Mood effects on ingratiation: Affective influences on producing and responding to ingratiating messages

Journal of Experimental Social Psychology 76 (2018) 186–197

Contents lists available at ScienceDirect

Journal of Experimental Social Psychology journal homepage: www.elsevier.com/locate/jesp

Mood effects on ingratiation: Affective influences on producing and responding to ingratiating messages Diana Matovic, Joseph P. Forgas

T

⁎,1

University of New South Wales, Sydney, Australia

A R T I C L E I N F O

A B S T R A C T

Handling editor: Ursula Hess

Can mood influence the way people produce and respond to ingratiating messages? Based on recent affectcognition theories we demonstrate for the first time that mild negative mood increased communicators' use of ingratiatory tactics such as flattery, conformity and self-promotion (Exp. 1). Experiment 2 further confirmed that ingratiatory messages written in a negative mood were more effective and resulted in more positive interpersonal evaluations than messages written in a positive mood. Experiment 3 found that negative mood also improved recipients' willingness to accept realistic ingratiation. An analysis of response latencies (Exps. 1 and 3) and recall (Exp. 3), and mediational analyses showed that these effects were consistent with negative mood promoting longer and more attentive processing by both senders and recipients. The theoretical implications of these results for recent affect-cognition theories are considered, and the practical implications of these findings for everyday strategic communication and interpersonal behavior are discussed.

Keywords: Affect Strategic communication Ingratiation Mood Social cognition

Humans are an intensely social species, and we use language as the primary means of establishing and maintaining positive relationships with others (Dunbar, 1996). Much social communication is designed to promote cohesion and to create and maintain positive impressions to achieve our strategic objectives (E. E. Jones, 1964). Social philosophers such as Machiavelli (1532/1961) recognized almost five hundred years ago that ingratiation is a highly effective strategy for accumulating and maintaining interpersonal influence and power. 1. Strategic communication and ingratiation Classic work on ingratiation by E. E. Jones (1964) showed that there are four major strategies that ingratiators use to create positive impressions, including (1) other-enhancement, or flattery, (2) conformity in opinion and behavior, (3) self-enhancement, and (4) rendering favors. There is considerable evidence supporting E. E. Jones' (1964) model of ingratiation showing that ingratiation is an effective strategy even when the ingratiator's ulterior motives are quite obvious. For instance, restaurant customers tip more after receiving flattery (Seiter, 2007), and interviewees who use self-promotion are more likely to get a job (Proost, Schreurs, De Witte, & Derous, 2010). Surprisingly, even though affect is a fundamental dimension of interpersonal behavior (Zajonc, 1981), the influence of affective states on the way people produce, and respond to ingratiation has not been

⁎

1

studied before. Both constructing and interpreting ingratiation are highly elaborate social cognitive tasks (E. E. Jones, 1964), and thus there is good reason to assume that moods should play a significant role in how these tasks are performed (Forgas, 1995; Forgas & Eich, 2013). The absence of prior work on affect and ingratiation is all the more surprising given accumulating evidence for mood-induced processing differences in interpersonal behavior (Bodenhausen, Kramer, & Susser, 1994; Forgas, 1999, 2007, 2011a). This paper reports three experiments that demonstrate, for the first time, that negative affect can improve both the production (Exps. 1, 2), and the reception (Exp. 3) of ingratiating communications, consistent with recent evidence for the information processing consequences of mood states (Forgas, 2013). The likelihood of a bidirectional link between ingratiation and affect is also suggested by prior work showing that both producing and receiving ingratiating messages can influence subsequent mood. For example, Dunn, Biesanz, Human, and Finn (2007) found that people engaging in self-promotion also feel more positively themselves. Further, Vonk (2002) found that recipients' mood was improved when they were the target of flattery and conformity. We explore a complementary effect here, the influence of antecedent moods on subsequent ingratiation. We hypothesize that negative mood should promote a more attentive and externally oriented information processing style compared to neutral mood (Bless & Fiedler, 2006). Such increased attention and

Corresponding author. E-mail address: [email protected] (J.P. Forgas). Support from the Australian Research Council to Joseph Forgas is gratefully acknowledged.

https://doi.org/10.1016/j.jesp.2018.02.002 Received 10 October 2017; Received in revised form 22 January 2018; Accepted 3 February 2018 0022-1031/ © 2018 Elsevier Inc. All rights reserved.

