Orgaizacija, Volume 47 Research papers Number 2, May 2014 DOI: 10.2478/orga-2014-0009 Exteded Techology Acceptace Model for SPSS Acceptace amog Sloveia Studets of Social Scieces Aleka Brezavšček, Petra Šparl, Aja Židaršič Uiversity of Maribor, Faculty of Orgaizatioal Scieces, Kidričeva cesta 55a, 4000 Kraj, Sloveia, aleka.brezavscek@fov.ui-mb.si, petra.sparl@fov.ui-mb.si, aja.zidarsic@fov.ui-mb.si Backgroud ad Purpose IBM SPSS Statistics is amog the most widely used programs for statistical aalysis i social scieces. Due to may practical values it is frequetly used as a tool for teachig statistical cocepts i may social sciece uiversity programs. I our opiio, motivatio to lear ad to use SPSS durig the studyig process plays a sigificat role i buildig a positive attitude towards SPSS which iflueces its usage at the professioal level after fiishig study. Desig/Methodology/Approach The aim of this paper is the developmet of the model for aalysig the acceptace of the SPSS amog uiversity studets of social scieces. The model is based o the widely kow Techology Acceptace Model (TAM). I additio to the traditioal compoets of the TAM, six exteral variables were icluded. The model is tested usig the web survey o the uiversity studets of social scieces from seve faculties at three Sloveia uiversities. Results The evaluatio of the questioaire was performed. Descriptive statistics were calculated. The depedecies amog the model compoets were studied ad the sigificat depedecies were poited out. Coclusio The results of the empirical study prove that all exteral variables cosidered i the model are relevat, ad directly ifluece both key compoets of the traditioal TAM,»Perceived Usefuless«ad»Perceived Ease of Use«. Therefore, our model is useful to study the adoptio ad cotiuous utilizatio of SPSS amog the studets of social scieces. The obtaied results are useful for educators, ad ca help them to improve the learig process. Keywords: uiversity educatio, statistics, statistics axiety, SPSS, acceptace, TAM model 1 Itroductio Courses i statistics are importat to all busiess majors because they represet the formal exposure to statistical aalyses ad research methods which may studets may fid useful i their careers. Aalytical skills ehace studets ability to read, iterpret, sythesize ad use reported results. O the other had, research productio skills eable studets to desig ad iitiate origial research (Ravid ad Leo, 1995). However, studets persoal experieces toward statistics are ofte a source of axiety producig egative perceptios. May researchers have idicated that courses i statistics are amog those that cause the most axiety, especially for studets i o-mathematicsorieted disciplies (e.g. Zeider, 1991; Baloğlu, 2003; Owuegbuzie, 2004; Pa ad Tag, 2004, DeVaey, 2010). Statistics axiety is experieced by as may as 80% of graduate studets i the social ad behavioural scieces ad is at least partly resposible for the procrastiatio of stu- Received: 6 th October 2013; revised: 27 th November 2013; accepted 25 th Jauary 2014 116
Orgaizacija, Volume 47 Research papers Number 2, May 2014 dets erollig i required statistics courses (Owuegbuzie ad Wilso, 2003). Our several years teachig experieces at higher educatio courses prove that the situatio at the Sloveia uiversities seems to be quite similar. We thik that studets motivatio for learig statistics ad usig statistical software is ot at a desirable level. The developmet ad psychometric properties of statistics axiety scales ad the factors affectig statistics axiety have bee extesively studied for more tha twety years, but few studies focused o how to reduce the statistics axiety for graduate studets i the social scieces. A comprehesive summary of the literature o statistics axiety is provided by Owuegbuzie ad Wilso (2003). The results of some previous studies idicate that usig computers i a statistics class has bee geerally successful i lowerig the statistical axiety (Stickels ad Dobbs, 2007). Amog the most widely used programs for statistical aalysis i social sciece is IBM SPSS Statistics. Itegratio of SPSS ito higher educatio courses i statistics ca have may positive effects such as (IBM, 2010): Focuses studets attetio o learig cocepts rather tha o formula maipulatio. Eables teachig of real-life problems, egagig studets more actively i the learig process. Icreases studets cofidece i beig able to lear ad uderstad what they iitially aticipated as beig a complicated ad difficult subject. Due to may practical values SPSS is frequetly used as a tool for teachig statistical cocepts i a majority of social sciece programs at Sloveia uiversities. I our opiio, the motivatio to lear ad the use of SPSS durig the studyig process plays a sigificat role i buildig a positive attitude towards SPSS ad to statistics itself which iflueces the usage of SPSS at the professioal level after fiishig the study. The aim of this research is to 1. idetify the exteral factors which may ifluece the adoptio ad cotiuous utilizatio of the statistical package SPSS amog the uiversity studets of social scieces; 2. examie the direct ad idirect iflueces of selected exteral variables o studets behavioural itetio which directly affect their actual use of SPSS i the future. For this purpose a research model, based o Techology Acceptace Model (TAM) will be developed. The model will be tested o uiversity studets of social scieces from seve differet faculties at three Sloveia uiversities. 