No measure is perfect, but some measures can be quite useful

The comment articles in this issue by Friese and Fiedler (F&F) and by Rothermund and Wentura (R&W) offer perspectives on the validity of the Brief Implicit Association Test (BIAT) (Sriram & Greenwald, 2009; S&G). F&F concluded that construct validity of the BIAT can be established only by conducting studies that experimentally manipulate association strengths. We suggest that this conclusion overvalues experimental strategies and undervalues correlational validation strategies. R&W's critique was predicated on their use of a semantic-network theoretical understanding of the concept of association. In contrast, S&G offered the BIAT as a technique for measuring association strengths in the context of a broader concept of association that has roots in antiquity--and remains widely used in psychology. With this broader understanding of association, some of the phenomena that R&W treated as threats to the BIAT's validity are viewed, instead, as contributors to its validity.