Flipbook

CHAPTER 5 126 fear does not eliminate safety behavior (Vervliet & Indekeu, 2015). However, it remains unclear whether safety behavior after fear extinction has detrimental effects on fear. This is not only theoretically, but also clinically relevant (Treanor & Barry, 2017). Using safety behavior after fear extinction logically again causes a misattribution of safety to the behavior instead of to the innocuous CS. Moreover, by signaling threat, safety behaviors may trigger a return of threat expectancy. Retention of safety behavior could then predict relapse. Accordingly, we tested whether using safety behavior after fear extinction triggers a return of fear. We used a within-subjects fear conditioning paradigm adapted from Lovibond et al. (2009) to investigate this hypothesis. Similar to Lovibond et al. (2009), the paradigm was a mix of classical and instrumental conditioning. Pavlovian acquisition took place with two CS+ (A and C) and one CS- (B). However, instead of shock, the US was a loud noise presented through headphones. Next, instrumental conditioning occurred in the Safety behavior acquisition phase. Participants learned to use safety behavior by unplugging the headphones during presentation of all three CSs, which cancelled the loud noise that would otherwise follow stimulus A and C. Then, during a fear Extinction phase, A and C were no longer followed by the US. After the Extinction phase, safety behavior was again made available during C trials, but not during A and B trials. We hypothesized that this would increase threat expectancy for C in a subsequent Test phase, in which C was presented without safety behavior. More specifically, we hypothesized that, compared to stimulus A and B, threat expectancy for stimulus C would increase from the final Extinction phase trial to the Test phase. Furthermore, we explored whether safety behavior affected evaluative conditioning in the current paradigm. Evaluative conditioning is the change in the valence of a CS after it has been paired with a US (De Houwer, Thomas, & Baeyens, 2001). A CS itself becomes more negative after it has been paired with an aversive US. Safety behavior may also directly make a stimulus more negative (see Centerbar & Clore, 2006). We therefore compared valence ratings of A, B, and C after the experimental task, and hypothesized that stimulus C would be rated more negative than stimulus A and B.