In early 2014, with the leadership of Deputy Superintendent

Transcription

1 2014 TELL Oregon Survey Research rief Winer TELL Oregon Survey Suden Achievemen and Teacher Reenion Analyses Inroducion In early 2014, wih he leadership of Depuy Superinenden of Public Insrucion Rob Saxon, a group of sakeholders ha included he Oregon Educaion Invesmen oard, Oregon Deparmen of Educaion, Oregon Educaion Associaion, Confederaion of Oregon School Adminisraors, Oregon School oards Associaion, and he Chalkboard Projec collaboraed wih he New Teacher Cener (NTC) o adminiser he Oregon Teaching, Empowering, Leading and Learning Survey (TELL Oregon Survey). The survey assesses wheher educaors across he sae repor having he resources and suppors necessary o encourage effecive eaching. The TELL Oregon Survey is a full-populaion survey based on he NTC TELL Survey firs developed in he Norh Carolina Governor s Office in I has since been replicaed in more han 12 saes and 10 disrics o collec criical daa on eacher working condiions. The TELL Oregon Survey is a full populaion survey based on he NTC TELL survey firs developed in he Norh Carolina Governor s Office in The TELL Survey is designed o repor educaors percepions abou he presence of eaching and learning condiions organized ino he following eigh consrucs: Time, Faciliies and Resources, Professional Developmen, School Leadership, Teacher Leadership, Insrucional Pracices and Suppor, Managing Suden Conduc, and Communiy Suppor and Involvemen. A series of NTC briefs provide resuls from he 2014 TELL Oregon Survey, summarizing he insrumen design and psychomeric properies and describing preliminary findings and group comparisons. These resources can be found on he TELL Oregon websie under he Research ab (hp://elloregon.org/research). This brief esablishes he research foundaion specifically linking eaching condiions as measured by he NTC TELL Survey o suden achievemen and eacher reenion oucomes, provides informaion on response raes o he 2014 TELL Oregon Survey, and ess he associaion beween 2014 survey daa and suden and eacher oucomes. The purpose of his repor is o help sakeholders beer undersand he relaionship beween eaching condiions and oucomes of ineres wihin he sae conex. The curren policy conex, wih is increasing emphasis on eacher and principal evaluaion, demands a more nuanced undersanding of he associaion beween eaching and suden learning. Sakeholders wan o beer undersand under wha condiions eachers conribue o suden learning (Hanushek & Rivkin, 2007; Seele, Hamilon, & Secher, 2010) as a growing body of research indicaes ha school environmens can encourage or consrain good eaching ( Johnson & he Projec on he Nex Generaion of Teachers, 2004; McLaughlin & Talber, 2001). This work is summarized below as background o he TELL Oregon analyses

