care should be taken to assess this (Haigh and Williamson, 2009, Mental Capacity Act, 2005). Iacono and Murray (2003), posit that whilst it is crucial to protect vulnerable participant groups and that excessive restrictions should not be placed on researchers so as to preclude valuable research. This, however, seems to ignore the rigorous ethical considerations research bodies go through before agreeing to proceed with a research project (Appendix 8.3). Ensuring that individuals are neither manipulated nor misled is critical to ensuring accurate data capture serving to elucidate the lived experience. The danger is that some could include learning disabled participants without their knowledge or exclude them all together making the data unreliable or skewed (Cameron and Murphy, 2007). In this case, candidates all disclosed that they had a formal diagnosis of dyslexia, and were required to disclose the year of their assessment in the questionnaire. All were asked to review and agreed to the information as laid out in the information sheets (Appendices 8.5, 8.6, 8.7 and 8.8). It was felt sufficient as the document only reports results that occurred under testing conditions and explain the impact of the dyslexia on reading and writing English. Additionally, many adults with dyslexia develop compensatory strategies to overcome some of the effects of dyslexia meaning that the content of reports may be outdated (Spafford and Grosser, 2005). Participants were recruited who had studied A/BSL for at least 6 months and had previously been diagnosed as having dyslexia. A postal questionnaire link was sent to interested participants designed to answer questions about their learning experience and language anxieties in both SL and English. This website link provided by the researcher s employer enabled them to enter information anonymously. Minna Moffatt-Feldman

34 Knowing that participants had dyslexia, care had to be taken regarding the compilation of the questionnaires that they were written in a dyslexia-friendly way. For example, a sansserif font was used in an appropriate size, which could be amended by using the zoom function on the point of use on all web-browsers (Rello, 2013). There was one respondent who disclosed that filling in forms caused him distress and therefore he was given the opportunity to sign/speak his answers via Skype to be entered by a SL interpreter although they never took up this offer despite the researcher following it up and reassuring the respondent to allay distress. 3.6 Data Collection Questionnaires Data collection was undertaken through the use of questionnaires containing closed and open-ended questions with the majority of data coming from the open, questions as it was believed that they would elicit detailed, deep qualitative data (Appendix 8.10). However, such data can be difficult to analyse whereas closed questions can offer quantitative data; which may be easier to analyse, but elicit less detail (Sociology Central, 2015). The use of either depends on the information being sought. A negative factor presented with questionnaires is the potential low return rates verses the time taken analysing the data (Sivo et. al., 2006). Mandatory participation in research is unlikely to be as forthcoming in comparison to voluntary respondents. Researchers should be careful of instances of the Hawthorne Effect (Cherry, 2013) and therefore should keep hypotheses confidential whilst also reinforcing that their responses will remain anonymous and confidential. Closed questions with a Likert scale to analyse (McLeod, 2008) was considered but deemed ineffective for this research due to the differing emotional impact dyslexia presents to individuals. Minna Moffatt-Feldman

35 As previously discussed phenomenological research aims to describe the Lebenswelt through looking at qualitative data for which questionnaires may be a good option, it must be remembered that realistically answers should be viewed with a touch of scepticism given a risk of bias (McNiff and Whitehead, 2005) and belief-bias (Davies, 2010). Additionally, questionnaires do not give researcher the opportunity to elicit additionally details after submission hence also have their draw-backs. Marriam, (1998) describes Epoche as the process that researchers should conduct in order to become aware of or remove bias statements from research. Although this can support the data analysis it is also dangerous in that one may be removing important data from the same, in addition the bias nature of the responses may in themselves provide important data. However, the setting aside of the researcher bias is crucial to conducting sound phenomenological research (ibid). To do this the questionnaires were submitted anonymously so that no persons known to the researcher could be identified and no preconceptions were considered during data analysis, rather the raw data alone was focused on. Open questions generally give opportunities for more in depth answers which may be compared to established literature on the subject, however to date there is very little formal research linking dyslexia and SL and therefore comparisons must be made with foreign/second language acquisition when dyslexia is present and SL acquisition. It was (and is) important that research questions are not leading in order to minimise influence and bias (McNiff, et.al, 2003). It is also important, as with interviews that responses elucidate fundamental beliefs and attitudes rather than superficial answers (Hannan, 2007). An effective way of achieving this is via the laddering interview technique eliciting attributes, consequences and core values from respondents (Hawley, 2009). Piloting serves to allow for research methods to be tested and amended in order to gain the specified data (Dawson, 2009). The questionnaire was piloted with 3 students currently Minna Moffatt-Feldman

36 known to the researcher, from which some questions were identified by the researcher, as being superfluous and non-specific leading to only yes/no answer. Questions were therefore modified to elicit more details within the final draft. Individuals who answered the questionnaires are referred to as respondents WRAT4 Assessments The original plan was to conduct assessments with 12 adults with dyslexia who had been studying BSL for 6-48 months. Individuals who took part in the assessments are referred to as participants and were tested using the WRAT4 Single Word Reading and Spelling tests in English and BSL. However, after the first 5 assessments a clear pattern emerged showing where by every participant scored lower in SL than in English. At this point the comparison with ASL was considered. The degree of flexibility and identification of trends in using Baker et.al s (2002) qualitative research model (stage four), led the researcher down a path of new discovery. After a discussion with the research supervisor it was felt that the assessments in BSL should cease and the remaining should progress using ASL because all tests were producing the same answers. It was also around this time that the research question 5 was added comparing ASL and BSL. For this a rigorous exercise of recruiting participants from the USA began. This was challenging given the scope of finding participants who fitted the criteria, who were willing to take part and that could take part at a mutually agreeable time given the time zone differences. Social media played a crucial part in enabling contact, but due to time constraints for the research only two assessments were carried out with ASL participants. More would have been welcomed. Minna Moffatt-Feldman

37 Prior to the assessment the participants were provided with the information sheets (Appendix 8.7) and given the Skype contact details and time of assessment. The participants then called via Skype to begin the assessment at the specified time. The participants were greeted, introduced to the researcher and the PATOSS assessor, put at ease and then the five stages of the assessment were explained: 1. WRAT4 Reading Test (Green) conducted in English Conducted by the PATOSS Assessor. 2. WRAT4 Spelling Test (Green) conducted in English Conducted by the PATOSS Assessor. 3. WRAT4 Reading Test (Blue) conducted in A/BSL Conducted by the researcher with careful supervision by the PATOSS assessor. 4. WRAT4 Spelling Test (Blue) conducted in A/BSL Conducted by the researcher with careful supervision by the PATOSS assessor. 5. Post assessment questions and invitation to fill out post assessment questionnaire. 1. Administration of the standard WRAT4 Single Word Reading Test (Green form) As the writing on the WRAT4 testing card was too small to present to participants via webcam, they were instead presented through flash cards. First 15 letters were presented and then the 55 English words the presented in order of increasing difficulty as ordered on the WRAT4 card. As with the spelling test it is normal not to include the reading of individual letters for English speaking adults, but was included for the same reasons expressed above. Participants were asked to read the letters and words aloud so that PATOSS assessor could note down their answers on the standard WRAT4 testing sheets. The participants were given 10 seconds to vocalise the word, after which they were asked to move to the next word. In line with the testing procedure, testing was discontinued after 10 consecutive errors. A point was awarded for each correct response cumulating in a raw score. This Minna Moffatt-Feldman

38 was then converted into a standardised score by using age appropriate tables within the WRAT4 manual. 2. Administration of the standard WRAT4 Spelling Test (Green form) The spelling sub-test had 2 sections, 13 dictated letters and 42 one-word spellings, which increase in difficulty. It is not usual to include part 1 for English speaking adults, however for the purpose of this research it was included. This was because when the modified test was conducted it was important to establish whether the adults could produce the signed letters and for continuity between both. The target word was dictated to the participant by the PATOSS assessor, then presented in a sentence to provide context before being read again, after which the participants wrote down the words in English. When the list was completed they were asked to present the words to the camera for checking. A point was awarded for each correct response cumulating in a raw score. This was then converted into a standardised score by using age appropriate tables within the WRAT4 manual. 3. Administration of the modified WRAT4 Single Word Reading Test (Blue form) The 15 letters and 55 English words were all manually produced by the researcher in A/BSL as required. The participants were asked to then verbally confirm the letters and words. Participants were permitted 10 seconds to respond and were allowed to request repetitions as many times as the time period allows, after which the participants were asked to move to the next word. The researcher was careful not to elicit whether the words were correct or incorrect so as not to create any anxiety for the participant. A point was awarded for each correct response cumulating in a raw score. This was converted into a standardised Minna Moffatt-Feldman

39 score by using age appropriate tables within the WRAT4 manual. It is acknowledged that standardised scores in this instance would not be a proven standard result due to the test being modified. Scores would simply act as quantitative guides to compare performance between the standard and modified tests. At the end of the assessment the participants were asked to repeat the Single Word Reading test of the blue WRAT4 in English for a comparison. It was also recognised that the participants had effectively seen the words twice, one in SL, then again in written English. This result cannot be relied on due to the participants already being exposed to the vocabulary during the signed Single Word Reading test. Results were of qualitative interest to see how the participants processed the vocabulary with all letters in vision simultaneously. It was therefore expected that their score would improve given their second reading. 4. Administration the modified WRAT4 Spelling Test (Blue form) The 13 dictated letters were manually produced by the participants in their specific SL. Participant s productions were assessed for errors by the researcher and results were noted by the PATOSS assessor. Participants were permitted to self-correct. The 42 one-word spellings were presented in the same way. The target word was dictated, presented in a sentence to provide context and then repeated again. A point was awarded for each correct production of the target word using the correct manual letter formations. The researcher was careful not to elicit whether the words were correct or incorrect so as not to create any anxiety for the participant. The raw score was then converted into a standardised score by using age appropriate tables within the WRAT4 manual. It is acknowledged that standardised scores in this instance would not be a proven standard result due to the test Minna Moffatt-Feldman

40 being modified. Scores simply acted as quantitative guides to compare performance between the standard and modified tests Informal Interviews Care should be taken with vulnerable adults and young people at the end of the testing procedure, especially as it can elicit negative memories and language anxieties. For this reason a few minutes were taken to converse with the participants afterwards to ensure emotional stability and also to attain feedback over the assessments. The original plan for participants completing the WRAT4 assessments was that 50% would be formally interviewed. After the assessments, participants were asked to complete a more in-depth questionnaire (Appendix 8.8) that specifically related to the informal interviews. These responses were also used during coding along with the researcher s memorandums to ensure appropriately interpreted responses; however only four out of seven were returned. Bland (2008) explains that the length of questionnaires can negativity impact on the proportion of returned responses and on reflection the post assessment questionnaire included some questions already covered in the postal questionnaire and discussed after assessment and therefore could have deterred participants returning them. Additionally, these post assessment questions were not anonymous which also may have impacted the return rate (ibid) Memorandums The use of memorandums enabled ideas and concepts to be noted down as they occurred and later were used to support analysis and theme identification. Given the pioneering nature of the research it was also felt that a stringent procedure could have led to important themes such as diamonds in the rough being missed (Corbin and Strauss, 2008, p.67). Although GTM was not strictly followed given the preconceived research questions, the Minna Moffatt-Feldman