2 Techology Acceptace Model TAM Techology Acceptace Model (TAM) is oe of the most widely used coceptual models i explaiig ad predictig the adoptio behaviour of iformatio techology (Hsu et al., 2009). TAM is widely kow ad it has received strog theoretical ad empirical support i literature, havig bee cited more tha 700 times (Padilla-Melédez et al. 2013). It was developed by Davis (1986, 1989) to explai the ature ad determiats of computer usage. The pricipal scheme of the origial TAM is show i Figure 1. The origial TAM postulates that»perceived Usefuless«ad»Perceived Ease of Use«are key costructs i determiig users acceptace of techology. As articulated by Davis et al. (1989) these costructs are defied i the followig way:»perceived Usefuless«is referred to as»the degree to which a perso believes that usig a particular system would ehace his/her job performace«.»perceived Ease of Use«is referred to»the degree to which a perso believes that usig a particular system would be free of effort«. Both,»Perceived Usefuless«ad»Perceived Ease of Use«have ifluece o»behavioural Itetio to Use«. The user of IS/IT iteds to use techology if the user feels the techology will be useful for them ad they feel it is easy to use. Despite that,»perceived Ease of Use«also iflueces Figure 1: The pricipal scheme of the origial TAM 117
Orgaizacija, Volume 47 Research papers Number 2, May 2014»Perceived Usefuless«but ot vice versa. Therefore, i TAM applicatios three covetioal relatioships are usually formulated i the followig research hypothesis (Lee ad Letho, 2013):»Perceived Usefuless«positively affects»behavioural Itetio to Use«.»Perceived Ease of Use«positively affects»behavioural Itetio to Use«.»Perceived Ease of Use«positively affects»perceived Usefuless«.»Perceived Usefuless«ad»Perceived Ease of Use«could be iflueced by several exteral variables which may affect attitudes toward usig the system (e.g., documetatio, system feature, traiig, user support, etc.). There are may exteral variables which are used with practical applicatios of TAM. Yousafzai et al. (2007) divided them ito four categories of orgaizatioal, system, users persoal characteristics, ad other variables. The TAM has bee validated over a wide rage of systems ad has bee idetified as a useful model i a relatively large umber of applicatios over the past two decades. A comprehesive overview of the TAM ca be foud i Legris et al. (2003) ad Chuttur (2009). Some iterestig applicatios i system dyamics ca be foud i Kljajić et al. (2012) ad Wag ad Liu (2005). I recet years several papers have bee published o the cotext of applicatio of TAM i higher educatio (e.g. Teo, 2009, 2010, 2011a, 2011b). A umber of studies have used TAM to examie learers willigess to accept e-learig systems (e.g., Al-Adwa et al., 2013; Shah et al., 2013; Sharma ad Chadel, 2013; Shroff et al., 2011; Tabak ad Nguye, 2013) or to predict learers itetios to use a olie learig commuity (Liu et al., 2010). Some papers are focused to validate TAM o some specific software which is applied i higher educatio. For example, Escobar-Rodriguez ad Moge-Lozao (2012) use TAM for explaiig or predictig uiversity studets acceptace of Moodle platform, while Hsu et al. (2009) performed a empirical study to aalyze the adoptio of statistical software amog olie MBA studets i Taiwa. 3 Research model ad hypothesis The mai goal of the paper is to idetify the ifluetial factors (the TAM exteral variables) that may facilitate or hamper the adoptio of the commercial statistical package SPSS amog the uiversity studets of social scieces. The basis for our study represets the empirical study of Hsu et al. (2009). The authors developed a exteded TAM icludig three exteral variables which are cosidered also i our model. These variables are:»spss Self Efficacy«it ca be defied as the belief that oe has the capability to perform a statistical aalysis usig SPSS. Idividuals who have high SPSS self efficacy are more likely to use SPSS ad would feel a higher level of mastery over SPSS applicatios.»computer Attitude«it ca be defied as a idex of the degree to which a perso likes or dislikes about computers. A umber of empirical studies have foud sigificat relatioships betwee attitudes about computers ad usage of them (Hsu et.al, 2009). It is postulated i our model that computer attitude affects the perceived usefuless ad the perceived ease of use of SPSS, which i tur, affects the itetio of usig SPSS i the future.»statistics Axiety«it refers to the feelig of axiety experieced by those takig a statistics course or udertake statistical aalyses. It cosists of six dimesios: worth of statistics, iterpretatio axiety, test ad call axiety, computatioal self-cocept, fear of askig for help, ad fear of statistics teacher. Worth of statistics refers to studets perceptios of relevace ad usefuless of statistics, ad is a key source of statistics axiety (Hsu et.al., 2009). It is hypothesized that lower level of statistics axiety icreases the level of perceived usefuless as