2 Providing Teachers wih he es Opporuniy o e Effecive Connecions eween Teaching Condiions and Suden Learning Teacher success is faciliaed by a posiive school conex, leadership, and a collaboraive working environmen. In paricular, research shows ha srong, rusing relaionships boh inernal and exernal and supporive school leadership are linked o improved suden achievemen ( Johnson, 2006; ryk & Schneider, 2002). Oher research demonsraes he imporance of communicaion and collaboraion for improving suden achievemen. For example, in schools where eachers alk o each oher abou heir work and principals communicae wih he communiy, sudens have higher reading and mahemaics es scores han sudens in schools where hese condiions were no as prevalen. Addiionally, hese condiions have a greaer impac on es scores han he experience or credenials of he saff (Leana & Pil, 2006). A 2009 analysis by Ladd ha used NTC survey daa also shows ha eaching condiions are linked o suden performance and can predic as much as 15 percen of school aggregae achievemen resuls. Also using NTC survey daa, Johnson, Kraf, and Papay (2011) find ha posiive condiions conribue o improved suden achievemen. Specifically, heir research shows ha in low-income, high-minoriy schools, percepions of more posiive eaching condiions are associaed wih beer suden academic oucomes. More recen research describes how he condiions assessed by he 2012 TELL Massachuses Survey heoreically and empirically link o imporan oucomes, including suden learning. Ferguson wih Hirsch (2014) demonsrae significan connecions beween eaching condiions and suden valueadded gains. In paricular, he auhors find ha four areas assessed by he NTC survey suden conduc managemen, demands on ime, professional auonomy, and professional developmen are linked o he prerequisie condiions for achievemen gains (e.g., suden percepions of suppor and rigor). Thus, posiive educaor percepions in hese four areas are associaed wih facors linked o improved suden engagemen and learning. Addiional recen work by Kraf and Papay (2014) also uses suden-eacher linked daa and school-level eaching condiions as measured by he NTC survey. The researchers find ha eachers who work in more supporive environmens become more effecive a raising suden achievemen on sandardized ess over ime han do eachers who work in less supporive environmens, afer conrolling for suden characerisics, prior es scores, and eacher and school characerisics. Teachers in schools ha had he mos posiive eaching condiions (in he 75h percenile as measured by 24 quesions in NTC s TELL Survey) were 38 percen more effecive afer a decade han eachers in schools in he 25h percenile. Over wo years, eachers were 11 percen more effecive if hey worked in schools wih posiive eaching condiions. Connecions eween Teaching Condiions and Teacher Reenion A hos of large-scale empirical sudies provide evidence ha conexual facors also maer in eachers decisions abou saying or leaving schools. In a mea-analysis of 34 sudies, orman and Dowling (2008) sugges ha eaching and learning condiions influence eachers career pahs more han previously documened. oyd e al. (2011) demonsrae ha eachers percepions of he school adminisraion have he greaes influence on eacher reenion decisions. Oher work finds similar effecs (see, for example, Pogodzinski, Youngs, Frank, & elman, 2012). Sudies also find saisically significan relaionships beween eachers percepions of school faciliies and heir plans o say or leave (Loeb, Darling-Hammond, & Luczak, 2005; uckley, Schneider, & Shang, 2004). Similar o he suden learning oucomes described previously, exernal researchers using NTC survey daa from an insrumen similar o he TELL Oregon Survey also demonsrae associaions beween eaching condiions and eacher reenion. Johnson, Kraf, and Papay (2011) find ha eachers are more saisfied and plan o say longer in schools wih posiive eaching

3 condiions. Their work suggess ha condiions such as a rusing amosphere, principal leadership, and collaboraive colleagues are as imporan, or more imporan, han condiions such as faciliies and resources in influencing eachers decisions o say in schools. This finding holds rue afer conrolling for suden and school characerisics such as he percenage of sudens caegorized as low income. Ladd (2009), also using TELL daa, documens ha eaching and learning condiions predic eacher plans o leave a school, independen of school demographics. This robus research foundaion demonsraes a consisen link beween eaching condiions and boh suden achievemen and eacher reenion oucomes. This brief adds o his work by analyzing 2014 TELL Oregon Survey daa. The brief provides a summary of survey paricipans and analyses of sae and school-level daa o help sakeholders undersand which eaching condiions maer mos in promoing eacher and suden success TELL Oregon Survey Paricipans NTC adminisered he 2014 TELL Oregon Survey o all school-based licensed educaors in early The daa for hese analyses include responses from more han 19,300 educaors in Oregon, yielding a response rae of 59 percen. Respondens include several caegories of educaors: 90 percen are eachers, 3 percen are principals, 1 percen are assisan principals, and 6 percen are oher licensed educaors such as librarians and school psychologiss (Table 1). TALE 1. RESPONSE RATE Y PARTICIPANT TYPE Teacher Principal Assisan Principal Oher Educaion Professional Toal Respondens* Spring 2014 Response Rae (N) 89.9 (17,418) 3.0 (588) 1.2 (236) 5.8 (1,131) 19,373 *Noe. The responden caegory eachers includes insrucional coaches, deparmen heads, lieracy specialiss, ec. The responden caegory Oher Educaion Professionals includes school counselors, school psychologiss, social workers, ec. Response raes also vary by school ype. As Table 2 demonsraes, he sample of paricipans includes 61 percen of elemenary school educaors, 63 percen of middle school educaors, 57 percen of high school educaors, 48 percen of educaors a special schools educaors, and 41 percen of educaors a combined schools. TALE SURVEY RESPONSE RATE Y SCHOOL TYPE Elemenary Middle High Special Combined Schools Toal School Type Headcoun 15,708 5,851 9, ,464 32,609 Responded 9,637 3,660 5, ,373 Percen Responded Of he 1,265 schools across he sae of Oregon, 784 me or exceeded he 50 percen minimum response rae hreshold (wih a leas five respondens) o have access o individual school-level repors on heir survey resuls. Those resuls can be accessed a How Oregon Teaching Condiions Impac Suden Learning The goal of hese analyses is o beer undersand how eaching condiions inersec wih suden performance and eacher reenion in he conex of Oregon schools. Do schools wih beer eaching condiions have beer suden performance, greaer academic growh, and/or higher eacher reenion? A brief summary of oucomes and approaches follows, wih a deailed mehodology in Appendix A. The Oregon Assessmen of Knowledge and Skills (OAKS) is used here o measure suden performance in erms of absolue achievemen. Teacher reenion is esimaed based on responses o quesion 10.1 on he TELL Oregon survey ( Which of he following bes describes your immediae professional plans? ), wih he percenage of respondens responding coninue eaching a my curren school considered as reained for hese analyses. The - 3 -