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

interpersonal behaviors. Extrapolating from this literature, Experiments 1 and 2 predicted that negative mood should produce more successful ingratiating messages and greater use of all of E. E. Jones' (1964) ingratiation strategies. When receiving ingratiation in Experiment 3, negative mood should promote greater attention to ingratiatory messages, and so increase their impact. Further, both latency and memory data will be collected to index mood-induced differences in processing style (Bless & Fiedler, 2006). A mediational analysis will assess the predicted pattern of correlations between processing style and the production and interpretation of ingratiating messages.

external focus on one's partner and on communicative norms while composing an ingratiatory message should improve the effectiveness of ingratiating communication (Schlenker & Leary, 1982). The role of more attentive processing in effective ingratiation is also suggested by Leary and Kowalski's (1990) model of impression management, stating that effective ingratiation requires the careful consideration of external factors such as role constraints and target features. Consistent with this view, successful ingratiators need to engage in attentive processing in order to carefully tailor their messages to the demands of the situation (Baumeister, 1982; R. G. Jones & Jones, 1964), and consider the context, their role and the relevance of the issue to the recipient (E. E. Jones, Gergen, & Jones, 1963). Even strategies such as self-enhancement – communicating about the self - must be sensitive to external factors such as ingratiators' target audience accordingly, self-enhancement is more likely to be used with strangers than friends (Tice, Butler, Muraven, & Stillwell, 1995). Thus, a more attentive and externally oriented and accommodative processing style promoted by a negative mood (Bless & Fiedler, 2006) is predicted to enhance communicators' ingratiation effectiveness compared to neutral and positive mood. Negative mood should also enhance the effectiveness of received ingratiation, as more accommodative processing increases attention to the details of incoming messages (Bless & Fiedler, 2006). The processing benefits of mild dysphoria in promoting accommodative processing apply to many kinds of social cognitive tasks where attention to external details is beneficial (Forgas & Eich, 2013).In contrast, positive mood participants are expected to be less effective in producing, and less influenced by received ingratiation due to their less attentive, and more assimilative processing style (Bless & Fiedler, 2006).

3. Experiment 1 3.1. Method 3.1.1. Overview, design and participants Mood effects on ingratiation were explored in a realistic ‘getting to know you’ exercise using a cyber interaction and e-mail messages. Participants first received background information about their ‘partner’, Laura. Next, they viewed positive, negative or neutral mood-inducing film clips before composing an open-ended ingratiating message to Laura. Messages were rated by two independent coders for the presence of E. E. Jones' (1964) four basic ingratiating strategies. Processing latencies (the time taken to produce a message) were also recorded. A complete debriefing concluded the experiment, indicating no awareness of the mood manipulations or hypotheses. A power analysis indicated that to achieve 80% power for a smallmedium effect (d = 0.30) with three mood conditions, approximately 100 participants will be required. Accordingly, participants were 100 psychology students (63 females, 37 males; Mage = 19.39, SDage = 4.66) who received course credit for their participation. Seven participants were excluded due to poor English, technical difficulties with the computer program, or confusion about the experiment (Npos = 30, Nneut = 32, Nneg = 31). In all three experiments described here, all data collected and removed have been fully reported.

2. Affect and strategic communication Affect can exert a subtle but important influence on social behaviors by influencing the information processing strategies people employ (Forgas & Eich, 2013). Specifically, several lines of evidence indicate that negative moods can trigger a more detailed and attentive processing style, a pattern broadly consistent with Clark and Isen's (1982) mood maintenance/mood repair hypothesis, Forster and Dannenberg's (2010) global-local processing model, Fredrickson's (2001) broadenand-build theory, and Schwarz's (1990) cognitive tuning model. The processing effects of moods have recently been integrated in the assimilative-accommodative model proposed by Bless and Fiedler (2006). This model suggests that moods perform an important evolutionary signaling function, such that positive mood, signaling a safe and familiar situation, promotes a more assimilative, top-down, and internally focused processing style. In contrast, negative mood functions as a mild alarm signal, and recruits more accommodative and bottom-up processing, increasing attention to external stimulus information. A growing number of recent experiments support these predictions. For example, negative mood improved attention to Gricean communication norms and helped the detection of linguistic ambiguity in messages (Koch, Forgas, & Matovic, 2013; Matovic, Koch, & Forgas, 2014), reduced the incidence of various judgmental errors (Forgas, 2011b, 2011c), improved social memory (Forgas, Goldenberg, & Unkelbach, 2009), and produced more effective persuasive arguments (Forgas, 2007). In turn, assimilative thinking produced by positive mood (Bless & Fiedler, 2006) results in greater reliance on pre-existing internal scripts and schemas when forming impressions (Bless & Fiedler, 1995), and frees up cognitive resources for improved performance on a second task in a dual-task paradigm (Bless et al., 1996). Positive mood also promotes creativity, and the greater use of mental shortcuts and heuristics (Bodenhausen et al., 1994; Forgas, 2011b; Isen, Daubman, & Nowicki, 1987; Ruder & Bless, 2003). Thus, assimilative and accommodative processing induced by positive and negative mood appears to produce important dichotomous consequences for social thinking and

