Although attempts have been made to relate the four ba~ic dimensions in a mathematical manner (e.g. Wallston and Wallston. 1984), the Janz and Becker review and most of the primary studies have discussed each dimension separately. For purposes of consistency and comparability. we analyzed each dimen!oion separately. Janz and Becker assessed the predictive power of the HBM by calculating ·significance ratios', i.e. dividing the significant positive findings by the total number of findings. Such reliance on significance t~ting from original studies is subject to the influence both of effect size and sample size, as well as the homogeneity of variances across studies (Light and Pillemer. 1984: Hedges and Olkin. 1985).Rosenthal (1984) demonstrated that the Pearson product-moment correlation (r) can be used as a common estimate of effect size (the magnitude of a relationship) and gave fonnulae for converting other Statistical results tor. As the distribution of r's S