4 eaching condiions measures include boh an overall indicaor ha combines all eigh consrucs as well as separae measures of each consruc. The Oregon Deparmen of Educaion provided oher variables. Using saisical approaches appropriae for school-level daa, hese analyses isolae he effec of eaching condiions from oher facors ha research suggess are relaed o suden academic performance, such as eacher and suden background characerisics. The analyses combine school-level daa across elemenary, middle, and high schools for sae-level findings. See Appendix A for a full explanaion of saisical modeling and variables. Teaching Condiions and Suden Performance Analyses In schools where educaors repor beer eaching condiions, higher percenages of sudens achieve proficiency on OAKS. Specifically, hree condiions predic suden achievemen: a schools wih srong or sufficien communiy suppor, faciliies and resources, and ime, more sudens achieve proficiency on OAKS. These resuls are imporan because hey show he impac of eaching condiions while conrolling for facors such as suden povery, oal enrollmen, and eacher experience a he overall sae level. Significanly, he conribuion of communiy suppor and professional developmen o suden achievemen is sronger han he conribuion of he percenage of sudens caegorized as receiving free or reduced-price lunch. See Appendix for sae-level model saisics. Teaching Condiions and Teacher Reenion Analyses Summary These analyses show ha beer eaching condiions are consisenly associaed wih beer oucomes in erms of suden performance and eacher reenion. Specifically, srong communiy suppor is relaed o higher suden achievemen and lower eacher ariion, and he impac on achievemen is sronger han he influence of he percenage of sudens caegorized as economically disadvanaged. Implicaions Togeher, hese analyses demonsrae ha many facors ha are wihin he conrol of sakeholders and policymakers conribue o creaing environmens where srong eaching and learning can occur. These findings sugges ha communiy involvemen, professional developmen, and suden behavior managemen play key roles in improving suden achievemen. Sakeholders may consider addiional analyses o beer undersand he inersecion beween hese condiions and oucomes of ineres a differen school levels, especially a he middle and high school levels, where posiive suden managemen sysems and srong communiy involvemen are less ypical han in elemenary schools. This evidence suggess ha, overall, eaching condiions are consisenly relaed o improved learning and eacher reenion. ased on hese findings, local educaion agencies and campuses should review heir TELL disric- and campuslevel repors. These daa can faciliae conversaions abou how o mainain and improve he eaching condiions ha analyses demonsrae help eachers and sudens succeed. Higher perceived communiy suppor and eacher leadership are relaed o fewer eachers leaving heir schools. Even afer including he conribuions of oher facors such as suden and eacher background characerisics, daa sugges ha in schools where eachers repor higher levels of suppor from he communiy and school leaders, fewer eachers choose o leave he classroom. See Appendix C for full models