3.1.2. Mood induction Participants watched funny, neutral or sad 9-min film excerpts (Borat, a documentary about the creation of the internet, and My Life, respectively) to induce mood, described as part of a separate experiment (Fiedler, 2001; Forgas & Eich, 2013). 3.1.3. The ingratiation task Participants were instructed to engage in a short online getting-toknow-you encounter, with the objective of trying to get their partner, Laura, “to like you, by presenting as positive an impression about yourself as possible”. Participants were told that Laura is located in a neighboring room and will start the conversation by writing a short introductory e-mail about herself. Laura was described as either extraverted and not studious, or introverted and studious (see below). These counterbalanced descriptions were included to increase external validity, and were not expected to produce any main or interaction effects, as confirmed in subsequent analysis: “Hey, my name is Laura. I just started uni this semester, and I'm enjoying it so far. I'm a fairly extraverted person, who is not studious./I'm a fairly introverted person, who is studious. I'm really into working out and hanging out with my friends. I don't like cleaning or cooking. My main focus at the moment apart from uni is saving up for a trip to Europe and America. I'm hoping to go mid-year. I can't wait to travel!! Actually, maybe you could help me with this… I'm not really sure where stuff is on campus, and I need a few sheets of paper to take notes in my lecture right after this experiment. Do you know somewhere nearby where I can buy a notebook, or get a few sheets of paper? My lecture is right after this experiment… Anyway, I don't think I'm supposed to write 187

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

(2,221) = 165.38, p < .001, ηp2 = 0.60. Participants in the positive mood condition rated their mood as significantly better [Mpos = 4.78, 95% CI (4.52, 5.03), SDpos = 1.08, Mneutral = 4.43, 95% CI (4.21, 4.65), SDneutral = 0.97, t(221) = 2.06, p = .040, d = 0.34], and the negative condition felt significantly worse than the neutral group, Mneg = 1.93, 95% CI (1.68, 2.18), SDneg = 1.06, t(221) = 14.88, p < .001, d = 2.42.

a lot, so I'll leave it at that for now. Cheers, Laura”. The message was designed to give scope for participants to employ all four of E. E. Jones' (1964) ingratiation strategies by disclosing Laura's likes and dislikes, allowing participants to use (1) other-enhancement and (2) conformity. Participants were also free to use (3) self-promotion and the request for help with obtaining paper offered participants an opportunity to (4) do her a favor.

3.2.2. Preliminary analyses One rater coded all the data for each dependent variable and a second independent rater coded 20% of the data using the same criteria to establish reliability. They achieved significant intra-class correlation coefficients (ICCs) on all measures, with an overall ICC(20) = 0.85. Describing Laura as either extraverted or introverted did not influence ingratiation, as predicted; Pearson's correlations between the two versions of the introductory message on the four ingratiation categories were: rLaura,flattery(91) = −0.03, p = .810, rLaura,conformity(91) = 0.01, p = .929, rLaura,self-promotion(91) = 0.05, p = .608, rLaura,favors(91) =0.05, p = .611. There was also no evidence for any interaction between mood and the two counterbalanced descriptions of Laura on the four measures of ingratiation: flattery, F(2,87) = 0.49, p = .612, ηp2 = 0.01, conformity, F(2,87) = 0.56, p = .575, ηp2 = 0.01, self-enhancement, F(2,87) = 0.42, p = .662, ηp2 = 0.01, favors, F (2,87) = 0.74, p = .480, ηp2 = 0.02. Thus, responses to the two types of introductory messages were pooled.

