Stevens, JoeKennedy, Patrick2014-06-172014-06-172014-06-17https://hdl.handle.net/1794/17919Teaching is a complex and fundamentally collaborative process, through which knowledge and skills are acquired as a result of repeated interactions between teachers and students. The effectiveness of these interactions depends on both the instructional environment created by the teacher and the extent to which students engage with that environment. The current study combines these dimensions of teaching to (a) evaluate the construct validity of two instruments: the Middle School Intervention Project Classroom Observation Tool (MSIP-COT) and the Student Engagement Instrument (SEI), and (b) test the extent to which these measures predict differences in reading proficiency for middle school students who struggle with reading. Observation, engagement, and reading achievement data were collected for a sample of 1,446 reading intervention students from 25 middle schools in the Pacific Northwest participating in an evaluation of state and local education programs. Instruments were evaluated by fitting a series of measurement models to the observation and engagement data. The results of the best fitting models were then used in a cross-classified, multilevel structural equation model to predict differences in reading proficiency and evaluate the direct and mediational effects of engagement and instructional environment. Results provided reasonably strong evidence for both measures as indicators of their respective constructs but limited support for the direct and mediational effects of observed instructional environment and self-reported student engagement on reading proficiency for these students. Limitations and directions for future research are discussed.en-USAll Rights Reserved.Classroom observationsConstruct validityInstructional evaluationMiddle schoolReading achievementStudent engagementElectronic Thesis or Dissertation