5 Appendix A. Model Specificaion and s Model Specificaions Saisical models appropriae for school-level daa es he relaionship beween eaching condiions and suden achievemen using Ordinary Leas Squares (OLS) regression. The OLS equaion assumes here is a linear associaion beween he oucome variable and he independen variable. For example, OLS assumes changes in eaching condiions are associaed wih changes in suden achievemen and beer eacher condiions are associaed wih beer suden achievemen. An advanage of OLS is ha i allows he relaionship beween eaching condiions and oucome variables o be isolaed by conrolling for oher facors, such as eacher and suden background characerisics. The following equaion (1) specifies he regression model using percenage proficien on OAKS as he oucome variable: (1) Yi = β0 + β1(suden) + β2(school) + β3(teacher) + β4(teaching Condiions) + βi All variables are a he school level. The oucome variable Yi in model (1) is he percen of sudens scoring proficien or above on OAKS. The β0 represens he value of he oucome variable when all he independen variables are a zero. The independen variables are represened by β1-4 and include blocks of characerisics abou sudens, schools, eachers, and eaching condiions. s included in each block follow, and full descripions are provided below. The eaching condiions measure consiss of he average of he eigh consruc means for each school. The β, or beas, are values, one for each explanaory variable, ha represen he srengh and ype of relaionship he independen variable has o he dependen variable. If he β is posiive, hen as he independen variable increases, he oucome variable increases. If he β is negaive, hen as he independen variable increases, he oucome variable decreases. The βi is he error erm or he difference beween he expeced value generaed by he regression equaion and he observed value in he daa for each school in his case. The eacher reenion regression model (2) follows a similar equaion as presened for he suden oucome models. The rae of eachers leaving classrooms is he oucome variable Yi. (2) Yi = β0 + β1(suden) + β2(school) + β3(teacher) + β4(teaching Condiions) + βi Oucome s Suden Performance OAKS Online is a compuer-based adapive es in which iems are seleced according o each suden s demonsraed abiliy. All of he es iems have been developed by Oregon eachers and reviewed by Oregon expers. OAKS Online has incorporaed a number of feaures and updaes based on inpu from Oregon eachers familiar wih he sysem. Sudens may ake assessmens via OAKS Online up o hree imes per year in a seven-monh esing window, lasing from approximaely November hrough May. The Oregon Deparmen of Educaion and is es vendor, American Insiues for Research, developed several enhancemens o he OAKS Online Sysem for he school year. Tes Adminisraors (TAs) were required o review he user guides prior o adminisering any OAKS Online assessmens. Teacher reenion: Teacher reenion is he rae a which eachers responding o he TELL Oregon Survey indicaed ha hey inend o remain eaching in heir school. The esimaed eacher reenion variable was calculaed from quesion 10.1 on he TELL Oregon Survey. Independen s Considered in he Models School Characerisics Change in Enrollmen: Change in enrollmen a he school building year over year School Enrollmen: Toal number of sudens enrolled a a school Average Number of Languages Spoken in a School - 5 -

6 Teacher Characerisics : Average salary of eachers a he school level Percen of Classes Taugh by Highly Qualified Teachers Percen Teachers Provisionally Licensed Suden Characerisics Percen of Sudens wih Disabiliies: Percenage of sudens a he school level paricipaing in special educaion services Percen of Whie Sudens: Percenage of sudens a he school level ha idenify wih he Caucasian race Suden Aendance: Number of sudens in aendance in heir classrooms annually aggregaed o he school level Percen of English Language Learners: Percenage of sudens idenified as eiher non-english proficien or limied English proficien. Non-English proficien is defined as a suden who speaks a language oher han English and does no comprehend, speak, read, or wrie English. Limied English proficien is defined as a suden who comprehends, speaks, reads, or wries some English bu whose predominan comprehension or speech is in a language oher han English. Disrics mus provide language services o all limied English proficien sudens. Percen Economically Disadvanaged: Percenage of sudens qualifying for eiher he free or reduced-price lunch program. The Federal Naional School Lunch Ac esablishes eligibiliy for he reduced-price lunch program for families wih incomes up o 185 percen of he federal povery level (in 2005, his amoun was $35,798 for a family of four). Families wih incomes up o 130 percen of he federal povery level qualify for he free lunch program (in 2005 his amoun was $25,155 for a family of four)