3.1.4. Dependent measures The use of the four ingratiation strategies was coded by two trained raters (one rater coding all the data and the second rater coding 20% of the data to establish reliability; see below) as follows: Other-enhancement was defined and coded in terms of: (1) number of direct compliments, (2) number of questions showing interest in the partner, (3) the intimacy of those questions (rated on a 5-point scale and averaged across questions for each participant), (4) seeking social contact (0 = no interest, to 4 = high interest in contact), and (5) number of positive emoticons used. As these five measures featured different scale properties, they were first converted to z-scores and then averaged to construct a single flattery or ‘other-enhancement’ index for each participant. The second ingratiation strategy, conformity was rated as: (1) the number of instances of direct conformity with Laura's self-description, (2) conformity to general politeness norms (rated on a 7-point politeimpolite scale), (3) conformity to positivity norms (rating the message valence on a 7-point negative-postive scale), and finally (4) conformity to norms of modesty (number of self-deprecating statements; Schutz, 1997). The four measures of conformity were converted into z-scores and averaged for each participant. The third ingratiation strategy, self-enhancement was measured in terms of the number of instances of (1) direct self-promotion, as well as self-descriptive statements indicating (2) immediacy, (3) independence and autonomy, and (4) positivity, (5) intimacy and (6) breadth of selfdisclosure (rated on 7-point bipolar scales for each self-disclosing statement and averaged for each participant; see also Forgas, 2011a). The inclusion of self-disclosure as a form of self-enhancement is also supported by a meta-analysis demonstrating that more intimate and broad self-disclosures are ingratiating (Collins & Miller, 1994). The six measures of self-promotion were also converted into z-scores and averaged to create a ‘self-enhancement’ index for each participant. The fourth ingratiating strategy, rendering favors in response to Laura's needing notepaper was assessed on a 5-point scale, from ‘no help offered’ to ‘maximum help offered’ (e.g., offering paper or escorting Laura to a stationery shop). As these four ingratiation categories are conceptually distinct and were based on a priori theorizing, they were analyzed separately. The correlations between these measures were small or non-existent, indicating that the measures were complementary rather than redundant: rflattery,conform(91) = 0.22, p = .038, rflattery,self-enhance(91) = 0.06, p = .597, rflattery,favors(91) = 0.25, p = .017, rconform,self-enhance(91) = 0.29, p = .005, rconform,favors(91) = 0.10, p = .334, rself-enhance,favors(91) = −0.11, p = .285.

3.2.3. Mood effects on ingratiation strategies Given multiple dependent variables, a one-way MANOVA of mood effects on the four ingratiatory categories was first conducted to control the Type 1 error rate (Cramer & Bock, 1966), confirming that mood had a significant overall influence on ingratiation, Pillai's Trace = 0.20, F (8,176) = 2.50, p = .014, ηp2 = 0.10. Next, a series of one-way ANOVAs explored mood effects on the use of each of the four ingratiation categories. As expected, mood significantly influenced the use of flattery or other-enhancement, F(2,90) = 3.27, p = .042, ηp2 = 0.07, such that participants in a negative mood employed significantly more flattery [Mneg = 0.19, 95% CI (−0.01, 0.38), SDneg = 0.53] than in the positive [Mpos = −0.10, 95% CI (−0.26, 0.06), SDpos = 0.43, t(90) = 2.22, p = .029, d = 0.57] and neutral mood conditions, Mneut = −0.09, 95% CI (−0.28, 0.10), SDneut = 0.53, t(90) = 2.21, p = .029, d = 0.56, with no significant difference between the positive and neutral groups, t (90) = 0.04, p = .969, d = 0.01 (see Fig. 1). Mood also significantly influenced the use of conformity, F (2,90) = 4.99, p = .009, ηp2 = 0.10. Those in a negative mood used more conformity [Mneg = 0.19, 95% CI (−0.01, 0.38), SDneg = 0.54] than did those in a positive mood [Mpos = −0.16, 95% CI (−0.27, −0.05), SDpos = 0.29, t(90) = 3.12, p = .002, d = 0.80] and a neutral mood [Mneut = −0.03, 95% CI (−0.19, 0.12), SDneut = 0.43, t (90) = 2.02, p = .047, d = 0.51], and there was no significant difference between the positive and neutral conditions, t(90) = 1.14, p = .257, d = 0.29 (Fig. 1). Self-enhancement was also significantly influenced by mood, F (2,90) = 4.19, p = .018, ηp2 = 0.09. As predicted, participants in a negative mood employed significantly more self-promotion [Mneg = 0.22, 95% CI (0.01, 0.43), SDneg = 0.58] than did those in the positive [Mpos = −0.12, 95% CI (−0.28, 0.04), SDpos = 0.42, t (90) = 2.55, p = .012, d = 0.65] and neutral mood conditions, Mneut = −0.10, 95% CI (−0.30, 0.09), SDneut = 0.54, t(90) = 2.46, p = .016, d = 0.62. There was no significant difference between positive and neutral groups, t(90) = 0.14, p = .890, d = 0.04 (Fig. 1). We found no significant mood effects on the fourth measure of ingratiation, rendering favors, Mneg = 1.48, 95% CI (1.00, 1.97), SDneg = 1.31, Mneut = 1.75, 95% CI (1.27, 2.23), SDneut = 1.32, Mpos = 1.53, 95% CI (1.06, 2.01), SDpos = 1.28, F(2,90) = 0.37, p = .691, ηp2 = 0.01, possibly because this strategy required much