well as the level of perceived ease of use. Therefore, such studets are likely to be more comfortable usig SPSS i class ad later i their daily jobs. The model of Hsu et al. (2009) was tested o olie MBA studets, where may of them are full time employees. However, the results were ot compared to studets who completed a traditioal i class course. Sice the olie MBA studets appear to be more capable, savvy ad demadig tha traditioal studets (Bisoux, 2002), it is our opiio that the model for traditioal studets eeds some adjustmet. Therefore, we supplemeted the model of Hsu et al. (2009) by ivolvig the followig additioal exteral variables:»statistics Learig Self Efficacy«Similar to SPSS self efficacy, this variable ca be defied as the studets belief i their ow ability to perform well i statistics learig tasks. It is hypothesized that higher level of self efficacy icreases the level of perceived usefuless as well as the level of perceived ease of use of SPSS.»Statistics Learig Value«The value of statistics learig is to let the studets acquire problem-solvig competecy, experiece the iquiry activity, stimulate their ow thikig, ad fid the relevace of statistics withi daily life. If they ca perceive these importat values, they will be motivated to lear statistics. I our opiio the icrease of statistics learig value decreases the level of statistics axiety ad icreases the level of statistics learig self efficacy. Higher statistic learig value also icreases the level of the perceived usefuless ad the perceived ease of use of SPSS.»Satisfactio with Achievemets«As studets icrease their competece ad achievemet durig 118
Orgaizacija, Volume 47 Research papers Number 2, May 2014 learig statistics they feel satisfactio. This results i a higher level of the perceived usefuless ad the perceived ease of use of SPSS. The compoets of our TAM-based exteded model are preseted i Figure 2. Arrows i Figure 2 represet the depedecies betwee the model compoets, where sigs»+«ad»-«idicate the positive or the egative depedece, respectively. Regardig the model descriptio, the followig hypotheses are postulated: H1a:»SPSS Self Efficacy«has positive effect o»perceived Usefuless«. H1b:»SPSS Self Efficacy«has positive effect o»perceived Ease of Use«. H2a:»Computer Attitude«has positive effect o»perceived Usefuless«. H2b:»Computer Attitude«has positive effect o»perceived Ease of Use«. H3a:»Statistics Axiety«has egative effect o»perceived Usefuless«. H3b:»Statistics Axiety«has egative effect o»perceived Ease of Use«. H4a:»Statistics Learig Self Efficacy«has positive effect o»perceived Usefuless«. H4b:»Statistics Learig Self Efficacy«has positive effect o»perceived Ease of Use«. H5a:»Statistics Learig Value«has positive effect o»perceived Usefuless«. H5b:»Statistics Learig Value«has positive effect o»perceived Ease of Use«. H5c:»Statistics Learig Value«has positive effect o»statistics Learig Self Efficacy«. H5d:»Statistics Learig Value«has egative effect o»statistics Axiety«. H6a:»Satisfactio with Achievemets«has positive effect o»perceived Usefuless«. H6b:»Satisfactio with Achievemets«has positive effect o»perceived Ease of Use«. H7:»Perceived Ease of Use«has positive effect o»perceived Usefuless«. H8:»Perceived Usefuless«has positive effect o»behavioural Itetios to Use«. Figure 2: The exteded TAM for aalysig the acceptace of SPSS amog the studets of social scieces 119
Orgaizacija, Volume 47 Research papers Number 2, May 2014 H9:»Perceived Ease of Use«has positive effect o»behavioural Itetios to Use«. 4 Methodology 4.1 Istrumetatio Our TAM-based exteded model was tested o the uiversity studets of social scieces i Sloveia. For this purpose we prepared a questioaire where every model compoet is described with a particular costruct which is represeted by several variables. I additio to the questioaire give by Hsu et al. (2009) we cosidered also the questioaire published i Tua et al. (2005). The umber of variables withi a particular costruct of our questioaire is as follows (see Table 1):»Statistics Learig Self Efficacy«- 3 variables,»behavioural Itetios to Use«ad»Perceived Ease of Use«- 4 variables,»perceived Usefuless«,»SPSS Self Efficacy«,»Computer Attitude«,»Statistics Learig Value«ad»Satisfactio with Achievemets«- 5 variables,»statistics Axiety«- 7 variables. All costruct variables are measured o the 5-poit Likert type scale of agreemet, where 1 meas»strogly disagree«, ad 5 meas»strogly agree«. 4.2 Populatio, sample ad data collectio method The web survey was performed from Jue 2013 till February 2014. Seve faculties of three Sloveia uiversities collaborated: Uiversity of Ljubljaa (Faculty of Admiistratio, Faculty of Educatio, Faculty of Arts), Uiversity of Maribor (Faculty of Orgaizatioal Scieces, Faculty of Crimial Justice ad Security, Faculty of Arts), Uiversity of Primorska (Faculty of Tourism Studies). The total umber of completed questioaires is 329. 