Patterns of Engagement and Professional Growth in an Online Mentoring and Induction Program for Beginning Science and Math Teachers
Authors: Irene Grimberg, Elisabeth Swanson, Larry Bice, DeAnna McAleer, Peggy Taylor, Anthony Villar

Contents
3. Design, Data & Analysis
Print Poster
« Back to Poster Hall
3. Design, Data & Analysis
Next »

The exploration of these claims was accomplished through a mixed-methods research approach (Creswell, 2003). The qualitative methods included: a) analysis of the online dialogue and teacher portfolios using a constant comparison approach (Patton, 2002) to trace both pre-selected and emergent themes, Toulmin's (2003) to examine patterns of argumentation (assertions and evidence about teaching) present in the dialogue, and Levin et al.'s (1990) graphical representations to show relationships in the content of multiple posts; b) teacher case studies including analysis of interview transcripts, online dialogue sampled across a teacher's tenure in the program (1-4 years), and portfolios assembled by the teachers that provided samples of their work as a discussion facilitator, mentor or mentee, and their interpretive comments; c) semi-structured phone interviews with teachers and faculty, and subsequent transcript analysis; d) teachers' responses to open-ended survey questions and constant comparative analysis . The quantitative aspect of the research study included: a) the use of Likert-style surveys for teachers' self-reported activities and online participation habits; b) the analysis of online dialogues before and after selected interventions using coding rubrics, descriptive statistics and statistical analysis; c)  tallying posting and reading behavior of mentors and mentees.  When surveys, interview protocols or coding rubrics were created by eMSS evaluators or researchers, relevant experts external to the project reviewed and validated item content, and the data collection or analysis tools were extensively piloted and revised.  Inter-rater reliability for coding rubrics was ensured through training procedures, and monitored once coding was underway.