3.2. Results 3.2.1. Validation of the mood induction The effectiveness of the mood induction was established in a separate pilot study (N = 224). A separate sample drawn from the same population was used, as collecting these judgments from the experimental sample after a complex and involving ingratiation exercise with likely affective consequences (Dunn et al., 2007) may have confounded their responses. Participants in the mood validation sample viewed the same mood induction videos and then rated their mood on a 7-point sad-happy scale immediately after the mood induction. For these and subsequent analyses, partial eta squared and Cohen's d are used as estimates of effect sizes rather than confidence intervals. An ANOVA confirmed that the mood induction was highly effective, F 188

Journal of Experimental Social Psychology 76 (2018) 186–197

enhancement

D. Matovic, J.P. Forgas

3.2.5. Mediational analyses The theoretically predicted relationship between processing style and the production of ingratiating messages was assessed in mediational analyses using a bootstrapping approach (Preacher & Hayes, 2004). For other-enhancement, the standardized regression coefficient between mood (dummy coded to compare negative mood to both neutral and positive mood) and processing latencies was significant, β = 0.39, 95% CI (0.19, 0.56), p < .001, as was the standardized regression coefficient between processing latency and other-enhancement, β = 0.27, 95% CI (0.08, 0.47), p = .009 (see Fig. 2). The standardized indirect effect was β = 0.39 × 0.27 = 0.11, 95% CI (0.03, 0.24), and was tested for significance using bootstrapping procedures. Standardized indirect effects were computed for each of 2000 bootstrapped samples, showing a significant indirect effect of mood on flattery via processing latencies, p = .005. The direct effect of mood on flattery changed from significant, β = 0.26, 95% CI (0.04, 0.44), p = .010, to non-significant, β = 0.16, 95% CI (−0.08, 0.36), p = .140, when the mediator was also included (Fig. 2). This suggests that processing latencies may fully mediate the mood effect, consistent with our theorized model, although such patterns need to be treated with caution, as other causal factors may also influence the tri-variate pattern reported (Fiedler, Schott, & Meiser, 2011). For conformity, the standardized regression coefficient between mood and processing latencies was significant, β = 0.39, 95% CI (0.19, 0.56), p < .001, but the standardized regression coefficient between latencies and conformity was not, β = 0.15, 95% CI (−0.07, 0.36), p = .174. The standardized indirect effect, β = 0.39 × 0.15 = 0.06, 95% CI (−0.02, 0.17), computed over 2000 bootstrapped samples, was not significant, p = .132, although there was a trend in the predicted direction. For self-enhancement, both the standardized regression coefficients between mood and processing latencies, β = 0.39, 95% CI (0.19, 0.56), p < .001, and latencies and self-enhancement were significant, β = 0.41, 95% CI (0.21, 0.58), p < .001 (Fig. 2). The standardized indirect effect, β = 0.39 × 0.41 = 0.16, 95% CI (0.07, 0.30), computed over 2000 bootstrapped samples, was significant, p < .001, indicating that mood may influence self-enhancement through processing latencies, consistent with more accommodative processing in a negative mood (Bless & Fiedler, 2006). The inclusion of the mediator reduced the effect of mood on self-enhancement from significant, β = 0.29, 95% CI (0.04, 0.49), p = .003, to non-significant, β = 0.14, 95% CI (−0.12, 0.37), p = .175 (Fig. 2). These results are consistent with a theorized model that predicts that mood effects on flattery, conformity and self-promotion are significantly mediated by processing style, with negative mood recruiting slower and more externally oriented and accommodative processing (Bless & Fiedler, 2006). However, caution is needed when interpreting meditational analyses. Processing latencies may be also influenced by additional, extraneous variables, and mediational analysis can only be suggestive in revealing essentially correlational links, rather than causally effective mediators (Fiedler et al., 2011). To further explore this pattern, in Experiment 3 memory data reflecting qualitative rather than quantitative differences in processing style (Bless & Fiedler, 2006) will be used as a further index of information processing styles.

