An attribute-treatment interaction study was conducted with two groups of teachers exhibiting contrasting natural teaching styles, teaching a Biological Sciences Curriculum Study (BSCS) based biology course in a natural school setting. Annotated transcripts of lessons were analyzed to provide data for the validation and description of the contrasting styles. Multiple regression techniques provided a parsimonious statement of relationships between 15 individual predictors (general, ability, four personality factors, two cognitive preferences, treatment, and seven attribute-treatment interactions) and each criterion variable (four attitude scales and two achievement tests). For three outcomes, treatment interactions were found with personality characteristics and/or cognitive preferences. (Author/CP)