41 method of analysing data was viewed as an appropriate way of establishing themes. There were some concerns about the sample size being small but recruiting candidates from the USA was far more difficult than anticipated despite various advertising attempts; all those who made themselves known to the researcher were used. Phenomenological research does not need to focus on sample sizes, rather the themes that emerge from the lived experience; this however was difficult with a population of two although technically possible. Discourse with individuals can be the most effective method of qualitative data collection enabling the exploration of attitudes, behaviour and experiences (Dawson, 2009). Asking the right questions is an important skill for any researcher, and therefore it is important for researchers to know the topic area adequately and pre-plan questions to elicit the most detail. Research questions were carefully considered, refined and piloted before distribution. Discourse during research was both face-to-face (via Skype ) and by means of written communication. Asking people to commit answers into writing can support qualitative data collection although it is possible to record and transcribe as it can easily be re-visited for analysis and comparison. Given the native language difference between researcher and participants (English/SL), this was deemed the most accurate way of recording data whilst not causing extra anxiety to the participants. This is especially important when following phenomenological research as theme-spotting can easily be tracked on paper. Researchers who create memorandums from impromptu discussions also add opportunity for comparison. Quantitative research generates statistics, which may be used to support qualitative data (Dawson, 2009) but requires different methods on analysis and rarely elicits personal experiences through the data. Typically, a large sample size is needed to look for themes and abnormalities. Structured or semi-structured interviews are other effective ways of elucidating perspectives. Open questioning always gives the opportunity to bring together multiple Minna Moffatt-Feldman

42 viewpoints, which can be used as a basis of reflection to find a way of understanding why different approaches work (Cook, 2009). It is important to allow interviewees to speak for themselves and not influence their responses with leading questions, thus introducing bias (McNiff et.al., 2003). Oppenheim (1992) argues that any change in wording, context and emphasis undermines reliability of interviewing as it ceases to be the exact same question for each respondent. In this research the participants were asked pre-prepared questions (Appendix 8.9) immediately after assessment as a way of ensuring they were unaffected by the assessment experience. Participants were asked to commit answers to text and submit formally. 3.7 Ethical Issues Within action research there are many areas of ethics to consider regarding the researchers behaviour towards the subject of the research and those affected by it (Grey, 2006). Although some may be considered common sense, other aspects of ethics may be vague between the dichotomies of right and wrong (Korac-Kakabadse et. al., 2002). Confidentiality must be maintained throughout and access from the appropriate bodies, stakeholders and participants sought together with permission to withdraw at any point (McNiff et. al., 2003). Researchers may unwittingly put pressure on individuals to take part due to the eagerness to progress within the research (Korac-Kakabadse et. al., 2002), an ethical researcher must be self-monitoring in order to avoid this. At all times strict ethical guidelines as laid out by the British Educational Research Association (British Educational Research Association, 2015) were adhered to, together with gaining appropriate ethical approval from the participating University of Bolton (2014) (Appendix 8.3). Information was provided to all participants in English seeking permission and included a statement about confidentiality, which was both in the advertisement for recruiting participants and on the questionnaire submission page (Appendices 8.5 and 8.6). Minna Moffatt-Feldman

43 Advertising through a SL video was considered but in doing so could have deterred learners of SL who could not fully comprehend the information. Research also suggests that very few deaf individuals are diagnosed as having dyslexia given the issues with recognising sounds (Herman et. al, 2014). The questionnaires were set up through a website; submissions were automatically ed to the researcher through the website s hosting company which did not capture any personal data from the respondents in order to retain anonymity. Regarding, Skype assessments, confidentiality was maintained through requesting that participants transmitted in a room where they could not be seen, overheard or disturbed and were asked to subsequently destroy their written contributions in order to protect the content of the WRAT4 test; this was confirmed with participants at the start of Skype sessions. Additionally, the assessor adhered to her professional code of ethics, which includes confidentiality of participant information, for which written confirmation was acquired (PATOSS, 2015), (Appendix 8.12). Confidentiality was further maintained by referring to participants by code including reference of country of origin; B1, A1, B2, A2 B meaning BSL user and A meaning ASL user. The country of origin and the age of the participant are the only personal information held due to the need for discrimination of signed languages and translating data in the WRAT4 scales. Minna Moffatt-Feldman

44 3.8 Piloting the Research Piloting the WRAT4 tests was invaluable in that many teething problems were eliminated. There were some difficulties keeping track of scores achieved in each stage of testing so the way forward was to simply used colour sheets that categorised the data to reflect the colour of the test sheets. When participants were assessed via Skype the WRAT4 sheets could not be disseminated due to strict testing and copyright regulations. The English single word-reading test was presented in the form of flash cards (green form) to the webcam to enable the participant to read the words aloud. A computer and Microsoft Word was originally used to mark down fingerspelling patterns but the software s auto-correct spell check kept changing the entered letters. For this reason the software choice was switched to Microsoft Excel which does not auto-correct spellings. Regarding the spelling test, piloting helped to overcome some of the logistics in assessing spelling via Skype. Participants were asked to take a sheet of paper and number it The assessor presented the spelling test in the same format but the participants wrote down the words and upon completion were asked to hold their answers to the webcam in order for them to be marked. To protect the format and requirement of the test, they were then asked to destroy their answer sheet. The first pilot was conducted using BSL fingerspelling with no accompanying English lippattern although this is not how it is presented in real-life. The next piloted participant was presented BSL fingerspelling along with English lip-pattern reflecting a real-life situation. It was decided to continue this through to the formal research in order to reflect the natural production of fingerspelling. Minna Moffatt-Feldman

45 3.9 Data Analysis Method Remenyi (2014) explains that the central theme of Grounded Theory is that the researcher must decide how to group data (code) in order to produce meaningful results. Thematic analysis looks for patterned examples of data, which may be grouped together as themes (Braun and Clarke, 2006). This philosophy was carried forward in this research along with the search for frequencies of each code (Scott, 2015) throughout the entire project. The researcher coded incidents relevant to the research question from conversations with participants, posts read on social media by participants and other research papers that have been read. The aim of this was to create descriptive, comparative and relational statements through which results could be displayed (Scott, 2015). Coding is an important part of research analysis but care should be taken to ensure that results are viewed in a holistic way which is why creating memorandums was an important part of the data collection (Remenyi, 2014). Braun and Clarke (2006) explain that Thematic Analysis is a widely-used qualitative analytic method but is rarely-acknowledged nor explained. It allows for key themes to be extrapolated from the data set whilst creating a path for researchers to follow during analysis. Thematic analysis describes a six-step process for analysing descriptive qualitative data (Braun and Clarke, 2006). This procedure was used in reviewing data collected through the research period. The six steps are as follows: 1. Familiarising yourself with your data: Read all of the responses in order to acquire a feeling of the descriptions and understand them. 2. Generating initial codes: Return to each response and exact phrases, which pertain to the investigated phenomenon. 3. Searching for themes: this enables analysis to be refocused using only the codes now collected from the original data set. Minna Moffatt-Feldman

46 4. Reviewing themes: organise the formulated themes and remove any that do not have enough data to support them. 5. Defining and naming themes: define and further refine the themes for analysis ensuring the essence of meaning is captured. 6. Write up report: Using the themes to create an interpretative analysis and outcomes of the research. (Braun and Clarke, 2006). Research may change over time so using codes to categorise themes is important method that enabled changes to be tracked and considered. Researchers must take care to insure that codes are not so grand that meaningful data is lost or overlooked and similarly not too much detail that the researcher is subject to data-asphyxiation (Remenyi, 2014, p.19). The key to identifying themes is time and patience; rushing and can cause major differences in results although an over-extended time frame can cause researchers to lose sight of their objectives (ibid). By doing this, data was analysed to answer the first three research questions: 1. To what extent, if any, is dyslexia a disadvantage to learning sign language? 2. Do people with dyslexia experience the same 'language-anxieties' over their first language and sign language, and if so are they equal? 3. Fingerspelling - do people with dyslexia feel that understanding fingerspelling presents the same problems to them as understanding written words? Here we have two variables. The Independent variables are the use of A/BSL or English and the dependent variables are the single word reading and spelling scores achieved on the WRAT4 Test. It seems clear that a useful analysis tool was to use a correlation Minna Moffatt-Feldman

47 coefficient; in this case Pearson s Correlation Coefficient was chosen. This served to measure the strength, if any, of linear correlation between English and SL during the WRAT4 assessments (Stangroom, 2015). The control element of the testing was the scores achieved when conducted in English. In this case, variables were reading and spelling in English and separating in A/BSL fingerspelling to address research questions four and five: 4. Are there any significant differences between single word reading and spelling comprehension of those with dyslexia using English or Sign Language? 5. Are there any significant differences between single word reading and spelling comprehension of those with dyslexia using the Two-handed British or One-Handed American fingerspelling systems? For these two questions it was assumed that the null hypotheses is: Question 4: H0 = There is no significant relationship between single word reading and spelling comprehension of those with dyslexia using English or Sign Language. H 1 = There is a significant relationship between single word reading and spelling comprehension of those with dyslexia using English or Sign Language. Significance level was set at 0.05 (95%). Question 5: H0 = There is no significant relationship between single word reading and spelling comprehension of those with dyslexia using the ASL or BSL fingerspelling systems. H 1 = There is a significant relationship between single word reading and spelling comprehension of those with dyslexia using the ASL or BSL fingerspelling systems. Significance level was set at 0.05 (95%). To test for a significant relationship as per the hypotheses laid out above, the data collected was analysed using Pearson s Correlation Coefficient. This measures linear association Minna Moffatt-Feldman

48 between two variables, where a result of r = 1 implies a perfect positive correlation and the result r = -1 implies a perfect negative correlation (Stangroom, 2015). This is then tested for significance by means of comparing the p-value (the probability of the null hypothesis (H 0) occurring by chance) with the set significance level (Fenton and Neil, 2012). In this case this is 95% and therefore the p value is compared to 0.05 and if it is greater than 0.05 then the result will not be significant and therefore H 1 will be rejected in favour of H Credibility and Validity Researchers often struggle to validate qualitative research due to its subjective nature with some preferring statistic figures, which may be manipulated to elicit meaning. Researchers bring with them their own ideals and preconceptions, which could influence results that must be avoided at all costs (Dawson, 2009). The results can only truly represent those involved within the sample and not be translated for a larger group (Scott, 2015). In contrast quantitative data collection requires the researcher to demonstrate that their method of theme extrapolation was appropriate (Dawson, 2009). Care must be taken to ensure that calculations are produced correctly as errors will affect the outcome of the research. For this reason the Pearson s Correlation Coefficient calculations were all completed using an online tool created by Stangroom (2015). This allowed for automated calculations and thus removed elements of human error from calculations. Minna Moffatt-Feldman

49 4.0 Results and Findings Through qualitative questioning and observations subjective themes were identified; these can be divided into the following five themes. Percentages indicate proportion of respondents whose answer matched the theme. Figure 1: Age of Respondents Respondents' Ages (years) 14.29% 14.29% 22.86% 48.57% Figure 2: Time Periods Other themes linked to time periods have been identified as follows: Reference Trend Time Periods a The average age of participants was 34.2 years old, with the age ranges spanning 34 years. The average age participants received a diagnosis of dyslexia was 21.3 years old, with the b diagnoses ages spanning 44 years. c 68.57% of participants received the diagnosis of dyslexia post millennium ( ) % of participants received the diagnosis of dyslexia after they had completed d compulsory education. There was no evidence of any trend surrounding the length of time participants had been e learning SL. Minna Moffatt-Feldman