4.3 Data aalysis Data gathered from the survey were aalysed i two stages. I the first stage the questioaire used i the survey was evaluated. The reliability of the istrumet was checked with Chrombach s α measure which was calculated for each questioaire costruct (i.e. model compoet). The samplig adequacy of the aalysis was checked with the Kaiser-Meyer-Olki measure ad Bartlett s test of sphericity. Sice our sample (N=329) is big eough (Field, 2013), a cofirmatory factor aalysis for each costruct was coducted. Furthermore, descriptive statistics of each costruct were calculated. I the secod stage of data aalysis the depedecies amog the model compoets were studied usig the regressio aalyses. Before the aalyses, we reversed scales of all three variables of the costruct»statistics Learig Self Efficacy«(which were egatively keyed i the origial questioaire). 5 Results 5.1 Questioaire evaluatio ad descriptive statistics The results of the first stage data aalysis are preseted i Table 1. I the first two colums the ames of the costructs ad the correspodig variables are listed. The third colum of the table represets the umber N of respodets that aswered all the questios withi the particular costruct. I the fourth colum Chrombach s α is show. The fifth colum combies the first two eigevalues together with the percetage of explaied variace (EV), while i the sixth colum the values of KMO ad Bartlett s test are give. I the seveth colum factor loadigs are show, ad fially i the last colum the average mea ad the average stadard deviatio (SD) for each costruct are preseted. We ca coclude from the results i Table 1 (colums from four up to seve) that our data reveals the same factors as proposed i our questioaire. It is evidet from the last colum of Table 1 that the highest rated costruct is»computer Attitude«with the highest av. mea 4,00 ad the lowest av. stadard deviatio 0,63. This shows that studets are accustomed to workig o computers, ad are aware of computer usefuless ad its importace owadays. Takig ito accout that all the questios of the costruct»statistics Axiety«were reverse phrased, it follows that the lowest rated costruct is»behavioural Itetios to Use«, with av. mea slightly above three ( = 3,05 ad SD = 0,99). O the other had, the estimates of the costruct»perceived Usefuless«were much higher ( = 3,66 ad SD = 0,85). Results for these two costructs suggest that studets actually are aware of the importace of SPSS ad they fid it quite useful for their job i the future, but at the momet they obviously have a lot of other priorities, ad SPSS does ot seem to be oe of them. Furthermore, the costruct»perceived Ease of Use«has the av. mea slightly above the itermediate value three ( = 3,21 ad SD = 0,89), which idicates that studets are ot very skilful at usig SPSS. But o the other had, from the values obtaied for the costruct»spss Self Efficacy«( = 3,75 ad SD = 0,67), it is evidet that studets believe that they could complete a statistical aalysis usig SPSS, if 120
Orgaizacija, Volume 47 Research papers Number 2, May 2014 Table 1: Results of the first stage data aalysis Factor / Costruct Behavioural Itetios to Use Perceived Usefuless Perceived Ease of Use SPSS Self Efficacy Questios - Variables I always try SPSS to coduct a task wheever it has a feature to help me perform it. I always try SPSS i as may cases/ occasios as possible. SPSS has lots of excitig fuctios that I ited to use. I ited to icrease my use of SPSS i the future. SPSS use ca improve my job performace. SPSS use ca make it easier to do my job. SPSS use i my job ca icrease my productivity. N Chr. α Eigevalues, Explaied Var. (EV) 289 0,73 λ 1 = 2,62 = 0,79 EV=65,36% 287 0,90 λ 1 = 3,62 = 0,56 EV=72,36% KMO, Bartlett s test Factor Loadigs 0,74 0,88 0,87 I fid SPSS useful i my job. 0,82 SPSS use would eable me to accomplish statistical aalysis more 0,74 quickly. I fid it easy to get SPSS to do what I wat it to do. My iteractio with SPSS is uderstadable ad clear. I fid SPSS to be flexible to iteract with. It is easy for me to become skilful at usig SPSS. I could complete a statistical aalysis usig SPSS...... if I had see someoe else usig SPSS before tryig it myself.... if someoe else had helped me get started....if someoe showed me how to do it first.... if I could call someoe for help if I got stuck.... if I had used similar software before this oe to do the same job. 290 0,90 λ 1 = 3,06 = 0,37 EV=76,56% 291 0,81 λ 1 = 2,99 = 0,80 EV=59,78% 0,84 0,80 0,91 0,84 0,55 0,89 0,91 0,88 0,85 0,91 0,88 0,86 0,67 0,87 0,86 0,83 0,59 Average Mea, SD = 3,05 SD=0,99 = 3,66 SD=0,85 = 3,21 SD=0,89 = 3,79 SD=0,67 121
Orgaizacija, Volume 47 Research papers Number 2, May 2014 Table 1: Results of the first stage data aalysis (cotiued) Factor / Costruct Computer Attitude Statistics Axiety Learig Statistics Self Efficacy Questios - Variables Computers are brigig us ito a bright ew era. The use of computers is ehacig our stadard of livig. There are ulimited possibilities of computer applicatios that have t eve bee thought of yet. Computers are resposible for may of the good thigs we ejoy. Workig with computers is a ejoyable experiece. I woder why I have to do all these thigs i statistics whe i actual life I ll ever use them. Statistics is worthless to me sice it s empirical ad my area of specializatio is philosophical. N Chr. α Eigevalues, Explaied Var. (EV) 298 0,83 λ 1 = 3,00 = 0,69 EV=59,91% 299 0,92 λ 1 = 4,76 = 0,64 EV=68,01% KMO, Bartlett s test 0,83 Factor Loadigs 0,81 0,82 0,73 0,77 0,73 0,92 0,82 I feel statistics is a waste of time. 0,87 I do t wat to lear to like statistics. 0,73 I wish the statistics requiremet would be removed from my academic program. 0,84 I do t uderstad why somebody i 0,83 my field eeds statistics. I do t see why I have to clutter up my head with statistics. It has o 0,87 sigificace to my life work. No matter how much effort I put i, I caot lear statistics. (R) Whe statistics activities are too difficult, I give up or oly do the easy parts. (R) Whe I fid the statistics cotet difficult, I do ot try to lear it. (R) 307 0,69 λ 1 = 2,18 = 0,50 EV=72,56% 0,69 0,82 Average Mea, SD = 4,00 SD=0,63 = 2,27 SD=0,82 0,81 = 3,69 SD=0,86 0,89 0,86 122
