Implementation and Effectiveness of the Content-Focused Coaching® Program

1. Implementation and Effectiveness of the Content-Focused Coaching® Program LINDSAY CLARE MATSUMURA HELEN GARNIER BRIAN JUNKER LAUREN RESNICK DONNA DIPRIMA BICKEL March 4, 2010 Society for Research on Educational Effectiveness

2. Context for the Study •Literacy coaching is widespread •Little evidence showsthat coaching influences instruction and student learning • Research shows that quality of coaching varies significantly across schools • Standards for coach qualifications often not followed (IRA, 2004; 2006) • What it means to be a “coach” is variably defined (Duessen et al, 2007) • Coaching resources used in a diffuse way

3. Content Focused Coaching • Intensive literacy-coach professional development program developed by the Institute for Learning (IFL) • 3 days a month over the academic year led by IFL fellows • Goals of the coach training • Develop coaching skills • Build subject matter knowledge and pedagogical skills to assist Ts to enact more rigorous reading comprehension lessons • Improving quality of class discussions about texts (Questioning the Author, Beck & McKeown, 2006)

4. Content-Focused Coaching® • Coaches work with IFL Trainers 3x month • Principals and District staff attend District Coaches work with Ts: • Weekly meetings in grade-level teams • Monthly meeting individually, modeling in classrooms, observing and co-teaching School Classroom Ts enact QtA lessons with Ss in their classroom

5. Study Design • Three year study (2006-2009) • Urban district in Texas • 91% of students eligible for free-lunch • 80% Hispanic, 15% African American • 40% English language learners •Lowest-performing schools randomly assigned to treatment (n=15) and comparison (n=14) conditions

6. Data Sources •Data sources include: • Teacher surveys (baseline and end of each study year) • Frequency of participation in literacy coaching • Satisfaction with coaching • Content of the coaching activities • Coach and principal interviews (once a year) • Classroom observations (fall and spring of each study year) • Quality of text discussions • Rigor of text discussions and lesson activities • Student test scores • Degrees of Reading Power assessment (fall and spring of each study year) • Texas Assessment of Knowledge and Skills (spring of each year)

7. Overview of the Talk • Part 1: Influence of a school’s social resources on the implementation of the Content-Focused Coaching (CFC) program (Y1) • Principal leadership • Norms for the professional community • Teacher experience • Part 2: Effectiveness of the CFC program (Y1 and Y2): • Ts participation in coaching • Observed text discussions • Ss reading achievement (all Ss; ELLs)

8. Part 1: Influence of a School’s Social Resources on Teachers’ Coaching Participation Analyses • Regression analyses based on T survey responses (N=96) • Social resources assessed at baseline • Principal leadership • Norms for the professional community • Professional experience • Teachers’ coaching participation assessed at the end of Y1 • Overall frequency of Ts’ participation in coaching • Ts perception of the usefulness of coaching • Content of coaching activities • Qualitative analyses of CFC coach interviews

9. Principal Leadership is Key •You take the principals to the [CFC] trainings, but honestly, it still comes down to if the principal doesn’t really want the coach to do these things, doesn’t value the coach doing these things, isn’t leading the way so the coach can follow, it just isn’t going to happen.

10. Principal Leadership •Principal’s willingness to share leadership predicted: • Greater frequency of T participation in coaching (p<.05 ) • Greater emphasisof coaching received by Ts • Planning and reflecting on instruction (p<.05 ) • Lesson enactment (p<.05) • Stronger T agreementthat the coaching they received had improved their practice (p<.01 )

11. Mechanisms by Which Principals Influence Coaches’ Work •Interviews with CFC coaches indicated that principals positively influenced their work by: • Actively supporting and participating in the CFC program • The day I came, the P introduced me to the faculty. She told them that CFC was vital for us to change our ways of thinking and that it was going to take some time…and that we would be very patient and not despair. They would get it and everybody is learning. She was learning…I was in a learner-student role and they were gonna be in the same role. • CFC isn’t the Ps agenda. It’s happening at her school and she knows that if she tries to block it she will get into trouble, but she is not going to pave the way for me. If the Ts are reluctant or hesitant she is not going to help. • Publicly identifying the coach as a resource for Ts • Referring Ts to coach for literacy related questions • Holding faculty-wide PD sessions, inviting coach to serve on leadership committees, watching the coach model lessons in Ts classrooms • Allowing coaches to manage their own schedules • Ps negative relationship with Ts impeded coaches’ work

12. Norms of the Professional Community •Strong existing culture of T collaboration predicted: • Less emphasis of coaching received by Ts on • Planning and reflecting on instruction (p<.05 ) • Lesson enactment (p<.05 ) • Stronger T disagreementthat the coaching they received had improved their practice (p<.05 )

13. Mechanisms by Which Ts Professional Community Influence Coaches’ Work •Interviews with CFC coaches suggested that: • In a few schools Ts were organized against coaching (n=2) • [This school] has a reputation to oust their coach within a year or two. They don’t like coaches at this campus so the longest a coach has been here is two years and then they’re out, they’re gone… • In a few schools with strong professional communities, reform goals were not aligned with CFC goals (n=3) • In some schools with a very weak professional culture, Ts were interested in working with coaches to alleviate their isolation (n=5) • Contrary pattern of collaborative professional community supporting coaching also found in some schools (n=5)

14. Teachers’ Years of Experience Teaching •Less experienced teachers participated more frequently in coaching (p<.05 ) • New Ts described by coaches as more receptive to coaching than more veteran teachers (n=10) • New teachers are really positive and appreciative of getting extra support. • I’m just another person coming into her school trying to save her school…She’s seen my kind so many times before she’s sick of us. So I don’t expect her to be my best buddy anytime soon.

