The original data met one of the three assumptions of MANCOVA noted by Garson (2010a). The rela- tionships between the dependent variables and the covariates were linear. Descriptive statistics revealed kurtosis of pretest intentions, posttest attitudes and intention scores were beyond the 12 to