50 Figure 3: Writing Hands Figure 4: Distinguishing Left from Right Minna Moffatt-Feldman

51 Figure 5: Learning to Fingerspell Right Handers Figure 6: Learning to Fingerspell Left Handers Minna Moffatt-Feldman

52 Figure 7: Fingerspelling in Left-handed Respondents Figure 8: Hand Dominance Other themes linked to hand dominance were identified as follows: Hand Dominance Reference Trend The participants who reported to switch dominant hand during fingerspelling all wrote with a their right hand. b No ASL users reported that they switched dominant hands during fingerspelling. Figure 9: Visual Processing Emerging themes linked to visual processing were identified as follows: Visual Processing Reference Trend a 85.71% of participants preferred to use their visual memory % of participants claim that dyslexia has had a positive impact on their learning b experience of SL due to the visual nature of the language % of participants felt the issue with the reception of fingerspelling was that letters are c viewed consecutively rather than simultaneously as in written text reading. Minna Moffatt-Feldman

53 Figure 10: WRAT4 Single Word Reading English test (Green Form) Reading English Participant Age Raw Score Standardised Score Percentile WRAT4 Classification of Achievement B th Average B th Low B th Average B th Average B th Average A rd Average A th Superior Figure 11: WRAT4 English Additional Single Word Reading test (Blue Form) Participant Age Raw Score Additional Reading English Approximated Standardised Score using WRAT4 tables Percentile WRAT4 Classification of Achievement B th Above Average B th Below Average B th Average B rd Average B th Average A th Upper Extreme A th Average As explained in Section 3.2 standardisation for SL scores are not proven to be standardised. Minna Moffatt-Feldman

54 Figure 12: WRAT4 English Spelling test (Green Form) Spelling English Participant Age Raw Score Standardised Score Percentile WRAT4 Classification of Achievement B th Below Average B th Low B th Average B th Low B th Average A rd Average A Below 23 rd Average Figure 13: WRAT4 English Averages and Range Reading English Spelling English Average Range Average Range UK USA Figure 14: WRAT4 Additional English Averages and Range Additional Reading English Average Range UK USA Minna Moffatt-Feldman

55 Figure 15: Perceptions of Impact on English Emerging themes about the perceptions of dyslexia on the English language were identified as follows: Reference a b c d e f g h i j Impact of Dyslexia on English Themes 46.67% of respondents stated that prior to learning SL their knowledge of phonics was poor % of respondents stated that their knowledge of phonics has improved since learning SL % of respondents stated that their knowledge of phonics remained the same since learning SL % of respondents stated that they preferred the use of English over SL % of respondents stated that they read for pleasure % of respondents stated that they felt dyslexia had an impact on their English ability. 51.4% of respondents stated that they felt dyslexia had a negative impact on their English ability. 80% of respondents stated that reading and/or spelling in front of others caused a negative emotional response % of respondents stated that they felt notable differences between themselves and their peers when using English. 60% of respondents stated that reading written English words is easier than reading fingerspelling. Figure 16: WRAT4 SL Single Word Reading test (Blue Form) Participant Age Raw Score Approximated Standardised Score using WRAT4 tables Reading SL B th B <1st Approximated Percentile WRAT4 Classification of Achievement Below Average Lower Extreme B th Low B th Low B th Low A st A rd Below Average Below Average As explained in Section 3.2 standardisation for SL scores are not proven to be standardised. Minna Moffatt-Feldman

56 Figure 17: WRAT4 SL Spelling test (Blue Form) Participant Age Raw Score Approximated Standardised Score using WRAT4 tables Spelling SL B st B st Approximated Percentile WRAT4 Classification of Achievement Below Average Lower Extreme B th Average B th Low B th Average A st Average A rd Below Average As explained in Section 3.2 standardisation for SL scores are not proven to be standardised. Figure 18: WRAT4 SL Averages and Range Reading SL Spelling SL Average Range Average Range BSL ASL Average standardised results and the range in the WRAT4 blue form SL tests. Minna Moffatt-Feldman

57 Figure 19: Perceptions of Impact of Sign Language Emerging themes linked to perceptions of dyslexia on SL have been identified as follows: Reference a b c d e f g h i j k l m n o p Impact of Dyslexia on SL Themes 82.85% of respondents stated that SL created a positive emotional response. 80% of respondents stated that they use SL for Pleasure % of respondents stated that they preferred to use SL over the English language % of respondents stated that they felt their confidence had improved since learning SL % of respondents stated that they enjoyed the structure SL % of respondents stated that felt that dyslexia had an impact on learning SL 34.29% of respondents stated that felt dyslexia has had a positive impact on learning SL 33.33% of respondents stated that they felt that dyslexia had enabled them to learn SL quickly % of respondents stated that they felt dyslexia had a negative impact on their fingerspelling ability % of respondents stated that they felt a notable difference between themselves when using SL % of respondents stated that they felt a difference in the speed of learning between themselves and their peers % of respondents stated that they felt that they needed to work harder and the learning process has taken longer in comparison with their peers % of respondents stated that fingerspelling (production and reception) in front of peers caused a negative emotional response. 40% of respondents stated that they felt no emotional response to fingerspelling in front of peers because they had previously disclosed their dyslexia diagnosis. 60% of respondents stated that they felt reading written English words is easier than reading fingerspelt words. 100% of the BSL assessment participants said their issue with reception of fingerspelling was comprehension of vowels Minna Moffatt-Feldman

58 Figure 20: WRAT4 Single Word Reading Test - Raw Scores Raw reading scores in English and SL highlights a line of best fit that shows positive correlation between the two. Figure 21: WRAT4 Signle Word Spelling Test - Raw Scores Raw spelling scores in English and SL, highlighting a line of best fit, which shows a positive correlation between the two.. Minna Moffatt-Feldman

59 Figure 22: WRAT4 Single Word Reading English Only Raw Scores This graph shows the raw reading scores in English on the blue and green forms. There is a line of best fit, which shows a positive correlation between the two. Figure 23: WRAT4 Single Word Reading Percentile Points Participant Age Single word reading WRAT 4 English Percentile point SL Percentile Point Difference in Percentile points B B <1 4 B B B A A Average This chart shows the percentile points given by the standardised scores for single word reading English and SL and the difference in percentile points for each participant. As explained in Section 3.2 standardisation for SL scores have not proven to be standardised. Minna Moffatt-Feldman

60 Figure 24: WRAT4 Single Word Spelling Percentile Points Participant Age Single word spelling WRAT 4 English Percentile point SL Percentile Point Difference in Percentile points B B B B B A A Average This chart shows the percentile points given by the standardised scores for single word spelling English and SL and the difference in percentile points for each participant. As explained in Section 3.2 standardisation for SL scores have not proven to be standardised. Figure 25: WRAT4 Single Words Reading English Only Percentile Points Participant Age English Percentile point Reading English WRAT4 English Percentile Point - 'Second' reading Difference in Percentile points B B B B B A A Average This chart shows the percentile points given for single word reading in English of the green form and then the second reading of the blue form, plus the difference in percentile points for each participant. Minna Moffatt-Feldman

61 Figure 26: Pearson s Correlation Coefficient Calculations a b c d e f Reading RS Reading SS Spelling RS Spelling SS 2nd English Reading RS 2nd English Reading SS X - Green Form Y- Blue Form x y µ:x µ:y SS x = (X - M x) 2 SS y = (Y - M y) 2 (X - M x) * (Y - M y) Minna Moffatt-Feldman r = ((X - M y) * (Y - M x)) / ((SS x)(ss y )) r 2 coefficient of determination Variability % that Y is linked to X p value Significant at 95% English SL % YES YES English SL % YES NO English SL % YES YES English SL % YES NO English English % YES YES English English % NO NO Key to symbols used in data calculations: r: Pearson s Correlation Coefficient R 2 : The coefficient of determination. α: significance level p: The p-value is the probability of obtaining a result equal to or more extreme than what was actually observed. RS: Raw Scores SS: Standard Scores (Stangroom, 2015) X: X Values Y:Y Values Mx: Mean of X Values My: Mean of Y Values X - Mx & Y - My: Deviation scores (X - Mx) 2 & (Y - My) 2 : Deviation Squared (X - Mx)(Y - My): Product of Deviation Scores Significant at 99%

62 5.0 Analysis and discussion No matter the volume or quality of data collected, theories do not jump to the attention of researchers without substantial analysis (Remenyi, 2014). The mixed approach to data collection during this research gives way for subjective data to be supported by quantifiable data. Although the two data sets aimed to look into different research questions, all are linked by the research project as a whole. It is interesting to see that the respondent average age was 34.2 years, the majority sitting within the year age mark, the average being skewed by the 34 year range of ages (Figure 1). Over half of the participants cited a dyslexia diagnosis post compulsory education with 68.57% post the year 2000, which could account for the average age being a strikingly low 34.2 years (Figures 2:c and 2:a). Genie Networks, a centre offering Signature examinations in Greater Manchester, state that learners attending their centre range in age from 0 to 79 (Genie Networks, 2015). In some cases students choose not to disclose their dyslexia for fear of being labelled or treated differently. There is a likelihood that the age of participants in the study was low because older SL learners may be unaware that they have dyslexia. The age range of assessment participants was only 11 years, which is relatively small given the questionnaire respondents age range spanned 34 years (Figure 2:a) % of participants received a diagnosis of dyslexia after they had completed compulsory education (Figure 2:d) (likely due to lack of funding and resources), which could lead one to assert that the individuals with dyslexia who fall into these age brackets have previously slipped through the net and unaware that they have dyslexia and thus did not come forward to take part. It would be interesting to conduct a study with older learners given the decline in BSL fingerspelling usage which has occurred in the last generation, a higher prevalence Minna Moffatt-Feldman

63 of fingerspelling exists with older generations but not with younger (Sutton-Spence and Woll, 1998). 5.1 Hand Dominance From the 35 Participants, 25% were sinistral (left handed) writers and 75% were dextral (right handed) writers (Figure 3). This is strikingly large given the national average of lefthanders is 10% (Live Science, 2012). It is wondered if the proportion of those learning SL who are left-handed reflects the population average or this is skewed to include more lefthanders too. Some argue that the proportion of left and right handed individuals with dyslexia are skewed in favour of left handers, whereas others report a roughly equal distribution (The Learning Solution, 2015), both views are rather interesting given the 10% in general population (Live Science, 2012). This proportion however is not reflected in the findings of this research (Figure 3). It is interesting how many sinistral students arrive in intermediate classes signing as if they are dextral. In the researcher s experience, their responses when questioned are always that it is easier to learn the same as their peers. A total of 28.57% of the sinistral participants signed as if dextral (Figure 7), which is high considering it is the non-natural method. Only 17.86% of respondents claimed to have a problem with switching hands during fingerspelling but about half (51.43%) cited general issues distinguishing their left and right hands (Figure 4), a common trait in dyslexia (Hope, 2008). One possible explanation for this much lower proportion is that fingerspelling is continuously one-hand dominant in BSL Minna Moffatt-Feldman

