ERIC Number: ED172425
Record Type: Non-Journal
Publication Date: 1979-May
Reference Count: N/A
Approaches to Data Analysis in Longitudinal Field Investigations of Educational Programs.
Jovick, Thomas D.
The federally funded longitudinal field study called Management Implications of Team Teaching (MITT) required a search for an appropriate strategy for analyzing through-time relationships among selected variables. The MITT project used questionnaires and interviews to collect data concerning the work, governance, attitudes, and orientation of teachers in fourteen multiunit elementary schools and thirteen conventionally organized elementary schools. In a study like MITT, one typically wishes to investigate gains or decrements in scores on variables as a consequence of either treatment differences (multiunit as opposed to conventional organization) or variation in quantitative variables. Two problems confront investigators who embark on longitudinal data analyses: one concerns whether to use original scores or "true score" estimates, and the other concerns what statistical procedures to apply to these scores. This document explains why and how the investigators in this case chose to use original scores and multiple linear regression analysis. While data from the MITT project are used to illustrate the process by which these methods were chosen, the findings of the original study are not presented here. (Author/PGD)
Publication Type: Reports - Research; Reports - Evaluative
Education Level: N/A
Sponsor: National Inst. of Education (DHEW), Washington, DC.
Authoring Institution: Oregon Univ., Eugene. Center for Educational Policy and Management.