Literacy Coaching as a Component of Professional Development

1
Literacy Coaching as a Component of Professional
Development

• Joanne F. Carlisle, PhD
• Coauthors
• Kai Cortina, Dan Berebitsky (University of
  Michigan), and Lauren Katz (Bowling Green State
  University)

2

• Current issues
• What are effective features of programs of
  professional development (PD) for teachers of
  early reading? Three proposed features
• Content to improve teachers disciplinary
  knowledge,
• Some way to engage teachers interest, motivation
  commitment to improving their teaching practices,
  
• Adequate (long-term) support for making and
  evaluating changes, implementing new approaches
• Are all three necessary to bring about
  substantive change in instruction and students
  achievement in reading?

3

• What features of PD motivate teachers to analyze
  their own teaching and seek ways to improve it?
• A discouraging finding is that few efforts to
  improve teachers professional knowledge have had
  detectable effects on instruction (Cohen
  Ball, 1999).
• Researchers have called for studies of components
  of PD programs that are likely to deepen
  teachers professional knowledge and bring about
  improvements in their instruction (Borko, 2004
  Desimone, 2009).
• Garet et al. (2008), PD with coach, without
  coach, and control group improvement in
  knowledge but no significant differences in
  instruction or students reading achievement

4

• Project inspired by Reading First as implemented
  in Michigan
• Knowledge Language Essentials for Teachers of
  Reading and Spelling (LETRS Moats, 2003)
• Guidance in using student assessments (e.g.,
  DIBELS) to evaluate the effectiveness of
  instruction
• Support for integration and implementation from
  literacy coach (weekly grade level meetings,
  class visits).
• If LETRS is considered a basic (traditional)
  model, what is the added value (if any) of the
  two other components? A question of importance
  for resource-limited schools.

5
(No Transcript)
6

• The first study compares teachers responses to
  the three models of PD
• Do perceptions of the effectiveness of their
  teaching change?
• Does instruction change across the year?
• The second study compares KEC and KE models
• Is coaching associated with greater improvements
  in students literacy?
• Does school climate mediated the effect of the PD
  on students improvement in reading?

7

• Study 1 Method
• Participants 111 first-grade teachers in
  schools in 9 districts (KEC 43 teachers KE 33 K
  35). No difference in 4th graders
  underachieving in reading by model. Teachers did
  not differ in ethnic background, years teaching
  KE had more teachers with masters degree.
• Treatment
• 9 seminars delivering Moats Language Essentials
  for Teachers of Reading and Spelling
• KEC and KE teachers taught to use DIBELS,
  guidance interpreting results.
• KEC teachers had in-school support from the coach
  (weekly grade-level meetings, classroom visits).

8

• Data collection
• Surveys
• Satisfaction with my work 3 factors--Self-Effica
  cy, Perceived Improvement in Teaching Reading,
  and Perceived Sources of Teaching Support.
• Language and Reading Concepts designed to assess
  knowledge gained from LETRS.
• Classroom observations fall, winter, and spring
  coding system--Teachers Instructional Practices
  (TIP). Of interest, practices influenced by
  LETRS
• time spent on components of reading (e.g.,
  phonics
• of time providing small group instruction and
  using learning centers (e.g., flexible grouping,
  guided reading).
• Coach survey (21 coaches) (in schools with
  coaches only)

9
(No Transcript)
10

• Study 1 Results
• Were there differential benefits in terms in
  improved performance on the teacher knowledge
  measure (LRC) by condition? No. Significant
  effect for time but no significant differences
  for teachers in the three models.
• Did teachers in the three conditions differ in
  their self-reported views about their teaching?
• Satisfaction With My Work Significant effect for
  time KE teachers showed significantly greater
  gains in self-efficacy in the winter and spring.

11

• Were there observed differences in key areas of
  instruction in year 1 (TIP variables) ?
• Phonics showed decrease over time for whole class
  lessonsmore rapid for KEC than KE or K. (Time
  spent on comprehension/vocabulary, fluency,
  writing did not differ by condition.)
• Centers/small group instruction Increase in time
  devoted to SGI across the year, KEC teachers
  spent more time providing instruction in small
  groups.

12
Whole class phonics broken line shows KEC spent
less time in phonics overall, big drop in winter.
13

• Do satisfaction, knowledge, and observed
  practices distinguish teachers in the three
  models? Discriminant function analysis (fall,
  winter, spring)
• The the fall, the first function was not
  significant for winter and spring, first
  function was significant.
• For the winter, time spent on centers/small group
  instruction), phonics, writing, and fluency
  significantly discriminated the groups.
• For the spring, the variables that significantly
  distinguished the groups were time spent on
  centers/SGI, fluency, and writing.
• Classification results ( correctly predicted by
  variables)
• Winter 76 KEC, 42 KE and 48 K
• Spring 72 KEC, 70 KE, and 30 K.

14

• Rationale for Study 2 Comparison of KE and KEC
• Experts believe that literacy coaching provides
  more intensive, on-demand support to extend
  teachers PD in their school and classroom (IRA,
  2004).
• School leadership and efforts to build a
  collaborative teacher community have been found
  to contribute to school improvement (e.g.,
  Sebring et al., 2006)
• The major research question
• Is teachers participation in the KEC or KE model
  associated with different outcomes for students
  word decoding, when taking into account attitudes
  toward the PD program, school climate, and time
  spent on key aspects of their reading
  instruction?

15

• Study 2, Method (year 2 of the study)
• 30 KEC and 39 KE teachers
• Measures from classroom observations (TIP)
• Four school climate teacher surveys (scales
  developed by Chicago Consortium)
• Communication Around Literacy (winter)
• Attitudes Toward PD (spring)
• rincipal Support for Change (spring)
• Reflective Dialogue (spring)
• DIBELS Nonsense Word Fluency (NWF) reading
  nonsense words in 1 min data collected by
  classroom teachers in the fall, winter and
  spring. Measures raw scores and risk status
  (low, medium, high).

16
(No Transcript)
17
(No Transcript)
18

• Study 2, Preliminary Results
• KEC and KE teachers did not differ in their
  attitudes toward the PD they received, the
  support of their principal, or opportunities for
  collaboration with other teachers.
• Differences in two areas of observed instruction
  (repeated measures ANOVAs controlled for fall
  NWF, teachers experience, and school climate
  measures)
• Phonics KE teachers time on phonics dropped off
  more across the year.
• Small group instruction/centers) KEC teachers
  spent more time the level of principal support
  positively and significantly influenced SGI.

19
(No Transcript)
20
Multilevel analysis effects of classroom
entering ability NWF, instruction (e.g., SGI),
and climate variables on change of risk
status. Only significant covariates were
teachers assessment of principals support for
change an NWF fall class average (others
dropped) Results showed that the probability of
high risk students being designated as at risk or
some risk in the spring decreased more for KEC
than KE
21
(No Transcript)
22

• What have we learned?
• Teachers responded positively to PD, whatever
  model they participated in. LETRS led to gains in
  knowledge, but that alone did not lead to
  noticeable changes in instruction.
• More intensive PD models with more support for
  teachers better than efforts to improve knowledge
  alone.
• Differences in changes in time devoted to areas
  of literacy instruction distinguished the
  teachers in the three models, winter and spring.
  
• Support for coaching over and above KE greater
  change in 1st graders decoding in KEC than KE
  classrooms. Results are unlike those of Garet et
  al (2008) perhaps because of methodology.

23
Questions? Comments? Contact
information Joanne F. Carlisle jfcarl_at_umich.edu