Teacher Participation Assessment Paper

In order to enhance teacher participation, I am offering principals of participating schools $200 pre-paid VISA gift cards to encourage their teachers to take the survey. To sweeten the deal for teachers—taking advantage of my very generous budget—I am offering every participating teacher a $50 prepaid VISA gift cards upon successful completion of each survey ($100 total). Since the online surveys will be fielded by a member of my research team on site, I will be able to control survey conditions and ensure that teachers do not discuss answers with each other during the survey.

MEASUREMENT

Variables As a general note, the most succinct definitions of each variable can be found in Table 1 of Appendix A. All variables will be measured at or

As noted in the “variables” section, I will be using multiple indicators to measure constructs like CTE, fairness of ratings, usefulness of ratings, and pressure to raise or maintain ratings. Using multiple indicators to capture these constructs enhance measurement validity. In all of the matrix tables, the order of items will be randomized to avoid any question order bias. Additionally, questions in the matrix tables are worded positively and negatively to control for the fact that the the overall positivity of question wording has an impact on responses (Goddard et al. 2000). Theoretically, since I am taking a census of all teachers in a school (and, under these ideal conditions, have a 100% response rate), non-response bias should not be a problem. The cover letters for my survey are written with enough information about the survey contents so as to properly inform participants but not with too much information, so as to avoid problems with internal validity. Although maturation is a threat to the internal validity of this project as teachers may naturally become more cohesive and more confident in their abilities to educate students over the course of a year, the use of a control group (schools that have no change in accountability rating from 2018 to 2019) would allow effects like maturation to be statistically controlled for. Similarly, although my project could suffer from differential attrition— e.g. teachers in low-SES schools leaving after the initial survey but before the follow-up survey—this threat to internal validity should be able to be controlled for with the teacher longevity variable; I would be able to predict differential attrition by looking at the correlation between teacher longevity and SES. The inclusion of several control variables should enhance non-spuriousness. As an added