64 and only uses one hand in ASL. This was reflected in the fact that none of the ASL users switched between hands during fingerspelling, as realistically expected. 5.2 Issues Surrounding Phonology The themes surrounding phonics are interesting given that a diagnostic element of dyslexia is a weak phonological awareness and that SL uses no sound yet a third of participants claimed that their knowledge of phonics improved since learning SL. The most plausible explanation for this is the use of mouth patterns in SL to add meaning to entity classifiers or to support fingerspelling (Sutton-Spence and Woll, 1998). Signers have reported not being able to imagine SL without some spoken mouth patterns (Brentari, 2001). The act of fingerspelling and producing the mouth pattern of the phoneme can be difficult for learners who tend to begin with lip patterns reflecting the name of the letters whilst manually producing the word, for example C.A.T rather the /k/æ/t/. This helps promote lip-reading and given that English is a second language to the majority of native Deaf SL users (Ladd, 2010) it is used as a method to support comprehension. In addition it is a requirement as laid out by the BSL examination board Signature (2014, p.7). Interestingly, Hope (2011) describes the difficulty that individuals with dyslexia have in understanding that letters written in capital and lower case are in fact the same letters. She explains that this deficit in phonetic discrimination can cause issues for some when reading individual letters. One of the most important results was that all of the BSL participants stated that vowels caused the most difficulties in comprehending words (Figure 19:p). Others went further adding that specific vowel combinations such as di-graphs and tri-graphs affected their reading ability and even more so when reading fingerspelling. After the BSL feedback the questions were asked to the ASL participants and although they confirmed a difficulty in this Minna Moffatt-Feldman

65 area, their responses differed in the level of frustration it caused. They sometimes had difficulty comprehending the relevant phoneme for the vowels internally, whereas the BSL participants cited that the speed and delivery of BSL fingerspelling affected their ability to hold onto all the letters. Specific vowels were sometimes missed, which left decoding words nearly impossible heightening their anxiety levels. Similar mistakes were also observed during the English reading assessments. Nearly half of the respondents (46.67%) stated that prior to learning SL their knowledge of phonics was poor (Figure 15:a) with some respondents reporting that they did not understand what phonics were and therefore could not answer the question. These two points are not surprising given their dyslexia diagnosis but as Figure 15:b shows, a third of respondents who clearly understood the term phonics believed their knowledge in this area had improved. This improvement could be a result of the use of phonology being separated from the written mode of language, with the only English related phonological input from the individual s inner-voice and response to the mouth pattern from the fingerspelling producer. This creates a link to Calvert et. al s (1998) statement that lip-reading causes the audio cortex of the brain to attempt to process information. The other important factor in learning SL is that learners are given extensive practice in the use of mouthing and recognising phonemes and hence stands to reason that their knowledge will improve. A little over half (53.33%) felt their knowledge of phonics remained the same since learning SL (Figure 15:c) with one respondent commenting as to the relevance of the question. This suggests that the respondent had not made the connection between the use of phonics when decoding fingerspelling. It would have been of interest to see this respondent s WRAT4 assessment, however questionnaires were all submitted anonymously with no way of tracing them back to individuals without publishing answers and breaching confidentiality. Minna Moffatt-Feldman

66 The questions surrounding phonics were included with a view of eliciting information of the cognitive process during fingerspelling production and reception. Specifically, were they sounding out the words using their inner voice or trying to remember the letters whilst simultaneously trying to lip-read the word? Of course this could not be explained to participants as this would have instantly affected their responses just as telling someone that their breathing is being observed, changes how they breathe. This method of using the internal speech generator was described in the research of Stein (Cited in Doyle, 2002, p ). Participants who took part in the WRAT4 assessment all commented on the use of lipreading in their decoding process which is just how a deaf person would decode the information; lip-reading being the primarily input and fingerspelling the second. Figure 9:c shows a theme of 30.77% of questionnaire respondents with the opinion that reading fingerspelling is difficult as letters (handshapes symbolising letters) are viewed consecutively rather than simultaneously. As those with dyslexia can exhibit symptoms of weak short-term and working memory and slow processing speeds (Miles, 1993) the issue of consecutiveness is understandable. From BSL teaching observation, letters that learners continuously produce incorrectly are C, F, G and X. These differ from the other letters in their phonology as seen here: Right hand variant ( 2013). Minna Moffatt-Feldman

67 C requires only one hand so leaves itself open to be produced with the non-dominant hand, especially for those who find it difficult to distinguish hands. In addition, the C produced with the right hand looks backwards for the producer and therefore seems sensible that the letter should be produced with the left hand rather than right hand. This, of course is incorrect given the rules on one hand dominance in SL but proposes a suggestion of why this is so prevalent in BSL learners (with and without dyslexia) and also why only right handed BSL users seem to produce this error. The signs for F, G and X are phonologically similar apart from palm orientation (which are only slightly different) and the handshapes used; the dominant hand should always be placed over the non-dominant hand. The confusion often lies by placing the non-dominant hand on top of the dominant hand rather than vice-versa. Interestingly those who reported (questionnaires) that they switched during fingerspelling were all right handed. This seems to reflect the typical right/left hand confusion amongst people with dyslexia and warrants future research. As previously stated, SL is a visual language with no audible element. The response of participants that 85.71% preferred visual input is a very strong theme (Figure 9:a). Even more interesting is that 91.43% reported they felt their dyslexia had a positive impact on their SL learning due to the visual nature of the language as a whole (Figure 9:b). This highlights the typical visual learning preferences of people with dyslexia (The Bright Hub, 2012) and explains why so many dyslexic learners seem to enjoy learning SL (Figure 19:b, 80%). Unfortunately, this cannot be translated into fingerspelling use, which, as previously stated is not technically part of SL, rather a feature to enable borrowing from English. Minna Moffatt-Feldman

68 5.3 American Sign Language and British Sign Language It was hoped during the initial stages of this research that a comparison could be made between the impact of dyslexia on ASL and BSL however due to problems recruiting ASL users this has not been possible. From the little information we do have it is possible to make some assertions but these would need substantial further research given only two ASL users took part in the research and clearly cannot represent the entire population. On first glance, the ASL results average higher in every part of the WRAT4 test but there are a few external factors may have caused this aside from the different fingerspelling method. One of the participants disclosed they had studied linguistics at university and is a doctorate holder, whilst the other disclosed that they came from a family of bibliophiles; with this information one would expect their vocabulary and spelling capability to be higher than the norm. There is also a case to be made for the different schooling systems in the UK and USA. The USA has high esteem placed on the achievement of their Standardised Test, much more so that the British equivalent. These factors could contribute to the high achievement in the English tests. With regards to the SL, the use of fingerspelling is more prevalent in ASL than in BSL resulting in more practice producing and receiving fingerspelling words as a natural part of SL discourse. Interestingly, Valli et.al (2002) describe that ASL allows a plethora of changes to fingerspelling structures to adhere to its own syntactical rules. It is not too far a leap to posit the idea that the acquisition of ASL fingerspelling requires far more work than that of BSL due to the additional rules students must learn in order to be proficient. Additionally, a large proportion of ASL fingerspelling patterns have become lexicalised producing new sign formations to follow to the internal rules of ASL manual spelling (ibid). This in turn suggests that those with dyslexia who master ASL fingerspelling may have a larger grasp of the production and reception of such in comparison to BSL users. Minna Moffatt-Feldman

69 Although it has not been possible to elicit themes linked to the different production style of ASL and BSL fingerspelling. Whilst ASL uses one hand and BSL uses two, another significance is the differences in vowel production. ASL uses five distinct handshapes to produce vowels (Appendix 8.1) whereas BSL uses the index finger of the dominant hand to point to the digits on the non-dominant hand, one for each vowel (Appendix 8.2). It is not difficult to recognise the potential for error in producing or understanding the vowels signed in BSL due to their similarity which adds another dimension for error for those using BSL. In BSL, Fingerspelling tends only to be used when there is no sign equivalent for the required meaning. The history and evolution of BSL demonstrates that fingerspelling patterns ultimately become replaced with established signs (Sutton-Spence and Woll, 1998). One example of such is the sign SON. Older generations tend to fingerspell this as standard s-o-n, whereas younger generations have adapted the sign BOY to be used to also mean SON. BSL does have some fingerspelling sequences as part of the core lexicon but these are few in comparison to ASL; for example the standard constrictions for months are fingerspelt, j-a-n for January, f-e-b for February etc. ASL and BSL both use single manual letter signs in some instances and as expected all participants achieved full parts for receptive and production of single manual letters. Fingerspelling in ASL tends to be used as an established construction rather than it being replaced by a sign. Valli et.al (2002) demonstrate this with the example w-h-a-t (p.63). Although, there is an established sign meaning WHAT, the fingerspelling production is also widely used and accepted as is. Minna Moffatt-Feldman

70 5.4 Emotional Responses The questionnaires and assessments all evidenced strong emotional responses, so much in fact that at the end of assessments it was clear that the participants were stressed and time was taken to attempt to bring them to a more settled emotion. Body language is an important and considerable part of SL and thus proficient signers are very good at reading body language. One participant commented that they felt as if they were back in school and that all the negative memories came back to them. A large proportion 80%, (Figure 15:h), described a negative emotional response towards reading and spelling English in front of others whereas only 57.14% (Figure 19:m) described a negative emotional response towards fingerspelling reception and production in front of others. However in general 82.85% of participants claimed a positive emotional response to SL (Figure 19:a) with 41.18% preferring to use SL over English (Figure 19:c). These statistics give an insight into the level of negativity that fingerspelling holds on its own yet this is not translated over to SL as a whole. Nijakowska (2010) explains that there exists a relationship between phonological proficiencies in L1 and subsequently learned languages which given the previously mentioned research by Calvert et. al (1998) could go some way in explaining why receptive fingerspelling is affected for those with dyslexia. Other interesting themes are that 57.14% of participants felt their general confidence had improved since learning SL (Figure 19:d) and that 80% sign for pleasure (Figure 19:b). The root of this could be that the use of SL does not hinge on the use of fingerspelling rather that fingerspelling is a small element that learners must reconcile themselves with but ultimately can develop coping strategies to avoid. Some participants cited favoured preferences for remembering signs in order to avoid fingerspelling. This is supported by 28.57% stating that they enjoyed the structure of SL (Figure 19:e). Interestingly 40% stated the reason they felt no emotional response to fingerspelling in front of peers was because they had previously disclosed their dyslexia diagnosis (Figure 19:n). This is interesting and leads one to consider if this acceptance from peers impacts on their confidence level, Minna Moffatt-Feldman

71 although 31.25% have reported that they felt a difference in the speed of learning between themselves and their peers (Figure 19:k) with 25.71% stating they needed to work harder than peers and the learning process was slower (Figure 19:l). These perceptions must also have an effect on the confidence of learners and has the potential to snow-ball into a bigger issue of anxiety. In essence SL is a very expressive language within which there is no room for shyness as without appropriately produced non-manual features meanings can be lost. 5.5 English or Sign Language A total of 71.43% felt that dyslexia has a negative impact on their fingerspelling ability (Figure 19:i) and 60% comment that reading English words was easier than reading fingerspelling (Figure 19:o) with 57.14% stating that using and receiving fingerspelling in front of peers created a negative emotional response (Figure 19:m). The results of the WRAT4 test show that every participant had a lower result for single word reading and spelling in SL than they did in English (Figures 10, 12, 16 and 17). As previously explained, the SL standardised scores and percentiles are only used as a guide given that the WRAT4 test was not designed nor has it been tested for SL use. However it does give us a basis for comparison given that individuals should achieve the same standard score regardless if they take the green or blue test in the WRAT4 test. This being true, any variant in score can be assumed to be caused by the use of SL given this is the only variant that has changed. The results present a strong case that there are large differences in results achieved in English and SL. The average of the difference in percentiles for single word reading is 37 percentile points but only two percentile points for single word spelling (Figures 23 and 24). In addition no participant achieved a score higher than below average for SL single word reading (Figure 16) as per the WRAT4 banding Minna Moffatt-Feldman