7 Appendix. Suden Achievemen Saewide Table -1 presens informaion from he OLS model (1) where he oucome variable is he performance on OAKS, eaching condiions is a composie measure across all eigh consrucs, and he elemenary, middle, and high school levels are combined. The unsandardized coefficien for he eaching condiions composie mean indicaes ha for every 1-poin change in he eaching condiions mean, he percenage of sudens score on OAKS would increase almos 5 percenage poins. Changes in he eaching condiions composie mean of half a poin or less are more common; however, o make model inerpreaion easier, he sandard 1-poin change in he mean is used. The able presens oher facors he model idenified as significan a he.05 level. TALE -1. MODEL SUMMARY EXPLAINING STATEWIDE STUDENT ACHIEVEMENT MEAN COMPOSITE (N=751) Sd. Error ea (Consan) Suden Aendance School Enrollmen Percen Economically Disadvanaged Percen of Sudens wih Disabiliies Percen of Whie Sudens Average Number of Languages Spoken in he School TELL: Mean Composie Adjused R 2 =

8 TALE -2. MODEL SUMMARY EXPLAINING STATEWIDE STUDENT ACHIEVEMENT CONSTRUCT AREA (N=751) Sd. Error ea (Consan) Suden Aendance School Enrollmen Percen Economically Disadvanaged Percen of Sudens wih Disabiliies Percen of Whie Sudens Average Number of Languages Spoken in he School TELL: Time TELL: Faciliies and Resources TELL: Communiy Suppor and Involvemen TELL: Professional Developmen Adjused R 2 =.491 School Level Models for elemenary and middle school levels esing he associaion beween he percenage of sudens passing OAKS and overall eaching condiions show posiive and significan associaions (see Tables -3 and -4). High schools failed o show significan differences beween he wo variables (Table -5). TALE -3. MODEL SUMMARY EXPLAINING ELEMENTARY SCHOOL STUDENT ACHIEVEMENT COMPOSITE (N=460) Sd. Error ea (Consan) Suden Aendance Percen Economically Disadvanaged Percen of Whie Sudens TELL: Mean Composie Adjused R 2 =

9 TALE -4. MODEL SUMMARY EXPLAINING MIDDLE SCHOOL STUDENT ACHIEVEMENT COMPOSITE (N=129) Sd. Error ea (Consan) Percen Economically Disadvanaged Percen of Sudens wih Disabiliies Average Number of Languages Spoken in he School TELL: Mean Composie Adjused R 2 =.526 TALE -5. MODEL SUMMARY EXPLAINING HIGH SCHOOL ACHIEVEMENT COMPOSITE (N=162) Sd. Error ea (Consan) Suden Aendance Change in School Enrollmen School Enrollmen Percen of Classes Taugh wih Highly Qualified Teachers Percen of Teachers Provisionally Licensed Percen of English Language Learners Percen Economically Disadvanaged Percen of Sudens wih Disabiliies Percen of Whie Sudens Average Number of Languages Spoken in he School TELL: Mean Composie Adjused R 2 =

10 School Level by TELL Consruc Models for each school level ha included he individual eaching condiions show ha, a he elemenary school level, Time and Managing Suden Conduc have a significan and consisenly posiive associaion wih suden learning. A he middle and high school level, Communiy Suppor has a significan and consisenly posiive associaion wih suden learning. Faciliies and Resources were also found o have a posiive influence on suden achievemen a he high school level. Professional Developmen a he elemenary level and School Leadership a he high school level have a significan and negaive associaion wih suden achievemen. For complee models, see Tables TALE -6. MODEL SUMMARY EXPLAINING ELEMENTARY SCHOOL STUDENT ACHIEVEMENT CONSTRUCT AREA (N=459) Sd. Error ea (Consan) Suden Aendance Percen Economically Disadvanaged Percen of Sudens wih Disabiliies Percen of Whie Sudens TELL: Time TELL: Managing Suden Conduc TELL: Professional Developmen Adjused R 2 =