Orgaizacija, Volume 47 Research papers Number 2, May 2014 Table 1: Results of the first stage data aalysis (cotiued) Factor / Costruct Statistics Learig Value Satisfactio with Achievemets Questios - Variables I thik that learig statistics is importat because I ca use it i my daily life. I thik that learig statistics is importat because it stimulates my thikig. I statistics, I thik that it is importat to lear to solve problems. I statistics, I thik it is importat to participate i iquiry activities. It is importat to have the opportuity to satisfy my ow curiosity whe learig statistics. Durig a statistics course, I feel most fulfilled whe I attai a good score i a test. I feel most fulfilled whe I feel cofidet about the cotet i a statistics course. Durig a statistics course, I feel most fulfilled whe I am able to solve a difficult problem. Durig a statistics course, I feel most fulfilled whe the teacher accepts my ideas. Durig a statistics course, I feel most fulfilled whe other studets accept my ideas. N Chr. α Eigevalues, Explaied Var. (EV) 311 0,77 λ 1 = 2,63 = 0,71 EV=52,68% 308 0,77 λ 1 = 2,73 = 0,97 EV=54,55% KMO, Bartlett s test 0,82 0,72 Factor Loadigs Average Mea, SD 0,734 = 3,20 SD=0,71 0,746 0,730 0,665 0,751 0,643 = 3,94 SD=0,66 0,763 0,725 0,788 0,764 they had appropriate support ad eough experieces. This holds true especially for the secod degree studets ( = 3,94 ad SD = 0,65). For the costruct»statistics Axiety«we obtaied = 2,27 ad SD = 0,82. These values mea that studets are ot too axious about statistics, although we would like the values to be eve lower. Similarly, the costruct»statistics Learig Value«has the av. mea slightly above the itermediate value three ( = 3,20 ad SD = 0,71), idicatig that a average studet is ot aware i what way statistics ca cotribute to his everyday activities ad critical thikig. The results for the costruct»satisfactio with Achievemets«( = 3,94 ad SD = 0,66) mea that studets are quite satisfied whe they achieve some good results regardig statistics. For the last costruct»statistics Learig Self Efficacy«we obtaied = 3,69 ad SD = 0,86. Takig ito accout that we aalyzed the recoded variables of this costruct, these values idicate that learig statistics is ot very easy for studets, but it does ot cause a isurmoutable obstacle for them either. 5.2 Regressio aalysis I the secod stage of data aalysis the regressio aalysis was performed. The ustadardized regressio coefficiets idicatig depedecies amog the model compoets are preseted i Figure 3. 123
Orgaizacija, Volume 47 Research papers Number 2, May 2014 Statistical sigificace of ustadardized regressio coefficiets: * deotes 5% statistical sigificace level ** deotes 1% statistical sigificace level *** deotes 0,1% statistical sigificace level Figure 3: Ustadardized regressio coefficiets of model compoets We ca see from Figure 3 that all predicted depedecies betwee ie model compoets are statistically sigificat at 5% sigificace level, except the depedece of»perceived Ease of Use«o»SPSS Self Efficacy«(B=0,113, R 2 =0,007) ad»perceived Usefuless«o»Computer Attitude«(B=0,058, R 2 =0,002), where the percetages of explaied variaces i both regressios are low ad the ustadardized regressio coefficiets are ot statistically differet from zero at 5% sigificace level. Therefore, the hypotheses H1b ad H2a could ot be cofirmed. The»SPSS Self Efficacy«has a positive effect o»perceived Usefuless«(B=0,192, R 2 =0,023) at 5% sigificace level which cofirms our first hypothesis H1a. The»Computer Attitude«has positive effect o»perceived Ease of Use«(B=0,258, R 2 =0,033) at 1% sigificace level which cofirms the hypothesis H2b. I the questioaire the egatively stated items for»statistics Axiety«were used which meas that higher scores represet higher level of statistics axiety. Therefore, we assumed that there exists a egative effect o both»perceived Usefuless«(H3a) ad»perceived Ease of Use«(H3b). The results cofirmed our expectatios sice both ustadardized regressio coefficiets are egative ad statistically sigificat at 0.1% sigificat level, while B=-0,494 (R 2 =0,237) for the»perceived Usefuless«ad B=-0,485 (R 2 =0,208) for the»perceived Ease of Use«. The»Statistics Learig Self Efficacy«has a positive effect o the»perceived Usefuless«(B=0,324, R 2 =0,110) 124
