D. B. Bailey and S. A. Palsha (1992) proposed two modified versions of the Stages of Concern Questionnaire for measuring teacher concerns during a reform effort. Their analysis suggested the use of a 5-factor model with 35 items or 15 items rather than the original 7-stage, 35-item Concerns Based Adoption Model (CBAM). The present study was carried out in two phases. The first phase analyzed data from 376 algebra teachers to evaluate reliability and validity issues of both the original Stages of Concern Questionnaire and the two proposed versions of Bailey and Pasha. Reliability estimates were lower for the 15-item revised version, but agreed with Bailey and Palsha for the original instrument and the 35-item, 5-factor version. A confirmatory factor analysis was conducted using the 5-factor, 35-item version. A less than optimal fit of the model to the data indicated continued problems with validity. From factor loadings and modification indices, a 27-item 5-stage model was proposed that more closely matched the original CBAM without the Awareness and Refocusing Stages. The second phase of the study employed a confirmatory factor analysis of the 27-item 5-stage model to examine data from a new sample of 237 algebra teachers. Reliability estimates were consistently higher than for other versions, but validity continued to be problematic. A broad range of issues arising from these results is discussed. (Contains 8 tables and 14 references.) (Author/SLD)