The target article seems to provide yet another illustration of the classic "It don't make no nevermind" principle in statistical analysis. In particular, relatively simple measures appear to do approximately as well as more complex measures do, even including indicators that represent nonlinear transformations of the simpler measures. The authors of the target article admit that their validity tests are limited--confined to correlations with career age and departmental reputation. At the very least, alternative citation measures must be tested against eminence measures, including awards and honors. Because the latter can be stratified into degrees of distinction, such measures can be highly discriminating with respect to citation impact. In this commentary, the author emphasizes a different form of validation, namely, using citation indicators of impact as criteria or dependent variables in predictive regression equations. A rather extensive body of research has accumulated that identifies dispositional traits and developmental events most strongly associated with scientific achievement. The dispositional traits include both cognitive and personality variables, whereas, the developmental events include such factors as family background, formal education, and direct mentoring. The question then arises: Which of the alternative impact criteria is most strongly predicted by the relevant dispositional and developmental variables? For instance, if the "h" index is the most predictable of all candidate citation criteria, and if the same measure correlates most highly with eminence and awards, then the case for its assessment superiority could not be stronger. The index would boast both psychological and sociological validity. In contrast, if it should prove the case that log-transformed citation counts (or some other alternative) are more strongly associated with both (a) award and eminence evaluations and (b) dispositional and developmental predictors, then th the "h" index would be weakened. One has no guarantee that one specific impact measure will emerge victorious. Indeed, some degree of divergence might be expected given that the various measures place somewhat different emphases on the relative contributions of quantity (publications) and quality (citations) and that these two components can have contrasting correlations with other variables, whether eminence criteria or dispositional and developmental predictors. Nevertheless, if a choice has to be made among competing measures, optimal solution would probably be to pick that indicator that correlates most strongly with substantive variables. In any event, it should be clear that the current comparative inquiry represents just a first step toward a more complete evaluation.