Vocational education researchers have relied heavily on opinionnaires to measure opinions and attitudes. Usually respondents are asked to indicate their degree of agreement to an attitude or opinion statement. Where significant statistical differences exist for opinions/attitudes there is still always the problem of confidence with which respondents respond to statements. This level of confidence is rarely determined in most studies. In this paper the use of the Certainty of Response Rating Technique (CRRT) in vocational education research is suggested as a measure of level of confidence. Its practical application is illustrated using a 1982 study of agricultural teachers concerning their opinions of international agricultural and rural development. The CRRT appears to have four major functions: (1) as a moderator variable, changing the kind and nature of relationship between two other variables; (2) to partition respondents into more homogeneous groups; (3) for interpreting research findings with a higher degree of confidence; and (4) to improve the initial correlations between predictor and criterion variables. (Author/BW)