0.2 0.1

Posive

0.0

Neutral Negave

-0.1 -0.2

Posive Neutral Negave

Conformity

0.3 0.2 0.1

Posive

0.0

Neutral

-0.1

Negave

-0.2 -0.3

Self-enhancement

Posive

Neutral Negave

0.3 0.2 Posive

0.1

Neutral

0.0

Negave

-0.1 -0.2

Posive

Neutral Negave

Fig. 1. Standardized scores and standard errors for mood effects on flattery (other-enhancement; top figure), conformity (middle figure), and self-enhancement (bottom figure): the use of all three ingratiating strategies was promoted by negative compared to positive and neutral mood.

greater investment of time and effort in the context of the present situation. 3.2.4. Mood effects on processing latencies Mood also significantly influenced processing latencies, F (2,90) = 8.09, p = .001, ηp2 = 0.15. As expected, consistent with more attentive processing, participants in a negative mood took significantly longer to produce their ingratiating messages than did positive [Mpos = 6 min 10s, 95% CI (5 min 8 s, 7 min 11 s), SDpos = 2 min 44 s, Mneg = 9 min 00s, 95% CI (7 min 44 s, 10 min 16 s), SDneg = 3 min 27 s, t(90) = 3.75, p < .001, d = 0.96] and neutral mood participants [Mneutral = 6 min 40s, 95% CI (5 min 44 s, 7 min 36 s), SDneutral = 2 min 34 s, t(90) = 3.13, p = .002, d = 0.79]. There was no significant difference between the positive and neutral conditions [t(90) = 0.67, p = .502, d = 0.17]. The longer processing in negative mood was not due to negative mood participants simply writing more (using more words). An ANOVA of text length showed that there was no significant difference in word count between conditions, Mneg = 150.16 words, 95% CI (136.69, 163.63), SDneg = 36.72 words, Mneut = 143.22, 95% CI (126.37, 160.06), SDneut = 46.72, Mpos = 128.13, 95% CI (114.27, 142.00), SDpos = 37.14, F(2,90) = 2.34, p = .102, ηp2 = 0.05. Thus, mood influenced the latency and quality of information processing rather than the quantity (word count) of the messages produced (Bless & Fiedler, 2006).

3.3. Discussion This experiment was successful in demonstrating for the first time that mood states have a significant influence on the use of ingratiation strategies. In particular, the results show that negative mood promoted the more extensive use of other-enhancement, conformity and self-enhancement. We also found that negative mood resulted in longer processing latencies, consistent with a more accommodative and attentive processing strategy. Further, these mood-induced processing differences significantly correlated with ingratiating strategies in mediational 189

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

Latencies = .39***

Latencies = .27**

= .15ns

= .39***

= .26** ( = .16ns)

= .30** ( = .24*)

Fig. 2. Mediational models of affective influences on the production of flattery, conformity and self-promotion (Experiment 1): negative mood produced longer response latencies that were positively associated with more extensive use of otherenhancement and self-enhancement in communicators' messages, consistent with the more accommodative processing of these messages when in negative mood (Bless & Fiedler, 2006). The standardized regression coefficients between mood and the three ingratiatory strategies, controlling for processing latencies, are in parentheses. ns = not significant; * p < .05; ** p ≤ .01; *** p < .001.

