Hugh E. McDonald’s research while affiliated with San Diego State University and other places

This study examined the effects of induced mood on susceptibility to question-framing effects. Participants were placed in either a happy, sad, or neutral mood and performed an impression-formation task under different phrasings of Martin, Ward, Achee, and Wyer’s (1993) stop rule instructions. For the enjoy rule, participants were told to stop reading behaviors either when they no longer enjoyed the task or when they became bored with the task. For the performance-based rule, participants were told to stop either when they had enough information to form an impression of the target or when they did not need to collect additional information. Results indicated that neutral mood participants were strongly influenced by the framing of the stop rule. Participants in valenced moods, however, were unaffected by framing, suggesting that they based their decisions about when to stop solely on the informational value of their moods. The implications of these results are discussed.

Previous research by Hirt, Melton, McDonald, and Harackiewicz (1996) found that mood effects on creativity were not mediated by the same mechanisms as were mood effects on quantitative measures of performance and evaluations of performance, suggesting that mood may simultaneously be working through different processes (dual process view). However, other research (Martin & Stoner, 1996; Sinclair, Mark, & Clore, 1994) supports a single process, mood-as-information model for similar effects of mood on processing. In the present research, we hypothesized that if a single, mood-as-information process accounts for mood effects on both creativity and quantitative performance, then all mood effects should be eliminated if participants are cued that their mood is irrelevant to the task (cf. Schwarz & Clore, 1983). We manipulated participants' moods prior to task performance and presented them with either an enjoyment-based or a performance-based stop rule; half of the participants were cued to the true source of their moods, half were not. Cueing participants eliminated mood effects on quantitative measures of performance (e.g., number generated). However, consistent with a dual-process view, the cueing manipulation did not affect creativity; happy participants generated the most creative responses regardless of stop rule or cue.

Two studies investigated the effects of motivational factors on expectancy use in reconstructive memory. Participants were given target's midterm grades for later recall; expectancies about the target's future performance were then manipulated. Participants' desires to see their expectancies confirmed were manipulated by making the target likable or unlikable. The authors hypothesized that when expectancy and liking "matched," participants would give significant weight to their expectancies at retrieval, resulting in expectancy-congruent distortion of the midterm grades. However, when expectancy and liking were "mismatched," expectancies would be discontinued, and participants would show little or no expectancy-congruent distortion. Results supported these predictions. Study 2 varied the order of the expectancy and liking information. Order affected the process by which mismatch participants discounted their expectancies. Results demonstrate that motivations not only may bias memory search but also may affect the reconstruction of existing memory traces.

Two studies investigated the effects of motivational factors on expectancy use in reconstructive memory. Participants were given a target's midterm grades for later recall; expectancies about the target's future performance were then manipulated. Participants' desires to see their expectancies confirmed were manipulated by making the target likable or unlikable. The authors hypothesized that when expectancy and liking ''matched,'' participants would give significant weight to their expectancies at retrieval, resulting in expectancy-congruent distortion of the midterm grades. However, when expectancy and liking were ''mismatched,'' expectancies would be discounted, and participants would show little or no expectancy-congruent distortion. Results supported these predictions. Study 2 varied the order of the expectancy and liking information. Order affected the process by which mismatch participants discounted their expectancies. Results demonstrate that motivations not only may bias memory search but also may affect the reconstruction of existing memory traces.

The authors examined the role of intrinsic interest in mediating the relationship among mood, processing goals, and task performance. Participants in induced happy, neutral, or sad moods generated similarities and differences between TV shows using performance-based, enjoyment-based, or no stop rule (cf. L. L. Martin, D. W. Ward, J. W. Achee, & R. S. Wyer, 1993). Pretask interest and both quantitative (time spent, number generated) and qualitative (creativity) performance were assessed. Happy participants spent more time and generated more items than other participants when using an enjoyment-based stop rule but spent less time and generated fewer items when using a performance-based stop rule. Happy participants also expressed greater pretask interest and were more creative than other participants regardless of stop rule. Regression-based path analyses indicated that pretask interest partially mediated the effects of mood on quantitative performance but not on creativity.

The authors examined the role of intrinsic interest in mediating the relationship among mood, processing goals, and task performance. Participants in induced happy, neutral, or sad moods generated similarities and differences between TV shows using performance-based, enjoyment-based, or no stop rule (cf. L.L. Martin, D.W. Ward, J.W. Achee, & R.S. Wyer, 1993). Pretask interest and both quantitative (time spent, number generated) and qualitative (creativity) performance were assessed. Happy participants spent more time and generated more items than other participants when using an enjoyment-based stop rule but spent less time and generated fewer items when using a performance-based stop rule. Happy participants also expressed greater pretask interest and were more creative than other participants regardless of stop rule. Regression-based path analyses indicated that pretask interest partially mediated the effects of mood on quantitative performance but not on creativity.

Two studies investigated the role of initial encoding in the relative weighting of the memory trace versus expectancy information in reconstructive memory. Ss were given a hypothetical student′s midterm grade report under either impression set, recall set, or comprehension set instructions. After a retention interval, Ss′ expectancies about the student′s future performance (improve vs decline) were manipulated. Ss then received outcome information, half consistent and half inconsistent with their expectancy. In both studies, impression set and comprehension set subjects gave significant weight to the expectancy information, recalling consistent information relatively accurately but displaying expectancy-congruent distortion of inconsistent information. Recall set subjects were unaffected by their expectancies and accurately recalled both consistent and inconsistent information. Additionally, the results of delay manipulation used in Study 2 indicated that longer retention intervals accentuated differences between encoding set conditions. Implications for the process by which expectancy information and the memory trace are integrated to determine recall are discussed.

Two studies investigated how expectancy-timing and expectancy-outcome consistency affect expectancy-guided retrieval. Ss were given a student's grade report for later recall. During the retention interval, Ss' expectancies about the student's future performance (improve vs decline) were manipulated. The expectancy information was presented either at the beginning (T1), halfway through (T2), or at the end of the retention interval (T3). Ss then received outcome information, half consistent and half inconsistent with their expectancy. In both studies, T2 and T3 Ss showed evidence of expectancy-guided retrieval, recalling consistent information accurately but displaying expectancy-congruent distortion of inconsistent information. T1 Ss showed no evidence of expectancy-guided retrieval and accurately recalled both inconsistent and consistent information. Further analyses suggested that T1 Ss appeared to reprocess the original information. (PsycINFO Database Record (c) 2012 APA, all rights reserved)