15. Part 2: Effectiveness of the CFC Program (Y1 and 2) •What is the influence of the CFC program on teachers’ coaching experiences, reading comprehension instruction and students’ reading achievement?

16. Participants • Students (N=1754) • 4th and 5th grade • 91% eligible for free or reduced price lunch • 80% Hispanic; 15% African American • 40% English language learners (ELLs) • Teachers (N=98) • 7 years average teaching experience • 38% heldmaster’s degree • 56% taught in both English and Spanish

17. Analyses •Hierarchical linear growth models: • Amount of coaching and focus of coaching received by Ts • T surveys (baseline and end of each year) • T belief that coaching helped improve their practice • T surveys (baseline and end of each year) • Quality of instruction • Observed text discussions (fall and spring of each year) • Student achievement • Texas Assessment of Knowledge and Skills (end of each year) • Degrees of Reading Power Assessment (fall and spring of each year)

18. Effect of CFC on T Participation in Coaching •Ts in the CFC schools compared to Ts in comparison schoolsby the end of Y2 • Participated more frequently in coaching (p=.000, ES=.89) • More strongly agreed that the coaching they received was useful to them for improving their practice (p=.000, ES=.95)

19. Figure 1: Coach Observed Teacher for 30 Minutes

20. Effect of CFC on Content of Coaching Activities •Ts in the CFC schools compared to Ts in comparison schools participated in coaching activities at the end of Y2 that more strongly emphasized • Building knowledge of the theory underlying effective reading instruction (p=.016, ES=.70) • Planning and reflecting on instruction (p=.002, ES=.94) • Lesson enactment (p=.001, ES=.91) • Differentiating instruction (p=.007, ES=.76)

21. Observations of Classroom Discussions •T and Ss participation • Percent of Ss participating in the discussion • T connects Ss contributions • Ss connect to each others’ contributions • T presses Ss to explain their answers using evidence from the text • Ss use evidence from the text to explain their answers • Rigor of the lesson • Quality (‘grist’) of text discussed • Ss opportunity to analyze and interpret a text

22. Effect of CFC on Observed Instruction •T and Ss participation observed in CFC schools compared to comparison schools • Greater proportion of Ss participating in the discussion(p=.005, ES=.35) • T more often connects Ss contributions (p=.003, ES=.46) • Ss more often connect to each others’ contributions (p=.025, ES=.38) • T more often presses Ss to explain their answers using evidence (p=.049, ES=.33) • Ss more often use evidence from the text to explain their answers(p=.006, ES=.43)

23. Effect of CFC on Observed Instruction • Rigor of the lesson observed in CFC schools compared to comparison schools • Higher quality (‘grist’) of the text discussed (p=.012, ES=.52) • More opportunities for Ss to analyze and interpret a text (p=.011, ES=.39)

24. Figure 2. Observation Ratings of Quality of Participation in Class Text Discussions,Fall 2006 to Spring 2008 (Cohort 1, N=98)Note. Significant change over time indicated in graph by **p<.01.Within-time comparisons indicated the following differences:Fall 2006, Spring 2007: No differences detected. Fall 2007, Spring 2008: CFC>Comparison, p<.01. **

25. * * Figure 3. Observation Ratings of Rigor of Class Text Discussions, Fall 2006 to Spring 2008 (Cohort 1, N=98)Note. Significant change over time indicated in graph by *p<.05.Within-time comparisons indicated the following differences:Fall 2006, Spring 2007, Spring 2008: No differences detected. Fall 2007: CFC>Comparison, p<.01.

26. Effect of CFC on Student Achievement • Average school-level achievement for students in CFC schools compared to students in comparison schools • Higher average gain in reading achievement for ELL students (p=.013, ES=.61) • No significant effect on the average level of reading achievement for all students

27. Dimensions of Instruction Associated with Improved Achievement for ELL Students • Higher average reading achievement of ELL students in CFC schools associated with: • Greater proportion of Ss participating in the discussion (p=.000, ES=.85) • Ss more often use evidence from a text to support their answers (p=.026, ES=.33) • Higher quality of the text discussed (p=.008, ES=.45)

28. Recent NAEP Findings (2007) • ELLs are the fastest growing subgroup in the U.S. • The reading achievement of ELL students is very low • 7.5% proficient • 70% below-basic

29. Summary of Findings • Contextual factors in schools significantly influenced the initial implementation of CFC • Principal leadership played a key supportive role • Less experienced Ts were more receptive to coaching • Stronger existing professional norms negatively impacted some of the coaches’ work with Ts • CFC produced positive results in desired teacher and student outcomes • Strong effect on Ts coaching experiences and attitude toward coaching • Moderate effect on reading comprehension instruction • Moderate effect on reading achievement of ELL students (40% of the sample)

30. Thank You! For further information about the study please contact me at: Lclare@pitt.edu