The study was a qualitative one designed over three phases to support a comparison of pre- and post-intervention measures. CGI was introduced as a vehicle for learning about children's mathematical thinking, and not as a course to be subsequently evaluated.
Data were gathered during each phase of the study.