Latencies = .39***

Mood

= .41***

= .29** ( = .14ns)

Self-promotion

SELF-ENHANCEMENT models. Possible alternative explanations (such as the negative film inducing mortality salience) will also be considered in the general discussion. Experiment 1 only showed that mood can influence the use, but not necessarily the quality of ingratiation. Experiment 2 will assess the effects of mood on the actual effectiveness of ingratiating messages produced.

previous getting-to-know-you experiment, and formed impressions of the writers. Participants also read the introductory messages from Laura used in Experiment 1. Based on extensive evidence suggesting that social evaluation and task evaluation, or competence are basic dimensions of impression formation (E. E. Jones, 1964), social evaluation was measured using three 6-point bipolar scales: (1) like/dislike, (2) friendly/ unfriendly, and (3) honest/dishonest. Competence was also assessed on three 6-point bipolar scales: (1) competent/incompetent, (2) self-confident/shy, and (3) extraverted/introverted. The order and polarity of the scales was counterbalanced.

4. Experiment 2 Experiment 2 examined the relative efficacy of the ingratiating messages produced by communicators in negative versus positive mood, by asking naïve participants to form impressions of messages written by negative and positive mood communicators.

4.2. Results 4.2.1. The effects of ingratiators' mood on impressions The scales measuring social evaluation (α = 0.79) and perceived competence (α = 0.71) were highly correlated for each construct and were averaged for further analysis. Ratings for the six messages written by communicators experiencing a negative or a positive mood were also averaged for these two dependent variables. A MANOVA was first conducted to control the Type 1 error rate (Cramer & Bock, 1966), indicating a significant mean difference in impressions due to the mood of the communicator, Pillai's Trace = 0.52, F(2,40) = 21.92, p < .001, ηp2 = 0.52. Although impressions on the social evaluation and competence measures were correlated of r(40) = 0.52, p < .001, they were analyzed separately based on strong a priori evidence for their complementary nature. Follow-up ANOVAs indicated that messages written by negative mood ingratiators produced significantly more favorable impressions on both the social evaluation [Mneg = 4.91, 95% CI (4.77, 5.04), SDneg = 0.44, Mpos = 4.64, 95% CI (4.50, 4.79), SDpos = 0.47, F (1,41) = 28.52, p < .001, ηp2 = 0.41] and the competence dimensions [Mneg = 4.60, 95% CI (4.44, 4.76), SDneg = 0.51, Mpos = 4.20, 95% CI (4.05, 4.36), SDpos = 0.51, F(1,41) = 27.73, p < .001, ηp2 = 0.40] compared to messages written by ingratiators experiencing a positive mood (Fig. 3). This result supports our prediction that negative mood produced not only more extensive, but also higher quality and more effective ingratiating messages, consistent with negative mood promoting a more externally oriented, attentive and accommodative

4.1. Method 4.1.1. Overview, design and participants The effectiveness of 12 randomly selected gender representative (3 males, 3 females per mood condition) ingratiatory messages written by positive and negative mood communicators in Experiment 1 was assessed by asking naïve recipients to form impressions of these writers. The introductory message by “Laura” was also included to provide participants with a standard communicative context. Each participant read all 12 messages in random order, in a single-factor within-subjects design, with author mood (positive, negative) as the independent variable. Impressions on six counterbalanced scales assessed impressions on social evaluation and competence as the dependent variables. A power analysis showed that to achieve 80% power for the same effect size as found in Exp. 1 (d = 0.50) in a two-group design a sample of at least 34 participants was needed. Actually 42 participants were run to increase power (23 females, 19 males; Mage = 19.05, SDage = 1.36) who participated in the study for course credit. 4.1.2. The impression formation task Participants were asked to read 12 randomly selected messages, six written by positive mood and six by negative mood participants in the 190

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

4.7

also the quality and focus of information processing. Thus, negative mood should enhance attention to external details and the recall of the received ingratiatory information. The ingratiatory messages in Experiment 3 will contain both selfrelated (information about the participant) and other-related (information about the partner) details. Consistent with more assimilative processing and self-directed attention, positive mood is predicted to improve selective recall of self-related information compared to neutral mood (Bless & Fiedler, 2006). Negative mood in turn should promote better recall of other-related information about the partner compared to neutral mood due to a more externally oriented, accommodative information processing style (Bless & Fiedler, 2006; Forgas et al., 2009; Forgas, Laham, & Vargas, 2005; Matovic et al., 2014; Storbeck & Clore, 2005).