Orgaizacija, Volume 47 Research papers Number 2, May 2014 ad the»perceived Ease of Use«(B=0,427, R 2 =0,174) at 0.1% sigificace level. Therefore, our hypotheses H4a ad H4b could be cofirmed. About»Statistics Learig Value«four hypotheses were postulated (see Figure 2). All of them ca be cofirmed. Namely, the»statistics Learig Value«has a positive effect o the»perceived Usefuless«(B=0,541, R 2 =0,217), the»perceived Ease of Use«(B=0,589, R 2 =0,226), ad the»statistics Learig Self Efficacy«(B=0,451, R 2 =0,141), ad a egative effect o the»statistics Axiety«(B=-0,706, R 2 =0,379). Therefore, all four hypotheses H5a, H5b, H5c, ad H5d ca be cofirmed at 0,1% sigificace level. The hypotheses H6a ad H6b explore the effects of the»satisfactio with Achievemets«o both the»perceived Usefuless«ad the»perceived Ease of Use«. Both regressios reveal positive effects of the»satisfactio with Achievemets«o two depedet variables. More precisely, the ustadardized regressio coefficiet for the»perceived Usefuless«is equal to B=0,426 (R 2 =0,109), ad the ustadardized regressio coefficiet for the»perceived Ease of Use«is equal to B=0,242 (R 2 =0,033). Sice both reported regressio coefficiets are positive ad statistically sigificatly differet from zero at 1% sigificat level, both hypotheses H6a ad H6b ca be cofirmed. The hypothesis H7 stated that»perceived Ease of Use«has a positive effect o»perceived Usefuless«ca also be cofirmed while the ustadardized regressio coefficiets is positive (B=0,482, R 2 =0,259) ad statistically sigificatly differet from zero at 0,1% sigificace level. Two of the highest three proportios of explaied variace were obtaied i two liear regressio models with the»behavioural Itetios to Use«as a depedet variable. To be precise,»perceived Usefuless«(B=0,643) ca explai 29,6% of variace of the»behavioural Itetios to Use«, while the»perceived Ease of Use«(B=0,618) ca explai 30,2% variace of the»behavioural Itetios to Use«. Accordig to positive ustadardized regressio coefficiets i both regressios, hypotheses H8 ad H9 ca be cofirmed. 6 Discussio The first objective of our paper was to idetify the exteral factors which may ifluece the adoptio ad cotiuous utilizatio of the SPSS amog the uiversity studets of social scieces. We defied six potetial factors (»SPSS Self Efficacy«,»Computer Attitude«,»Statistics Axiety«,»Statistics Learig Self Efficacy«,»Statistics Learig Value«ad»Satisfactio with Achievemets«) which represet the exteral variables of our exteded TAM. Sice all these variables are foud to have a direct ifluece o the»perceived Usefuless«ad/or»Perceived Ease of Use«, we ca assert that they also affect the behavioural itetios to use SPSS. Therefore, all these variables are relevat to be ivolved i aalysig the adoptio of SPSS amog the studets of social scieces. The secod objective was to examie the relatioships amog the model compoets. The results of our empirical study show that all three covetioal relatioships, usually formulated i TAM applicatios, ca be cofirmed. Namely, our results prove that both key compoets of the TAM,»Perceived Usefuless«ad»Perceived Ease of Use«positively ifluece studets behavioural itetios to use SPSS, while»perceived Usefuless«is also positively affected by»perceived Ease of Use«. I our opiio these results prove the applicability of TAM to our topic. Sice four of six exteral variables of our model are the same as the exteral variables cosidered by Hsu et al. (2009) it is iterestig to compare the results of both studies. I our study we foud out that»spss Self Efficacy«has a positive effect oly o»perceived Usefuless«, while the effect of this model compoet o»perceived Ease of Use«is ot sigificat. These statemets agree with the fidigs of Hsu et al. (2009). However, our results show that»computer Attitude«has a positive ifluece o»perceived Ease of Use«, while the effect of this model compoet o»perceived Usefuless«is ot sigificat. It is iterestig that the results of Hsu et al. (2009) are just the opposite. As we expected, it was foud that»statistics Axiety«has a direct egative impact o both,»perceived Usefuless«ad»Perceived Ease of Use«. This reflects i a egative ifluece o studets behavioural itetios to use SPSS. The results of Hsu et al. (2009) are similar. We agree with the authors that educators should try to elimiate studets axiety toward usig SPSS by itroducig a few carefully desiged activities ad by presetig real-world examples. I order to effectively reduce studets axiety i learig statistics, Pa ad Tag (2004) recommeded the combiatio of applicatio orieted teachig methods ad istructors attetiveess to studets axiety. The results of our study also idicate that statistics axiety ca be mitigated by icreasig the value of statistics learig. There are some limitatios that should be take ito cosideratio i the future research. For example, future research may examie whether demographic variables such as geder, age, educatioal level, etc. could potetially cofoud the observed relatioships. As previous researches suggest that the TAM ad the ed-user techology usage may differ across the cultural borders (Hsu et al., 2009), a reasoable ext step would be to extet this research to other coutries. 7 Coclusio I the paper, a exteded Techology Acceptace Model (TAM) for aalysig the acceptace of IBM SPSS Statistics amog the uiversity studets of social scieces was deve- 125