72 (Appendix 8.11). From this we can assert that dyslexia poses little impact on the production of fingerspelling but a substantial impact on the reception of fingerspelling. It should be noted that, although the percentile points are based on standard scores those for SL can only ever be an estimate given there has been no grand-scale testing or standardisation of ability in SL. If we hold them as following the same normal distribution as the English language results, the change in percentiles points is quite striking. This assumption however is quite a jump given the massive differences between written languages and fingerspelling comprehension, speed and processing. It could be argued that if a participant achieves result X in single word reading for English then the same participant should be able to achieve a correlating result Y in SL given the Pearson s Correlation Coefficient attained for raw scores. This argument is supported by the Pearson s Correlation Coefficient attained for the raw scores in single word reading which is r = with p < 0.01 (Figure 26:a). In other words there is a very strong positive correlation between the results with 99% chance that these results did not occur by chance and therefore are highly significant (Figure 20). In addition, the coefficient of determination (r 2 ) gives a value of 82.63% showing that the variability observed in the SL raw scores is directly linked to the English raw scores (Figure 26:a) (also termed as goodness of fit ). The results for the standard scores are slightly lower but, again, this was to be expected given the approximation made by adapting the English standardisation for SL. The Pearson s Correlation Coefficient dropped to ; still a strong correlation but is now only significant at the 95% level (Figure 26:b). This was the required significance as laid out in the initial stages of the research (Chapter 3). In addition the goodness of fit here is 70.71% (Figure 26:b); still a very positive result for this research. For single word spelling the results also prove to be significant at the 99% level for raw scores and 95% for standard scores. The Pearson s Correlation Coefficient for raw scores was (Figure 26:c), and that Pearson s Correlation Coefficient for the standard scores Minna Moffatt-Feldman

73 was (Figure 26:d). The goodness of fit was calculated for both at 88.21% and 80.34% respectively (Figures 26:c and 26:d). These Pearson s Correlation Coefficient results all contribute to answering the research questions about the impact of dyslexia on SL. It is clear that there is a strong positive correlation between the results of participants in English and in SL (Figure 26). It is also clear that participants found single word reading in SL significantly more challenging than in English (Figure 23) conversely the single word spelling results for English and SL were very similar (Figure 24). It must be recognised that participants B1 and B2 both attained a percentile point difference of far larger than the other participants in single word spelling. One may expect this if the test was successfully repeated on a larger scale the difference in percentile points may even drop closer to 1.4 (the average of the other scores). This massive drop in comprehension ability goes a long way to explaining why 78.13% of respondents reported feeling a notable difference between themselves and their peers when using SL (Figure 19:j). This however is lower than the percentage of those who reported a notable difference in the use of English, 91.43% (Figure 15:i). The difference here is likely due to the other areas of SL that those respondents feel confident with and positive about using. A third of participants claimed that they felt dyslexia had a positive impact on their learning of SL (34.29%) (Figure 19:g) with a third of respondents also stating that dyslexia enabled them to learn SL quickly (Figure 19:h). After the WRAT4 tests had been completed in English and SL, the participants were asked to read the words from the Blue WRAT4 sheet in English (these had previously been read in SL (Figure 11). The comparison between Green and Blue forms (Figure 25) shows an Minna Moffatt-Feldman

74 average change in result to be 9.71 percentile points. Figure 14 shows that the average standardised score has increased in comparison to Figure 13 but additionally the range of scores has increased. From this we can assert that the participants recognised the words previously produced via SL in such a way that helped them decode the words to vocalise them appropriately. Clearly this did not happen in participant A2 s case given they were the only participant to score lower on blue reading compared to the green so it is wondered if the results would change if repeated on a larger scale. Some participants stated that they felt familiar with the words which they could not understand during fingerspelling, but were unable to formulate them into intelligible words due to, in their opinion, short-term memory problems; again, this would be a good point for future research. Extraordinarily only 38.24% of participants reported as preferring to use English over SL (figure 15:d). Given that English is their first language this is much lower than expected and highlights the levels of anxiety faced by those with dyslexia when using English. This anxiety clearly does not continue over to SL given that 82.8% of participants stated SL gave a positive emotional response (Figure 19:a). Despite this, 77.14% still said they read for pleasure (Figure 15:e), which is interesting. One may assume that those with dyslexia shy away from the use of English but in this case clearly not. This percentage is not dissimilar from the 80% of respondents who said that they sign for pleasure (Figure 19:b). These themes show that the active use of SL or written English is not circumvented by those with dyslexia despite 94.29% stating they felt dyslexia had an impact on their English ability (Figure 15:f) and 91.43% felt the impact on SL (Figure 19:f). Interestingly, the proportion who felt a negative impact was lower for both; English reading and spelling 51.43% (Figure 15:g) and 71.43% reported a negative impact on SL fingerspelling (Figure:19i). The perception of a greater impact on SL could be expected given two factors. Firstly SL is a second language of which participants had been studying for various lengths of time. Those who felt that reading English words was easier than SL words was 60% (Figure 19:c). Given Minna Moffatt-Feldman

75 the previous results this number is interesting as a higher proportion of individuals felt their English was impacted in comparison to SL.. Secondly, fingerspelling is a small part of SL use, which enables a signer to borrow English words for which there is no sign such as a name. This means that if a person is reading or producing fingerspelling in a normal situation there are only a finite possible set of answers that fit into the context. For example, if Dorothy signs: [HELLO NAME I d-o-r-o-t-h-y] one would never expect the fingerspelt word to be anything other than a name. Similarly if Dorothy signs [m-o-s-t-o-n FROM I] one would automatically assume that the word fingerspelt was a geographical place name. Additionally, in support of these cues, the first letter fingerspelt is usually easily understood, again narrowing down the possible words being produced. Earlier, the two models of reading were discussed, data-driven and concept driven (Reid, 2003). From the examples above it could be argued that reception of fingerspelling uses a combination of the two models rather than only following one. Furthermore, the onlooker is able to view the (silent) mouth pattern of the word to support decoding the word produced. Some mouth patterns are easily distinguishable whereas others are not because the phonic it produces occurs in a part of the mouth which is not visible to the onlooker or the mouth pattern itself is similar to the pattern of another phoneme such as /m/, /b/ and /p/. It is also interesting to find that Calvert et. al (1998) state that looking on to a person who is speaking without voice not only activates the visual cortex of the brain but also the audio cortex despite no sound being heard. From this we can infer that the recognition of the phoneme rather that the name of a letter can improve understanding of receptive fingerspelling better. Equally one may ask, if this is true, then how do deaf people do it given that they may have no concept of what phonics are; research for another day. Minna Moffatt-Feldman

76 Analysis of the data pool has produced some interesting and thought provoking results. It would seem that future research would be necessary to qualify the quantitative data on a larger scale given the significances suggested from the results attained. It is felt that given the overwhelming proportion of respondents eliciting data describing similar phenomena, that repetition of a qualitative data collection would elicit similar data. It would be suggested that less focus is given to English language use and more given to SL as a whole rather than just fingerspelling. The questionnaire results also showed that participants feel dyslexia is not a disadvantage to learning SL as a whole but rather an advantage due to the visual nature and structure of the language. The disadvantages lay within the fingerspelling comprehension shown clearly in both the qualitative and quantitative results. A possible issue however, which has not been mentioned, is the possibility that learners with dyslexia had withdrawn from studying SL at the initial stages. Usually fingerspelling is learnt first as a foundation of communication through which signs maybe added into the learners lexicon. If a learner with dyslexia is overwhelmed by this use of fingerspelling and the pressure caused upon comprehension it could cause them to withdraw from learning or be overly anxious thus hindering language acquisition as per the Affective Filter Hypothesis posited by Dulay and Burt (1977). The avoidance of fingerspelling was mentioned by a few assessment participants citing that they favoured remembering signs in order to avoid repetitive fingerspelling, all these participants were using BSL. The level of emotional responses cited by participants were all strong; so much in fact that at the end of assessments there was visible evidence of elevated stress levels. If a learner is experiencing anxiety due to their environment they could perceive situations differently and thus their learning is negatively impacted. Influential social factors would differ according to environment (Ellis, 2003). Minna Moffatt-Feldman

77 If a learner is experiencing anxiety due to their environment they could perceive situations differently and thus their learning is negatively impacted. Influential social factors would differ according to milieu; classrooms versus a natural setting for instance (Ellis, 2003). Over three quarters of participants felt a difference between themselves and their peers whilst learning SL which is clear evidence of having the potential to cause anxiety; the WRAT4 results of reading comprehension demonstrates the massive difference between results. It should be noted that the seemingly small difference in percentile points seen in participant B2 is such as the individual already placed very low in the percentiles for English single word reading and subsequently dropped to the lowest level allowed by the scoring system. Questionnaire responses support the findings of the WRAT4 tests whereby 71.43% felt that dyslexia has a negative impact on their fingerspelling and 60% comment that English word reading is easier than reading fingerspelling. The two reasons that this research has found contributing to this is that fingerspelling is predominately different from reading text as the onlooker has the opportunity to see all letters in the word simultaneously to support the phonetic decoding of the words and not consecutively as in fingerspelling. In addition, the conflict between audio and visual input is not present during reading as it is during fingerspelling given the addition of mouth patterns, which support the phonic decoding of the letters as they are manually produced. This is supported by studies that show the brain usage of Deaf SL users presenting akin to hearing spoken language users (MacSweeney et. al, 2008). Additionally, the area used for understanding reading is the same as understanding fingerspelling (Waters et. al, 2007). The most striking results can be seen in Figures 12 and 13; the average difference in percentile points between English and SL single word reading is 37 percentile points but only an average of two percentile points for spelling. This massive drop in comprehension Minna Moffatt-Feldman

78 ability goes a long way to explaining why 78.13% of respondents reported feeling a notable difference between themselves and their peers when using SL and contributes to language anxieties. 5.6 Answering the Research Questions Question 1: To what extent, if any, is dyslexia a disadvantage to learning sign language? On the whole, dyslexia does not seem to pose a disadvantage to learning SL given the majority of participants citing a positive emotional response from SL and over 40% stating they preferred to use SL over English. However, having dyslexia is a clear disadvantage when trying to comprehend fingerspelling. This in turn can make the learning journey filled with anxiety shown by over half of respondents stating they had a negative emotional response from fingerspelling in front of peers. This however did not stop the participants from continuing with SL and 80% stated that they use SL for pleasure. From this we can assert that the disadvantage of learning SL would hinge on the use of fingerspelling by instructors when teaching. If however the process of language acquisition as described by (Krashen, 1987) was followed and learners were treated akin to child learners, fingerspelling use would be minimal. This is in the same manner that young children would be expected to learn new vocabulary, naturally from discourse before learning to spell and read. Additionally, it is important to note that 40% of respondents stated that informing peers of their dyslexia helped to allay any negative emotional responses. From this we can conclude that any disadvantage in learning SL is strongly linked to the quantity of fingerspelt vocabulary that individuals with dyslexia must comprehend before replacing fingerspelt words with an established sign. To help improve the process it would seem that informing peers and instructors would also help the experience. Minna Moffatt-Feldman