11 TALE -7. MODEL SUMMARY EXPLAINING MIDDLE SCHOOL STUDENT ACHIEVEMENT CONSTRUCT AREA (N=129) Sd. Error ea (Consan) Percen of Sudens wih Disabiliies Average Number of Languages Spoken in he School TELL: Communiy Suppor and Involvemen Adjused R 2 =.478 TALE -8. MODEL SUMMARY EXPLAINING HIGH SCHOOL STUDENT ACHIEVEMENT CONSTRUCT AREA (N=164) Sd. Error ea (Consan) Suden Aendance Percen of Sudens wih Disabiliies TELL: Faciliies and Resources TELL: Communiy Suppor and Involvemen TELL: Teacher Leadership Adjused R 2 =

12 Appendix C. Teacher Reenion Saewide condiions more posiively, he eacher reenion rae increases. Table C-1 presens oher facors he model idenified as significan as well. Table C-1 presens OLS model (3). The model demonsraes ha as eachers view heir school s eaching TALE C-1 MODEL SUMMARY EXPLAINING STATEWIDE ESTIMATED TEACHER RETENTION MEAN COMPOSITE (N=762) Sd. Error ea (Consan) Suden Aendance School Enrollmen 2.008E E TELL: Mean Composie Adjused R 2 =.195 School Level Teacher reenion is higher a all school levels where eachers perceive here are beer eaching condiions; in all cases, he influence of eaching condiions oupaces ha of any oher independen variable inroduced ino he model. Tables C-2 C-4 presen significan coefficiens. TALE C-2 MODEL SUMMARY EXPLAINING ELEMENTARY SCHOOL ESTIMATED TEACHER RETENTION COMPOSITE (N=465) Sd. Error ea (Consan) Suden Aendance E Percen of Classes Taugh wih Highly Qualified Teachers TELL: Mean Composie Adjused R 2 =

13 TALE C-3 MODEL SUMMARY EXPLAINING MIDDLE SCHOOL ESTIMATED TEACHER RETENTION COMPOSITE (N=130) Sd. Error ea (Consan) Suden Aendance Change in School Enrollmen TELL: Mean Composie Adjused R 2 =.303 TALE C-4 MODEL SUMMARY EXPLAINING HIGH SCHOOL ESTIMATED TEACHER RETENTION COMPOSITE (N=170) Sd. Error ea (Consan) Suden Aendance E TELL: Mean Composie Adjused R 2 =

14 School Level by TELL Consruc In elemenary schools, he Faciliies and Resources and School Leadership consrucs are associaed wih more eachers saying a a school. A he middle school level, Insrucional Pracices and Suppor effecs a eacher s decision o coninue eaching more han suden aendance and eacher salary combined. A he high school level, Communiy Suppor and Involvemen and School Leadership boh have a greaer influence on reaining eachers han salary. See Tables C-5 C-7 for full models. TALE C-5. MODEL SUMMARY EXPLAINING ELEMENTARY SCHOOL ESTIMATED TEACHER RETENTION CONSTRUCT AREA (N=467) Sd. Error ea (Consan) Suden Aendance Percen of Classes Taugh wih Highly Qualified Teachers TELL: Faciliies and Resources TELL: School Leadership Adjused R 2 =.189 TALE C-6. MODEL SUMMARY EXPLAINING MIDDLE SCHOOL ESTIMATED TEACHER RETENTION CONSTRUCT AREA (N=130) Sd. Error ea (Consan) Suden Aendance Change in School Enrollmen E TELL: Insrucional Pracices and Suppor Adjused R 2 =