Orgaizacija, Volume 47 Research papers Number 2, May 2014 loped. O top of the traditioal compoets of the TAM, the followig six exteral variables were icluded:»spss Self Efficacy«,»Computer Attitude«,»Statistics Axiety«,»Statistics Learig Self Efficacy«,»Statistics Learig Value«ad»Satisfactio with Achievemets«. The model was tested usig the web survey ivolvig the uiversity studets of social scieces from seve differet faculties at three Sloveia uiversities. The questioaire used i the survey was evaluated with a cofirmatory factor aalysis. The reliability of the scale with Chrombach s α was examied. The samplig adequacy for the aalysis with the Kaiser-Meyer-Olki measure ad Bartlett s test of sphericity was checked. Descriptive statistics of the model compoets were calculated, ad depedecies betwee the model compoets were studied usig the regressio aalyses. The empirical results prove that all exteral variables cosidered i our model are relevat, ad directly ifluece the»perceived Usefuless«ad»Perceived Ease of Use«which are the key compoets of the traditioal TAM. Both,»Perceived Usefuless«ad»Perceived Ease of Use«have direct impact o the studets behavioural itetio which affects their actual use of SPSS i the future. Therefore, we ca assert that all exteral variables icluded i our model represet the potetial areas where activities to reduce the studets axiety ad to stregthe the positive attitudes towards statistics ad SPSS ca be plaed. Therefore, we ca coclude that the aim of the paper has bee achieved. Our TAM-based exteded model is foud to be useful i studyig the adoptio ad cotiuous utilizatio of SPSS amog the studets of social scieces. Fidigs obtaied with the model applicatio are of great value for educators, ad ca help them to improve the learig process. I the ext stage of our research we are goig to cotiue the validatio of our model by expadig the survey to some faculties from other East Europea coutries coverig the social sciece studyig programs. I additio to regressio aalyses we ited to employ other applicable statistical methods. For example, structural equatio modellig which could eable us to ivestigate all causal coectios amog the model compoets simultaeously, or hierarchical clusterig combied with K-meas clusterig which may reveal clusters of studets with similar attitudes toward statistics ad SPSS. Refereces Al-Adwa, Am., Al-Adwa, Ah., & Smedley, J. (2013). Explorig studets acceptace of e-learig usig Techology Acceptace, Model i Jordaia uiversities. Iteratioal Joural of Educatio ad Developmet usig Iformatio ad Commuicatio Techology, (IJEDICT), 9(2), 4-18. Retrieved December 5, 2013 from http://ijedict.dec.uwi.edu/ viewissue.php?id=35 Baloğlu, M. (2003). Idividual differeces i statistics axiety amog college studets. Persoality ad Idividual Differeces, 34(5), 855 865, http://dx.doi.org/10.1016/ S0191-8869(02)00076-4 Bisoux, T. (2002). The e-learig equatio. BizEd, 40 45. Chuttur M.Y. (2009). Overview of the Techology Acceptace Model: Origis, Developmets ad Future Directios. Workig Papers o Iformatio Systems, 9(37), 9-37. Retrieved October 12, 2013 from http://sprouts.aiset. org/785/1/tamreview.pdf Davis, F. D. (1986). A techology acceptace model for empirically testig ew ed-user iformatio systems: theory ad results. Ph.D. Thesis, Massachusetts Istitute of Techology, Sloa School of Maagemet. Davis, F. D. (1989). Perceived Usefuless, Perceived Ease of Use, ad User Acceptace of Iformatio Techology. MIS Quarterly, 13(3), 319-340, http://dx.doi.org/10.2307/249008 DeWaye, T. A. (2010). Axiety ad Attitude of Graduate Studets i O-Campus vs. Olie Statistics Courses. Joural of Statistics Educatio, 18(1), 1-15. Retrieved October 12, 2013 from http://www.amstat.org/publicatios/jse/v181/derryberry.pdf Escobar-Rodriguez, T., & Moge-Lozao, P. (2012). The acceptace of Moodle techology by busiess admiistratio studets. Computers & Educatio, 58, 1085 1093, http://dx.doi. org/10.1016/j.compedu.2011.11.012 Field, A. (2013). Discoverig statistics usig IBM SPSS statistics. Lodo: SAGE. Hsu M. K., Wag S. W., & Chiu K. K. (2009). Computer attitude, statistics axiety ad self-efficacy o statistical software adoptio behavior: A empirical study of olie MBA learers. Computers i Huma Behavior, 25, 412 420, http:// dx.doi.org/10.1016/j.chb.2008.10.003 IBM (2010). Makig statistics relevat ad easy to lear. Retrieved October 13, 2013, from http://www.presidio.com/idustry_ solutios/academia/materials/case_studies/ytc03045usen. pdf Kljajić, M. et al. (2012). System approach to MIS ad DSS ad its modelig withi SD, I M. Mora, O. Gelma, A. Steekamp, & M. Raisighai (Eds.) Research Methodologies, Iovatios ad Philosophies i Software Systems Egieerig ad Iformatio Systems (340-359). Hershey, Iformatio Sciece Referece, http://dx.doi.org/10.4018/978-1-4666-0179-6. ch017 Lee, D.Y., & Lehto, M.R. (2013). User acceptace of YouTube for procedural learig: A extesio of the Techology Acceptace Model. Computers & Educatio, 61, 193-208, http://dx.doi.org/10.1016/j.compedu.2012.10.001 Legris, P., Ighamb, J., & Collerettec, P. (2003). Why do people use iformatio techology? A critical review of the techology acceptace model. Iformatio & Maagemet, 40, 191 204, http://dx.doi.org/10.1016/s0378-7206(01)00143-4 Liu, I-F. et al. (2010). Extedig the TAM model to explore the factors that affect Itetio to Use a Olie Learig Commuity. Computers & Educatio, 54 (2), 600-610, http:// dx.doi.org/10.1016/j.compedu.2009.09.009 Owuegbuzie, A. J. (2004). Academic procrastiatio ad statistics axiety. Assessmet & Evaluatio i Higher Educatio, 29(1), 3-19, http://dx.doi.org/10.1080/0260293042000160384 Owuegbuzie, A.J., & Wilso, V.A. (2003). Statistics axiety: Nature, etiology, atecedets, effects, ad treatmets A 126