79 Question 2: Do people with dyslexia experience the same 'language-anxieties' over their first language and sign language, and if so are they equal? Over half of respondents stated that dyslexia caused a negative effect on their English ability with 80% citing that reading and spelling in front of peers caused a negative emotional response. This is highlighted in the participants WRAT4 scores given that the majority of results sat in the low and average bands. These results are still superior to the achievement in SL yet the respondent themes cited that only 57.14% felt a negative emotional response for using fingerspelling in front of peers with 40% stating they had no emotional response at all. This suggests that any language anxiety is less than those present in English language use. Additionally, a third of respondents cited a perception that dyslexia helped them learn quickly, which clearly would produce a positive response. In fact 85.85% stated that using SL created a positive emotional response and 57.14% stated an improvement in confidence. Confidence does not grow in the presence of anxiety and therefore we can assert that language anxieties in SL are lower than English but there is a vast amount of evidence suggesting the presence of anxieties in fingerspelling (especially receptive fingerspelling). This anxiety appears to be more than that seen in English reading and spelling, which is highlighted by 60% of respondents stating that reading English words is easier than reading fingerspelling and nearly the same stated a negative emotional response. From this it seems that SL itself is not the cause of the anxieties rather just the use of fingerspelling, which as explained in chapter 1 is not strictly part of SL. This could be compared to the difference between speaking and listening, and reading and writing. SL is an additional language to all participants and respondents yet 80% said that they use SL for pleasure compared to 77.14% who read for pleasure. The difference here is marginal Minna Moffatt-Feldman

80 and shows that dyslexia does not need to be a deterrent for SL or English use or comprehension. To directly answer the research question all the above must be considered together with the subjectivity of the participants, respondents and the researcher. It is concluded however that the language anxieties in English and SL are linked to the comprehension and production of spelt words (fingerspelt or written) in front of peers. The participants WRAT4 test results gives us reason to understand why. Both the use of English and SL give opportunities for pleasure for the respondents with over a third feeling that dyslexia actually gave them an advantage in language use. To consider if the language anxieties are equal is a difficult task given the dynamics and nature of the languages. Many of the participants spoke about the troublesome time they had as children in education and that they still carry this with them but have also developed strategies to cope better in situations. This is not the case with SL; classes are informal and optional with adults who are empathic to the needs of others and the learning of SL. This is demonstrated in the theme of 40% citing that they were not negatively affected by using fingerspelling in front of peers given their disclosure of dyslexia. It would seem that the learning of L2 is built upon the linguistic foundations instilled during the acquisition of L1 (LIM). Question 3: Reception of fingerspelling - do people with dyslexia feel that understanding fingerspelling presents the same problems to them as understanding written words At first glance the easy answer to this would be no. However, this comparison question was not posed to individuals directly, rather a series of questions were asked systematically to elicit answers. Respondents were asked questions that focused on phonological awareness, elements of the English language they found difficult and reading/spelling in front of peers; similar questions were asked about SL. It is clear from the WRAT4 tests and Minna Moffatt-Feldman

81 the questionnaire results that individuals strongly felt disadvantaged in the reception on fingerspelling, clearly demonstrated in every participant scoring lower in SL than they did in English. A common theme, which emerged throughout the research was that respondents and participants felt the difference between reading English text and reading SL fingerspelling was due to the consecutive nature of fingerspelling. The longer the words, the more information has to be remembered and held to gain the whole. The weaknesses in processing and working memory seen with dyslexia makes this a more challenging exercise. Having no audial input also means there are fewer pathways to the brain to store and recall all that is being conveyed. From results in questionnaires, respondents showed themes of having poor phonological awareness and feeling noticeable differences between themselves and their peers and that dyslexia has had a negative impact on their English reading/spelling ability. Conversely, respondents also stated that dyslexia had a positive impact on the SL ability but struggled with fingerspelling. Interesting that over 85% stated that they preferred to use their visual memory but it seems that it is this memory (together with their working memory and deficits in processing) that is letting them down when it comes to fingerspelling comprehension along with their deficits in knowledge of phonics and vowel patterns. Additionally, 60% of participants felt that comprehension of English words was easier than that of fingerspelt words. Although over half this figure seems low given the types of words generally fingerspelt (words that have no signs such as places or names). With regards to the WRAT4 assessments, participants were straight away put at a disadvantage given the words on the WRAT4 test included English words which do not directly translate into SL or Minna Moffatt-Feldman

82 are seldom used in general English discourse. Further research to elucidate strategies that individuals use when comprehending fingerspelling would be beneficial. It seems sensible to assume that as per the LIM, participants are trying to process the letters in English but the answers to these are outside the scope of this research Question 4: Are there any significant differences between single word reading comprehension and spelling in individuals with dyslexia whilst using English or sign language? Yes, there is a significant relationship between single word reading comprehension and spelling of those with dyslexia using English or Sign Language. Both the raw and standardised scores were significant at the set 95% level and in addition the raw scores were significant at the 99% level. The drop in single word reading WRAT4 percentile points averaged at 37 places. Spelling however showed less of a variation, averaging at a drop by 2 percentile points. This in itself is a significant finding showing that spelling ability is mainly affected. Generally, English spelling presented itself as an issue for respondents whereas the reception of fingerspelling was a specific, more often mentioned issue. Understandably so, given stark drop in comprehension ability highlight in the WRAT4 SL single word reading tests. This research faced some limitations from the offset given the traits of the researcher. Most of the participants were themselves right handed but the researcher was left handed, which put them at a disadvantage if they were not used to observing left handed fingerspelling. The question was never asked, although with hindsight should have been, which hand was dominant for the instructor and if they had had opportunity to practice with left and right handers. It would be interesting if research were undertaken focusing on comprehension of fingerspelling produced with the left and right hands. Minna Moffatt-Feldman

83 Additionally, the researcher is a BSL user and although has years of experience using ASL fingerspelling this does not compare to a native ASL user. This could have placed the ASL participants at a disadvantage so their comprehension results may have been higher if a native ASL user had produced the fingerspelling. Question 5: Are there any significant differences between single word reading comprehension and spelling of those with dyslexia using the two-handed British or onehanded American fingerspelling systems? It was unfortunate that in the time scale it was only possible to organise 2 tests with American participants, an instant issue with this proved to be the time zone difference plus perhaps a trust issue. Many individuals are wary about conversing with strangers, especially ones not on home territory. It was attempted, but proved difficult to try to involve educational institutions in the USA to recruit students. From the information that has been collected, at first glance it seems that there are differences in achievement in the WRAT4 test yet the data pool is not large enough to complete a Pearson s Correlation Coefficient. As per Chapter 5, it is not sure how representative these two participants were and substantially further research is required to answer this. Minna Moffatt-Feldman

84 6.0 Summary and conclusions This paper served to identify the impact dyslexia has on learning SL as a non-primary language. The study explored and highlighted the perceptions and experiences of people with dyslexia whilst learning SL with an additional focus of fingerspelling use within SL. The impression that dyslexia presents itself as a disadvantage in written language is reasonably well known but the impact on SL has attracted little research. It is well documented that learning additional languages brings any dyslexia type difficulties to the foreground, making second language acquisition difficult (Nijakowska, 2010). Despite this, there is very little, if any, previous research focusing on the transference of such language anxieties from first language to SL. Fingerspelling was thought to be the most akin to written language and so this was targeted for specific investigation. The research aimed to establish if there were any significant differences between the ability of individuals with dyslexia in single word reading comprehension and spelling via English and SL; additionally if there were differences between ASL and BSL. By using quantitative data collection methods the impact became measureable. Comparisons were made between English and SL using a standardised and modified WRAT4 test, often used by licensed professionals to ascertain the reading and spelling ability of individuals with specific learning difficulties. The available sample of ASL users was not equal to the sample size of BSL users, nor was it large enough to be representative of the population. Regardless, limited comparisons were drawn that provided enough insight to warrant further investigation. Minna Moffatt-Feldman

85 6.1 Implications of the study It has already been stated that dyslexia is a language processing disorder affecting the reading and writing of letters, numbers and symbols, to varying degrees (Schneider and Crombie, 2003). Whilst these symbols are generally understood to mean musical notes there is no evidence suggesting that this group could not include manually produced handshapes (signs). It has been explained that dyslexia exhibits issues due to the brain s conflict over processing auditory and visual information (ibid) and yet whilst SL is silent this can still be an issue. As Calvert et.al (1998) explained, when watching lip patterns, despite no sound being produced, both the visual and audio cortices of the brain were seen to activate. From this, it may be inferred that dyslexia becomes an issue during fingerspelling comprehension whilst simultaneously watching the mouth patterns of the signer, which serve to qualify homonym signs and nouns (Sutton-Spence and Woll, 1998). Although only the WRAT4 assessments were able to offer quantitative data with regards to fingerspelling comprehension, qualitative data was collected which supported the recognition of the deficit ability. The most striking results showed that the average difference in percentile place between English and SL single word reading was 37 percentile points but only an average of two percentile points for spelling. This massive drop in comprehension ability goes a long way to explaining why over three-quarters felt a notable difference between themselves and their peers when using SL and why it contributes to language anxieties. The BSL using participants all expressed that vowels presented specific problems during fingerspelling comprehension. The ASL participants went further and described issues specifically highlighting weaknesses in identifying di-graphs and tri-graphs. This would be a useful insight for further research to look at general sign comprehension of single signs and signed phrases. Minna Moffatt-Feldman

86 This finding also informs teachers to pay specific attention to teaching vowel combinations, remembering that people with dyslexia have weaknesses in this area. It is also essential that SL teachers have working knowledge regarding the effects of stress and anxiety on a person s performance, these must not be underestimated as they impair memory, reduce efficiency, affect clarity of thought and undermine reasoning skills (Bartlett and Moody, 2001). 6.2 Limitations of the Study Sample Selection The age of assessment participants only had a range of 11 years; relatively small given the questionnaire respondents age range spanned 35 years. A limitation to recruiting older participants could be the lack of older SL users not having a dyslexia diagnosis, or even being aware they could have it. From the data collected, one can assert that there is a subjective perception of disadvantage by those with dyslexia in the use of fingerspelling. ASL users were obviously very difficult to recruit. The late inclusion of the element of the research left time short to recruit these participants. It would have improved the reliability of the research to have more ASL users take part. BSL users were also more difficult to recruit, not because of a lack of interest but because many did not have a dyslexia diagnosis. In total only 35 respondents were recruited plus 7 participants; 50 respondents would have been preferable together with 12 participants. An even distribution between ASL and BSL users would have also been preferable. Minna Moffatt-Feldman

87 6.2.2 Data Analysis Pearson s Correlation Coefficient proved to be an effective way of processing/analysing the WRAT4 data. This enabled confidence intervals to be gained, demonstrating reliability. The WRAT4 test standardisation method, whilst beneficial in highlighting deficits in ability, is not tested or standardised for SL use and therefore was only used as a guide. Regardless, given the reasoning that SL was the only change in the variable, the WRAT4 test was deemed as an appropriate indication that participants faced greater difficulties whilst using SL. Baker et. al s (2002) qualitative four-stage research model was also used as a guide to aid the process of planning, being flexible and reformulating where necessary. This served to suit the research given the later inclusion of ASL as a variable Interviews Interviews would have undoubtedly elicited more information, as they allow deeper probing. Time restrictions, including work commitments, plus completing the WRAT4 assessments during the assessors given time frame, prevented further interviewing. Assessments took longer than anticipated especially as care had to be taken afterwards to ensure participants were emotionally calm. If repeated, it would be suggested that the WRAT4 tests and interviews be completed by different people or at different times. Minna Moffatt-Feldman

