Several years ago, I conducted a study to explore how dysphoria, a transient state of feeling blue, affects the effort that individuals expend when making decisions. I was particularly interested in investigating this relationship, as too few behavioral decision theorists had incorporated emotions into their theoretical perspectives. As I read the literature, I noted that several possible relationships between dysphoria and cognitive effort had been hypothesized and/or found. Some argued that dysphorics tend to be somewhat more apathetic in their general outlook, and as such they would be expected to provide lesser decisional effort. Others proposed that dysphorics can regain a semblance of control over their lives by providing greater commitment to a decision, and as such depressed individuals might be more thorough and exhaustive when making decisions. To further confuse things, some scholars had obtained an inverted-U relationship between the two variables whereas others had uncovered the "dreaded" null effect (i.e., no relationship between dysphoria and effort). In light of the disparate findings that had been documented in the literature, it was hardly obvious what the appropriate (if any) a priori hypothesis might be.
In my study, I used a computer interface (Saad, 1998) to track participants' behaviors as they chose between competing pairs of apartments to rent. Participants (n = 84) were administered the CES-D scale to measure their levels of dysphoria and were subsequently split into two groups as a function of their scores (sample sizes for the non-depressives and depressives were 52 and 32 respectively). Sixteen dependent variables were compared across the two groups, one of which was the number of attributes that were acquired (on the competing apartments) prior to making a choice (a proxy measure of cognitive effort). Incredibly, I obtained null effects for all but one of the dependent measures. In other words, the only robust finding was the veracity of the null effects!
At the time, I had submitted the paper to a special issue on emotions in decision making hosted in a top psychology journal. The guest editor got back to me and pronounced that whereas the paper was very interesting and the methodology was highly rigorous, the extent of null effects precluded its publication. This brings me to an important point that has been raised by numerous scholars across several disciplines, namely the harm that the null effects bias causes on the advancement of science. Ultimately, if only positive results are published, it provides an inaccurate picture of the totality of actual findings for a given phenomenon.
I am thinking of reviving this paper and sending it off for publication in the near future. Wish me luck as I tackle the null effects bias! Ciao for now.