Teaching residential segregation: an evaluation of the value of graphical representations presented with the SegMaps program

Abstract

In this thesis, I evaluated the effectiveness of a specific teaching strategy (use of thematic maps to depict segregation patterns) in teaching sociology students about the phenomena of residential segregation. The thesis design was unique because it included quantitative data measuring the effectiveness of the new teaching strategy, in addition to the qualitative information typically reported in such studies. The thesis used an experimental design commonly referred to as the non-equivalent research design, and included controls for all the normal aspects of a classroom interaction such as an assigned reading and a lecture that did not include the program illustrations being tested. While this thesis did not provide conclusive support for the effectiveness of SegMaps, based on global tests, it did provide evidence that further testing is merited. Both the pretest and lecture had significant positive effects on student learning outcomes (measured by the posttest score). The assigned reading was not significant to the outcome variable, but only 50% of students asked reported that they had in fact attempted the assignment. More troubling was the insignificant relationship between the SegMaps variable and the students' results on the posttest for the global tests (involving a composite index based on answers to all ten questions). However, more detailed analyses of the individual questions indicated the SegMaps variable was significantly related to the question expressly included to measure its effectiveness. The assigned reading, lecture, and posttest variables, also performed as expected in these individual tests, lend support to my hypothesis.