88 6.3 Dissemination of Findings There has already been a vast interest in the research from individuals on social media wanting to advance their knowledge base. The research and findings shall be disseminated in the following ways: Made accessible to those who took part in the research via and through Facebook, research website. Given the researcher is currently teaching SL, open discussions shall be held in the work place and with current learners with an option to present the findings on a wider scale, especially within the Deaf community. The research will be submitted for publication to allow greater access to the findings, either in an educational journal or through a self-publishing scheme now available online. The research will be disseminated to organisations such as The British Dyslexia Association, PATOSS and other specialist organisations such as The International Dyslexia Association whose work focuses on dyslexia, both in the UK and the USA. 6.4 Recommendations for Future Research Further research would be beneficial in the following ways: 1. Replicated with a larger sample size. 2. Data collection to be replicated by a right hand dominant SL user. 3. Further exploration of the types of mistakes made during fingerspelling comprehension and production. 4. Research to fully ascertain any differences between results of ASL and BSL users. Minna Moffatt-Feldman

89 6.5 Closing Remarks The aim of this descriptive study was to explore and highlight the perceptions and experiences of people with dyslexia whilst learning SL. Results from the qualitative and quantitative data pools have both served to highlight the anxieties faced by learners with dyslexia and quantify them as a true impact. It would seem that the critical theme identified is that SL users with dyslexia are heavily disadvantaged when it comes to the reception of fingerspelling in comparison to their ability in English but less so with fingerspelling production. It has also come to light that feelings of anxiety surrounding fingerspelling is somewhat allayed by the disclosure of dyslexia. Individuals with dyslexia who are currently learning, or considering learning SL should not be disheartened or feel that dyslexia will be a barrier. Preliminary research here suggests that with the exception of reading fingerspelling, many feel that having dyslexia has had a positive impact on their learning of SL. Minna Moffatt-Feldman

90 7.0 Bibliography Baby Centre. (2011) Baby Signing [online]. Available from: < [Accessed 25 th May 2015]. Barker, C., Pistrang, N. and Elliott, R. (2002) Research Methods in Clinical Psychology: An Introduction For Students And Practitioners [online] Hoboken, New Jersey: John Wiley and Sons, Inc. ebook Collection (EBSCOhost), EBSCOhost [Accessed 16 th June 2015]. Bartlett, D. and Moody, S. (2010) Dyslexia in the Workplace: an introductory guide. 2 nd ed. Oxford: Wiley-Blackwell. Bayley, R., Schembri, A. and Lucas, C. (2015) Variation and change in sign language. In: Schembri, A.C. and Lucas, C. (eds) (2015) Sociolinguistics and the Deaf Communities. Cambridge: Cambridge University Press. Pp Berke, J. (2010) Deaf Culture - Big D Small D Who is a Small D in the Deaf Community? [online]. Available from: < [Accessed 15 th May 2015]. Bernstein, L. and Auer, E. (2003) Speech Perception and Spoken Word Recognition. In: Marschark, M. and Spencer, P. (2005) Deaf Studies, Language and Education. New York: Oxford University Press Inc. pp Birks, M and Mills, J (2015) Grounded Theory: a practical guide. 4th ed. London: SAGE. Bland, M. (2008) Increasing Response Rates to Postal Questionnaires [online]. Available from: < [Accessed 28 th July 2015]. Blankfield, S. (2001) Think, problematic and costly? The dyslexic student on work placement. SKILL, issue 70, July, pp Braun, V. and Clarke, V. (2006) Using thematic analysis in psychology. Qualitative Research in Psychology [online]. 3 (2). pp Available from: < [Accessed 28 th July 2015]. Brennen, M. (2001) Making Borrowings Work in British Sign Language. In: Brentari, D (ed) Foreign Vocabulary in Sign Languages: A Cross-Linguistic Investigation of Word Formation. New Jersey: Lawrence Erlbaum Associates, Inc Publishers. pp British Dyslexia Association. (2015) Definitions [online]. Available from: < [Accessed 16 th June 2015]. British Educational Research Association. (2015) BERA [online]. Available from: < [Accessed 16 th June 2015]. Brotchie, J., Clark, E., Draper, J., Price, R. and Smith, P. (2008) Designing Healthcare Research: Study Guide. Milton Keynes: The Open University. BSLBT and Remark! (2011) The History of Deaf Education: Part 2 [online] Available from: < [Accessed 20 th June 2015]. Burm, J. (2015) The positive aspects of having dyslexia [online]. Available from: < [Accessed 28 th July 2015]. Minna Moffatt-Feldman

91 Calvert, G., Bullmore, E., Brammer, M., Campbell, R., Williams, S., McGuire, P., Woodruff, P., Iversen, S. and David, S. (1997) Activation of Auditory Cortex During Silent Lipreading. Science [online] 276 (5312), pp Available from: < [Accessed 9 th May 2015]. Cameron, L. and Murphy, J. (2007) Obtaining consent to participate in research: the issues involved in including people with a range of learning and communication disabilities. British Journal of Learning Disabilities, 35 (2), pp Cherry, K. (2013) Hawthorne Effect [online]. Available from: < [Accessed 15 th July 2015]. Chomsky, N (2015) The Architecture of Language. 12 th ed. New Delhi: Oxford University Press. Children of Deaf Parents UK. (2015) Children of Deaf Adults: PARENTING & COMMUNICATION ISSUES [online]. Available from: < 009.pdf> [Accessed 15 th July 2015]. Cohen, L., Manion, L. and Morrison, K. (2007) Research Methods in Education. 6th ed. London: Routledge. Colaizzi, P. (1978) Psychological research and the phenomenologist views it. In: Valle, R.S. and King, M. (eds). (1979) Existential Phenomenological Alternatives for Psychology. New York: Oxford University Press.. Coleman, J. (2015) Phonetics vs. Phonolog [online]. Available from: < [Accessed 28 th July 2015]. Cook, T. (2009) The purpose of mess in action research: building rigour though a messy turn. Educational Action Research [online]. 17:2, pp Available from: < [Accessed 9 th May 2015]. Corbin, J. and Strauss, A. (2008) Basics of Qualitative Research 3E. 3rd ed. London: SAGE. Cormier, K., Fenlon, J., Rentelis, R. and Schembri, A. (2015) Lexical frequency in British Sign Language conversation: A corpus-based approach [online]. Available from: < [Accessed 17 th July 2015]. Coughlan, S. (2015) Dyslexia not linked to eyesight, says study [online]. Available from: < [Accessed 28 th June 2015]. Crawford, S. (2013) HELP! My Child Isn't Reading Yet What Should I do? New York: Amazon. Crystal, D. (1980) A First Dictionary of Linguistics and Phonetics. London: André Deutsch. Cummins, J. and Swain, M. (1986) Bilingualism in Education. New York: Longman. Daly, J., Kellehear, A. and Gliksman, M. (1997) The public health researcher: A methodological approach. Melbourne, Australia: Oxford University Press. Dash, N. (2005) Selection of the Research Paradigm and Methodology [online]. Available from: < [Accessed 25 th July 2015]. Minna Moffatt-Feldman

92 Davis, W. (2010) The Simple Way to Understand Research Methods and Statistics in Psychology. United Kingdom: Amazon Kindle. Dawson, C. (2009) Introduction to Research Methods. 4th ed. Oxford: How to Books. Deaf Studies Research Unit, University of Durham (1992) Dictionary of British Sign Language / English. London: Faber and Faber Ltd. Delamont, S. (2002) Fieldwork In Educational Settings: Methods, Pitfalls And Perspectives [online]. London: Routledge. ebook Collection (EBSCOhost), EBSCOhost [Accessed 9 th May 2015]. Deuchar, M. (2013) British Sign Language. 5th ed. New York: Routledge. Doyle, J. (2002) Dyslexia: An Introductory Guide. 2nd ed. London: Whurr Publishers. Dulay, H., & Burt, M. (1977). Remarks on creativity in language acquisition. In M. Burt, H. Dulay, & M. Finocchiaro (Eds.), Viewpoints on English as a second language. New York: Regents. pp Dyslexia Action Shop. (2014) Wide Range Achievement Test 4 (WRT4) [online]. Available from: < [Accessed 25 th May 2015]. Ellis, R. (2003) The Study of Second Language Acquisition. 10th ed. Oxford: Oxford University Press. Emmorey, K. (2001) Language, Cognition, and the Brain: Insights from Sign Language Research. (2nd ed) [online] New Jersey: Lawrence Erlbaum Associates Inc. Available from < q&f=false> [Accessed date]. Equality Act 2010, c.2. London: HMSO. Eriksson, P. (1998) The History of Deaf People: A Source Book. Ӧrebro: Daufr. Feher-Prout, T. (1996) Stress and Coping in Families with Deaf Children. Journal of Deaf Studies and Deaf Education. 1 (3), pp Fenton, N. and Neil, M. (2012) Risk Assessment and Decision Analysis with Bayesian Networks [online]. Available from: < [Accessed 28 th July 2015]. Gass, S. & Selinker, L. (2000) Second Language Acquisition: An Introductory Course [online] Mahwah, New Jersey: Lawrence Erlbaum Associates, Inc. ebook Collection (EBSCOhost), EBSCOhost, [Accessed 28 th May 2015]. Genie Networks. (2015) Genie Communication Academy [online]. Available from: < [Accessed 9 th July 2015]. Goldin-Meadow, S. and Mayberry, R. (2001) How do profoundly Deaf Children Learn to read? Learning Disabilities Research and Practice. 16 (4), pp Grey, D. (2006) Doing Research in the Real World. 3rd ed. London: SAGE. Haigh, C. and Williamson, T. (2009) Research ethics: RCN guidance for nurses. 2nd ed. London: Royal College of Nursing. Minna Moffatt-Feldman

93 Hannah, A. (2007). Interviews in Education Research. Available from: [Accessed 31st July 2015] Hamann, C. (2014) Language Acquisition [online]. Available from: < sition.pdf> [Accessed 28 th July 2015]. Hatch, J.A. (2002) Doing Qualitative Research In Education Settings [online] Albany: State University of New York Press, ebook Collection (EBSCOhost), EBSCOhost, [Accessed 4 th June 2015]. Hawley, M. (2009) Laddering: A Research Interview Technique for Uncovering Core Values [online]. Available from: < [Accessed 28th July 2015]. Heidegger, M. (1962) Being and time (Translated by Maquarries, J. and Robinson, E.). Oxford: Blackwell. Herman, R., Roy, P. and Kyle, F. (2014) Reading, Dyslexia and Oral Deaf Children [online]. Available from: < data/assets/pdf_file/0014/212441/briefing-paper-reading-anddyslexia-in-oral-deaf-children-website-final-version-feb-2014.pdf> [Accessed 28 th July 2015]. Holloway, I. and Wheeler, S. (2010) Qualitative Research in Nursing and Healthcare. 3rd ed. London: Blackwell Publishing Ltd. Hope, K. (2008) A Trick to Help a Dyslexic Know Right from Left [online]. Available from: < [Accessed 9 th July 2015]. Hope, K. (2011) Dyslexic Preschoolers learn Capital Letters easier than Small Letters [online]. Available from: < [Accessed 28 th July 2015]. Hopwood, V. (2003) Deaf Children in hearing classrooms: teacher-pupil talk in secondary schools. In: Gallaway, C. and Young, A. (eds) Deafness and Education in the UK. Research Perspectives. London: Whurr Publishers Ltd. Husserl, E. (1962) Ideas: General introduction to pure phenomenology. New York: Collier. Iacono, T. & Murray, V. (2003) Issues of informed consent in conducting medical research involving people with intellectual disability. Journal of Applied Research in Intellectual Disability, 16 (1): pp Irlen. (2012) About Irlen Syndrome [online]. Available from: < [Accessed 10 th July 2015]. Irlen. (2015) Irlen Syndrome: Perceptual Problems [online]. Available from: < [Accessed 28 th June 2015]. Keane, J., Brentari, J. and Riggle, J. (2012) Coarticulation in ASL fingerspelling [online]. Available from: < [Accessed 17 th July 2015]. Knight, P. & Swanwick, R. (2002) Working with Deaf Pupils, Sign Bilingual Policy into Practice. London: David Fulton Publishers Ltd Minna Moffatt-Feldman

