Suppose we have a word recognition task, on the basis of which we compute the four rates defined in Signal Detection Theory as Hits, False Alarms, Correct Rejections, and Misses (HR, FAR, CRR, MR). Due to some reasons, a researcher decides to use only CRR as a dependent variable for the data analysis. What are the underlying processes which are tapped by CRR? Is it just accuracy? If so, then perhaps wouldn't it be complete if HR aren't included into it?

  • $\begingroup$ What do you mean by a process here, are you talking about a statistic such as accuracy or something else like memory, attention etc? $\endgroup$ – knk May 23 '19 at 15:00
  • $\begingroup$ If this is not part of a homework problem, please elaborate on the reason the researcher chooses to use correct rejections only. If this is part of a homework problem think hard about what a correct rejection signifies. $\endgroup$ – knk May 23 '19 at 15:03
  • $\begingroup$ Thanks for getting back. It's not part of homework. The reason why I decided to use CRR is because I ran a logistic regression model with it as a dependent variable so I needed a binary response which CRR provided (with values 1=correct; 0=incorrect).However, now I regret I did it because I think CRR does not capture "accuracy" in its most general sense. I.e. it shows that participants were some percent correct at rejecting foils. Therefore, wouldn't it be better to define accuracy as number of hits and number of correct rejections and then use it as DV for logistic regression? $\endgroup$ – Tasha May 23 '19 at 15:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.