Donn Byrne (1971) and his colleagues have developed a convincing argument in favor of a respondent conditioning interpretation of the linear relationship between attitudinal similarity and attraction. Attitudinal statements serve as UCSs in the sense that they elicit affect in subjects exposed to them (Byrne & Clore, 1967; Clore & Gormly, 1969). Previously neutral stimuli, when associated with attitudinal statements, become CSs and elicit similar affective responses. Successfully conditioned stimuli include slides of strangers or of randomly selected objects (Byrne & Clore, 1970; Sachs & Byrne, 1970), the experimenter, the apparatus and the experiment (Baskett, 1971; Griffitt & Guay, 1969). Related research suggests that attitudinal similarity is a determining variable in the evaluation of a stranger's task performance (Smith, Meadow & Sisk, 1970), in the hiring of teachers (Merritt, 1970), in grades given to college students (Hamlish & Gaier, 1954), in decisions on loan approvals (Golightly, Huffman, & Byrne, 1971) and personnel selection (Griffitt & Jackson, 1970) and in the length of sentences given to murderers (Mahaffey, 1969). In each case the relationship between attitudinal similarity and the criterion is positive and linear. The respondent hypothesis holds important implications for classroom teachers. It has been demonstrated that a significant other becomes a positive or negative CS on the basis of his attitudinal position. It seems reasonable to assume that, when the significant other is a teacher, cognitive material being simultaneously presented will comprise an additional set of conditioned