94 Korac-Kakabadse, N., Kakabadse, A. and Kouzmin, A. (2002) Ethical Considerations in Management Research: a "truth" seekers guide. In: Partington, D. (ed) Essential Skills for Management Research. London: SAGE. Krashen, S. (1987) Principals and practice in second language acquisition. 2nd ed. Hertfordshire: Prentice Hall Europe. Kyle, F. (2015) Studying Reading and Literacy Development in Deaf Children. In: Orfanidou, E., Woll, B. and Morgan, G. (eds) Research Methods in Sign Langauge. West Sussex: John Wiley and Sons Ltd Kyle, J. and Woll, B. (1985) Sign Language: The study of deaf people and their language. Cambridge: Cambridge University Press. Ladd, P. (2010) Understanding Deaf Culture, In Search of Deafhood. 7th ed. Clevedon: Multilingual Matters. Lado, R. (1957) Linguistics Across Cultures. Michigan: The University of Michigan Press Lane, H. (1987) Sign Language, American History. In: Van Cleve, J. (ed) Gallaudet Encyclopedia of deaf people and deafness. New York: McGraw-Hill Lang, H. (2003) Perceptive on the history of deaf education. In: Marschark, M. and Spencer, P. (eds) Deaf Studies, Language and Education. New York: Oxford University Press. Live Science Staff. (2012) Study Reveals Why Lefties Are Rare [online]. Available from: < [Accessed 5 th July 2015]. MacSweeney, M., Waters, D., Brammer, M., Woll, B. and Goswami, U. (2008) Phonological processing in deaf signers and the impact of age of first language acquisition. NeuroImage. 40 (3), pp Marshall, A. (2012) Positive Aspects of Dyslexia [online]. Available from: < [Accessed 28 th July 2015]. Mayer, C. and Wells, G. (1996). Journal of Deaf Studies and Deaf Education. 1 (2), p McLeod, S. (2008) Likert Scale [online]. Available from: < [Accessed 7 th April 2015]. McNiff, J., Lomax, O. and Whiltehead, J. (2003) You and Your Action Research Project. 2nd ed. Oxon: Routledge Falmer. McNiff, J. and Whitehead, J. (2005) Action Research for Teachers. A practical guide. London: David Fulton Publishing Ltd. Mental Capacity Act 2005, c.9. London: HMSO. Merriam, S. (1998) Qualitative Research Case Study Applications in Education. San Francisco: John Wiley and Sons, Inc. Miles, T. (1993) Dyslexia: The Pattern of Difficulties. 2 nd ed. London. Wurr Publishers. Mitchell, R. and Karchmer, M. (2004) Chasing the mythical ten percent: Parental hearing status of deaf and hard of hearing students in the United States. Sign Language Studies. 4(2), pp Minna Moffatt-Feldman

95 Moody, S. (2006) Dyslexia, How to survive and succeed at work. London: Vermilion. Morgan, M. (2009) Cross-Linguistic Typology of Argument Encoding in Sign Language Verbal Morphology. Paper presented at Association of Linguistic Typology, Berkeley Morse, J. (1991) Qualitative Nursing Research: A Contemporary Dialogue. London: Sage Publications. National Deaf Children's Society. (2010) Hands up for Help: Giving deaf children a fair chance at school. London: National Deaf Children's Society. Nijakowska, J. (2010) Dyslexia in the Foreign Language Classroom. Bristol: Multilingual Matters. O'Donoghue, T. & Punch, K. (2003) Qualitative Educational Research In Action: Doing And Reflecting. London: RoutledgeFalmer, ebook Collection (EBSCOhost), EBSCOhost, [Accessed 4 th June 2015]. Oppenheim, A. N. (1992) Questionnaire Design, Interviewing and Attitude Measurement. London: Pinter Publishers Limited. Padden, C. and Gunsauls, D. (2003) How the alphabet came to be used in a sign language. Sign Language Studies, (4). p Patton, M. (1990) Qualitative Evaluation Methods. California: SAGE. PATOSS. (2015) Code of ethics [online]. Available from: < [Accessed 16 th June 2015]. Pratt, J. (2012) The Problem of Grounded Theory [online]. Available from: < [Accessed 23 rd June 2015]. Ramus, F., Pidgeon, E. and Frith, U. (2002) The relationship between motor control and phonology in dyslexic children [online]. Available from: < [Accessed 30 th May 2015]. Reid, G. (2003) Dyslexia, a Practitioner's Handbook. 3rd ed. West Sussex: John Wiley and Sons Ltd. Rello, L. (2013) Good Fonts for Dyslexia [online]. Available from: < [Accessed 28 th July 2015]. Remenyi, D. (2014) Grounded Theory. 2nd ed. Reading: Academic Conferences and Publishing International Ltd Reynolds, C.R. and Fletcher-Janzen, E. (2007) Encyclopaedia of special education. 3 rd ed. Hoboken, New Jersey: John Wiley & Sons. Schneider, E. and Crombie, M. (2003) Dyslexia and Foreign Language Learning. London: David Fulton Publishers. Scott, G. (2015) The Researcher's Bible: An overview of key concepts and Methods in Social Science Research. United Kingdom: Amazon Kindle. Scriven, M. (1997) Truth and objectivity in evaluation. In: Chelimsky, E. and Shadish, W. (eds) Evaluation for the 21st century. London: Sage. pp Shield, A. & Meier, R. (2012) Author's personal copy Palm reversal errors in native-signing children with autism. Journal of Communication Disorders, 45 (6), pp Minna Moffatt-Feldman

96 Signature. (2014) Teacher Notes: Level 1 award in British Sign Language Sept 14 - Aug 15. Durham: Signature. Signature. (2015) Assessment Regulations [online]. Available from: < 102-and-103-Regulations-Sept pdf> [Accessed 28 th July 2015]. Sign Sounds. (2014) Is the Standard of English in deaf schools lower than those in mainstream schools? [online] Available from: < [Accessed 28 th July 2015]. Sivo, S., Saunders, C., Chang, Q. and Jiang, J. (2006) How Low Should You Go? Low Response Rates and the Validity of Inference in IS Questionnaire Research. Journal of the Association for Information Systems, 7 (6), pp Sociology Central. (2015) Questionnaires [online]. Available from: < [Accessed 16 th June 2015]. Sokolowski, R. (2008) Introduction to Phenomenology. New York: Cambridge University Press. Snowling, M. (1996) Developmental Dyslexia: An introduction and theoretical overview. In: Snowling, M. and Stackhouse, J. (eds) Dyslexia, Speech and Language. London: Whurr Publishers. Stangroom, J. (2015) Pearson Correlation Coefficient Calculator [online]. Available from: < [Accessed 25 th June 2015]. Starkiss, K. (2010) The Advantages of Being Dyslexic [online]. Available from: < [Accessed 28 th July 2015]. Stokoe, M (1991) Semantic Phonology. Sign Language Studies, 71. Pp Stokoe, W. (1960) Sign language structure, an outline of the visual communication systems of the American deaf. Studies in Linguistics, No 8. Buffalo NY: Department of Anthropology and Linguistics. State University of New York at Buffalo. Streubert, H. and Carpenter, D. (1995) Qualitative Research in Nursing: Advancing the Humanistic Imperative. Philadelphia: JB Lippincott Company. Sutton-Spence, R. and Woll, B. (1998) The Linguistics of British Sign Language. Cambridge: Cambridge University Press. The Bright Hub. (2012) Dyslexia and Visual Learning Strategies [online]. Available from: < [Accessed 28 th July 2015]. The Hearing Fund UK. (2015) Deaf Children: Statistics [online]. Available from: < [Accessed 15 th July 2015]. The Learning Solution. (2015) Common Misconceptions About Dyslexia [online]. Available from: < [Accessed 5 th June 2015]. The Makaton Charity. (2015) About Makaton [online]. Available from: < [Accessed 28 th July 2015] Urquhart, C. (2013) Grounded Theory for Qualitative Research. London: Sage Minna Moffatt-Feldman

97 Valli, C., Lucas, C and Mulrooney, J. (2002) Linguistics of American Sign Language: An Introduction. 3rd ed. Washington DC: Gallaudet Vickers, W. (2015) American Sign Language: Fingerspelling & Numbers: Introduction [online]. Available from: < [Accessed 14 th July 2015]. Vrooman, M. (2000) The Linguistic Interdependence Hypothesis and the language development of Yucatec Maya -Spanish bilingual children. Electronic Doctoral Dissertations for UMass Amherst. Paper AAI [online]. Available from: < [Accessed 25 th May 2015]. Warnock, H. (Chairman). (1978) Report of the Committee of Enquiry into the Education of Handicapped Children and Young People [online]. Available from: < [Accessed 28 th June 2015]. Waters, D., Campbell, R., Capek, C., Woll, B., David, A., McGuire, P., Brammer, M., and MacSweeney, M. (2007) Fingerspelling, signed language, text and picture processing in deaf native signers: The role of the mid-fusiform gyrus. NeuroImage, 35, pp Wilcox, S. and Perrin Wilcox, P. (1997) Learning to See: Teaching American Sign Language as a Second Language. 2nd ed. Washington DC: Gallaudet University Press. Wilkinson, G.S. and Robertson, G.J. (2006) WRAT4 Wide Range Achievement Test Professional Manual. Florida: Psychological Assessment Resources, Inc. Winstanley, C. (2010) Writing a Dissertation for Dummies. 2nd ed. West Sussex: John Wiley and Sons Ltd. Winzer, M.A. (1993) The history of special education: From isolation to inclusion. Washington, DC: Gallaudet University Press. Woll, B. & Ladd, P. (2003) Deaf Communities. In: Marschark, M. and Spencer, P. (eds) Deaf Studies, Language and Education. New York: Oxford University Press. Wundt, W. (1921) The Language of Guestures. The Hauge: Mouton. Minna Moffatt-Feldman

98 8.0 Appendices 8.1 American Sign Language Fingerspelling Right-hand variant of ASL fingerspelling (Vickers, 2015) Minna Moffatt-Feldman

99 8.2 British Sign Language Fingerspelling Right-hand variant of BSL fingerspelling ( 2013). Minna Moffatt-Feldman