Teacher evaluation and local control in the US: an investigation into the degree of local control afforded to districts in defining evaluation procedures for teachers in non-tested subjects and grades

Abstract

Many states only recently incorporated indicators of student achievement into teacher evaluation systems for Non-Tested Subjects and Grades (NTSG). This study examines how practices related to the inclusion of student achievement measures vary across states as to the discretion left to districts in defining and implementing evaluation systems for teachers in NTSG. For each state, information about current practices was obtained through document analysis and, when provided, feedback from state department representatives. We find substantial variation in state polices. Some states — notably those that received Race to the Top funding — afford districts with considerably less local control than others. Results presented here provide rich descriptive information and highlight the need for future studies related to local control and teacher evaluation.

Notes

1. Question 9 was also examined for inclusion in this index. However, state responses to this question proved difficult to classify. Furthermore, reliability testing found that the operationalized response to question 9 decreased the internal consistency of the index variable. For these reasons, it was not included.

2. Washington DC is excluded from these analyses, given that it is a district and not a state.

3. The number of groups and cut points between these groups could be considered somewhat arbitrary. One may have included only two groups, for instance, or set the cut points differently between high and moderate local control. Ultimately the decisions made with two criteria in mind: (1) preserve that natural groupings that seemed to emerge from the distribution of the NTSGLCI and (2) generate a somewhat-even distribution of states into each group. For example, there was a steep drop off between values of 8 and 9, so that was deemed an appropriate cut for high and moderate categories. The cut between moderate and low categories was made using the latter of the two criteria.