G. Bohner et T. Weinerth, Negative affect can increase or decrease message scrutiny: The affect interpretation hypothesis, PERS SOC PS, 27(11), 2001, pp. 1417-1428
Applying insights from the mood-as-information and mood-as-input models to
persuasion, it was hypothesized that negative (vs. neutral) affect would in
crease or decrease message processing depending on how recipients interpret
their affect. Recipients who are (vs. are not) likely to question the legi
timacy of the message were predicted to process to a low (vs. high) extent
when experiencing negative affect. High salience of a judgment-irrelevant c
ause for negative affect was predicted to render affect uninformative and h
ence discount its effects on processing. These hypotheses were supported in
two experiments with 388 university students, in which the likelihood of q
uestioning message legitimacy was operationalized as high versus low vested
interest (smokers vs. nonsmokers reading an antismoking appeal; Experiment
1) or by providing versus not providing a cue to the communicator being pr
opagandist (Experiment 2).