Orgaizacija, Volume 47 Research papers Number 2, May 2014 comprehesive review of the literature. Teachig i Higher Educatio, 8 (2), 195-209, http://dx.doi.org/10.1080/135625 1032000052447 Padilla-Melédez, A., Del Aguila-Obra, A.R., & Garrido-Moreo, A. (2013). Perceived playfuless, geder differeces ad techology acceptace model i a bleded learig sceario. Computers & Educatio, 63, 306-317, http://dx.doi. org/10.1016/j.compedu.2012.12.014 Pa, W., & Tag, M. (2004). Examiig the Effectiveess of Iovative Istructioal Methods o Reducig Statistics Axiety for Graduate Studets i the Social Scieces. Joural of Istructioal Psychology, 31(2), 149-159. Ravid, R., & Leo, M. R. (1995). Studets perceptios of the research compoet i master s level teacher educatio programs. Paper preseted at the Aual Meetig of the America Educatioal Research Associatio. Sa Fracisco, CA, April 18-22, 1995. (ERIC Documet Reproductio Service No. ED393840). Shah, S. A. M. et al. (2013). Employee Behavior Towards Adoptio of E-learig Courses: Validatig Techology Acceptace Model. Mediterraea Joural of Social Scieces, 4(14), 765-773, http://dx.doi.org/10.5901/mjss.2013.v414p765 Sharma, S. K., & Chadel, J. K. (2013). Techology Acceptace Model for the Use of Learig Through Websites Amog Studets i Oma. The Iteratioal Arab Joural of Iformatio Techology, 3(1), 44-49. Shroff, R. H., Deee, C. C., & Ng, E. M. W. (2011). Aalysis of the techology acceptace model i examiig studets behavioural itetio to use a e-portfolio system. Australasia Joural of Educatioal Techology, 27(4), 600-618. Retrieved December 12, 2013 from http://www.ascilite. org.au/ajet/ajet27/shroff.html Stickels, J.W., & Dobbs, R.R. (2007). Helpig Alleviate Statistical Axiety with Computer Aided Statistical Classes. The Joural of Scholarship of Teachig ad Learig, 7(1), 1-15, Retrieved October 12, 2013 from http://josotl.idiaa.edu/ article/view/1676/1674 Tabak, F., & Nguye, N. T. (2013). Techology Acceptace ad Performace i Olie Learig Eviromets: Impact of Self-Regulatio. MERLOT Joural of Olie Learig ad Teachig, 9(1), 1-11. Retrieved December 2, 2013 from http:// jolt.merlot.org/vol9o1/tabak_0313.pdf Teo, T. (2009). Modellig techology acceptace i educatio: A study of pre-service teachers. Computers & Educatio, 52, 302 312, http://dx.doi.org/10.1016/j.compedu.2008.08.006 Teo, T. (2010). A empirical study to validate the Techology Acceptace Model (TAM) i explaiig the itetio to use techology amog educatioal users. Iteratioal Joural of Iformatio ad Commuicatio Techology Educatio, 6(4), 1-12, http://dx.doi.org/10.4018/jicte.2010100101 Teo, T. (2011a). Factors ifluecig teachers itetio to use techology: Model developmet ad test. Computers & Educatio, (57), 2432 2440, http://dx.doi.org/10.1016/j.compedu.2011.06.008 Teo, T. (2011b). Techology Acceptace i Educatio, Research ad Issues. Rotterdam: Sese Publishers. Tua H. L., Chi C. C., & Shieh S. H. (2005). The developmet of a questioaire to measure studets motivatio towards sciece learig. Iteratioal Joural of Sciece Educatio, 27(6), 639 654, http://dx.doi.org/10.1080/09500690420003 23737 Yousafzai, S.Y, et al. (2007). Techology acceptace: a meta-aalysis of the TAM: Part 1. Joural of Modellig i Maagemet, 2(3), 251 280, http://dx.doi.org/10.1108/17465660710834453 Wag, W. T., & Liu, C. Y. (2005). The Applicatio of the Techology Acceptace Model: A New Way to Evaluate Iformatio System Success. I J. D. Sterma (ed.), N. P. Repeig (ed.), R. S. Lager (ed.), J. I. Rowe (ed.), & J. M. Yai (ed.) Proceedigs of the 23th Iteratioal Coferece of the System Dyamics Society. Albay: System Dyamics Society. Zeider, M. (1991). Statistics ad mathematics axiety i social sciece studets: Some iterestig parallels. British Joural of Educatioal Psychology, 61(3), 319 328, http://dx.doi. org/10.1111/j.2044-8279.1991.tb00989.x Ackowledgmet Authors express their thaks to Jasa Mažgo, La Umek ad Jaez Vogric from the Uiversity of Ljubljaa, Matevž Bre, Jaja Jerebic ad Jereja Šifrer from the Uiversity of Maribor, Daša Fabja ad Saša Plaic from the Uiversity of Primorska, who ecourage their studets to participate i the survey. Aleka Brezavšček is a Assistat Professor of Stochastic Processes at the Faculty of Orgaizatioal Scieces, Uiversity of Maribor. Her mai research iterests are stochastic processes, system reliability ad availability, ad iformatio system security. Petra Šparl is a Associate Professor of Mathematics at the Faculty of Orgaizatioal Scieces, Uiversity of Maribor, Sloveia. Her mai research iterests are graph theory, data aalysis ad studets performace i methodological courses. Aja Židaršč is a Assistat Professor of Quatitative Methods at the Faculty of Orgaizatioal Scieces, Uiversity of Maribor, Sloveia. Her mai research iterests are social etwork aalysis, micro-eterprises ad iformatio-commuicatio techology, ad studets performace i methodological courses. 127