15 TALE C-7. MODEL SUMMARY EXPLAINING HIGH SCHOOL ESTIMATED TEACHER RETENTION CONSTRUCT AREA (N=168) Sd. Error ea (Consan) E TELL: Communiy Suppor and Involvemen TELL: School Leadership Adjused R 2 =.206 References orman, G. & Dowling, N. (2008). Teacher ariion and reenion: A mea-analyic and narraive review of he research. Review of Educaional Research, 78(3). oyd, D., Grossman, P., Ing, M., Lankford, H., Loeb, S. & Wyckoff, J. (2011). The influence of school adminisraors on eacher reenion decisions. American Educaional Research Journal, 48(2). uckley, J., Schneider, M., & Shang, Y. (2004). The effecs of school faciliy qualiy on eacher reenion in urban school disrics. Chesnu Hill, MA: Naional Clearinghouse for Educaional Faciliies. ryk, A. S. & Schneider,. (2002). Trus in schools: A core resource for improvemen. New York: Russell Safe Foundaion. Corina, J. (2002). ig hings have small beginnings: An assormen of minor mehodological misundersandings. Journal of Managemen, 28(3). Ferguson, R.,(wih Hirsch, E.) (2014). How working condiions predic eaching qualiy and suden oucomes. In T. J. Kane, K. A. Kerr, and R. C. Piana (Eds.), Designing Teacher Evaluaion Sysems: New guidance from he Measures of Effecive Teaching projec (pp ). San Francisco: Jossey-ass. IDRA Research Technology Group. (n.d.) Annoaed SPSS oupu facor analysis. Rerieved November 20, 2012 from hp:// Johnson, S., Kraf, M. & Papay, J. (2011). How conex maers in high-need schools: The effecs of eachers working condiions on heir professional saisfacion and heir sudens achievemen. Teachers College Record, 114(10). Johnson, S. M. (2006, July). The workplace maers: Teacher qualiy, reenion, and effeciveness. NEA Working Paper. Washingon, D.C.: Naional Educaion Associaion

16 Johnson, S. M. & The Projec on he Nex Generaion of Teachers. (2004). Finders and keepers: Helping new eachers survive and hrive in our schools. San Francisco: Jossey-ass. Kraf, M. A., & Papay, J. P. (2014, Jan. 30). Can professional environmens in schools promoe eacher developmen? Explaining heerogeneiy in reurns o eaching experience. Educaional Evaluaion and Policy Analysis, 36(4), Ladd, H. (2009). Teachers percepions of heir working condiions: How predicive of policy relevan oucomes? CALDER Working Paper 33. Washingon, D.C.: Naional Cener for Analysis of Longiudinal Daa in Educaion. Leana, C. & Pil, F. (2006). Social Capial and Organizaional Performance: Evidence from Urban Public Schools. Organizaion Science. Vol. 17, No. 3, May June 2006, pp Loeb, S., Darling-Hammond, L. & Luczak, J. (2005). How eaching condiions predic eacher urnover in California schools. Peabody Journal of Educaion, 80(3). McLaughlin, M. W. & Talber, J. E. (2001). Professional communiies and he work of high school eaching. Chicago: Universiy of Chicago Press. New Teacher Cener. (2013). The Impac of Teaching and Learning Condiions on Suden Performance and Fuure Employmen Plans. Sana Cruz, CA: New Teacher Cener. Rerieved from hp:// Pogodzinski,., Youngs, P., Frank, K. & elman, D. (2012). Adminisraive climae and novices' inen o remain eaching. The Elemenary School Journal, 113(2). Seele, J., Hamilon, L. & Secher,. (2010). Incorporaing suden performance measures ino eacher evaluaion sysems. Sana Monica, CA: RAND Corporaion. Abou he New Teacher Cener New Teacher Cener focuses on improving suden learning by acceleraing he effeciveness of new eachers. NTC parners wih saes, school disrics, and policymakers o design and implemen sysems ha creae susainable, high-qualiy menoring and professional developmen; build leadership capaciy; work o enhance eaching condiions; improve reenion; and ransform schools in vibran learning communiies where all sudens succeed Fron Sree, Suie 400, Sana Cruz, CA I Fax: I info@neweachercener.org DOC-SATR-USOR-1412