4.6

5.1. Method

4.9 4.8

Posive

4.7

Negave

4.6 4.5

Competence

Posive

Negave

4.5 4.4

Posive

4.3

Negave

5.1.1. Overview, design and participants The effects of mood (positive, neutral, negative) on perceptions of an ingratiator were assessed on impression formation measures of social evaluation and competence, as in Experiment 2. Participants believed that they were exchanging e-mail messages with a partner in an online interaction. To increase the realism of the encounter, they also witnessed a female “partner” (in fact, a confederate) named Olga entering one of the neighboring cubicles. After an audiovisual mood induction, participants read an ingratiating e-mail, supposedly from Olga, containing all four of E. E. Jones' (1964) ingratiation strategies. They next rated Olga and her message. A power analysis indicated that consistent with Experiments 1 and 2 above, to achieve 80% power for small-medium effect (d = 0.35), approximately 27 participants were needed in each mood condition. Data were ultimately collected from 79 participants (46 females, 33 males; Mage = 18.94, SDage = 2.14) who participated in the experiment for course credit. Two participants were excluded due to reported confusion about the procedure (Npos = 25, Nneut = 24, Nneg = 28).

4.2 4.1 Posive

Negave

Fig. 3. The effectiveness of ingratiatory messages written by communicators in a negative versus a positive mood (Experiment 2): Negative mood produces ingratiatory messages that result in more favorable ratings of the communicators on measures of social evaluation and competence than does positive mood, reflecting more effective ingratiation.

processing style (Bless & Fiedler, 2006). 4.3. Discussion Experiment 2 was successful in demonstrating that messages written in a negative mood were actually more effective in ingratiating the communicator, indicated by more favorable evaluations of the communicator on both the social evaluation and competence measures. These findings complement the results in Experiment 1, showing that negative mood also resulted in the more extensive use of other-enhancement, conformity and self-enhancement than did positive mood. Experiment 2 established that messages written in a negative mood were also more effective and more successful in changing impressions. However, as yet little is known about the complementary question of how the mood of the recipient may influence the way ingratiating communication is received and interpreted. In the third experiment, we examined affective influences on recipients' evaluations of ingratiation, and sought to obtain further evidence about the role of mood-induced differences in processing strategies in producing these effects.

5.1.2. Mood induction In experiments using an audio-visual mood induction, there is a possibility that video content might confound mood effects. Using thematically different videos in a related series of studies helps to control for such confounds, and triangulate the underlying induced mood effect (Forgas, 1995; Westermann, Spies, Stahl, & Hesse, 1996). Accordingly, a new set of thematically different 8 min. mood induction videotapes were used here to induce mood (excerpts from a British comedy, Fawlty Towers, a documentary about minerals, and excerpts from Philadelphia, respectively), described as part of a separate experiment. 5.1.3. The ingratiation task Participants engaged in a short getting-to-know-you interaction with a “randomly assigned” partner named Olga whom they saw entering a neighboring cubicle. Participants were first asked to provide some background information about their personality (Ten Item Personality Inventory; Gosling, Rentfrow, & Swann, 2003) and values (a modified, short Schwartz value inventory; Lindeman & Verkasalo, 2005) to their ‘partner’, as a first step in the acquaintance process. They then underwent a mood induction while supposedly waiting for Olga to respond. Participants then received an ingratiatory e-mail from Olga featuring all four of E. E. Jones' (1964) strategies:

5. Experiment 3 Experiment 3 examined the complementary hypothesis that negative mood may also enhance the effects of ingratiation on recipients, by promoting more accommodative processing and greater attention to the details of plausible and realistic ingratiating messages (Bless & Fiedler, 2006; Forgas & East, 2008). Positive mood in turn should promote a less attentive and more assimilative, internally oriented processing style, potentially reducing message impact (Forgas & Eich, 2013). Processing latencies will also be measured here to evaluate mood-induced differences in processing style. Slower, more accommodative processing is expected in a negative mood, and faster, more assimilative processing predicted in a positive mood. The assimilation/accommodation model (Bless & Fiedler, 2006) predicts that negative mood should influence not only the duration, but

“Hi, I'm Olga. I have seen the information you completed about yourself. I noticed that we have the same ideas about many things. Especially on our values, we are very much alike. It seems to me that you are an easy person to get along with, and I'm sure we will have a lot to talk about. A bit about me… I too prefer a certain amount of change and variety, and become unhappy when hemmed in by restrictions and limitations. At 191

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

correlated (α = 0.94), and were combined into a single mood valence measure. An ANOVA indicated that the mood induction was highly effective, F(2,74) = 34.73, p < .001, ηp2 = 0.48; contrasts confirmed that participants in the positive mood condition rated their mood as significantly better [Mpos = 5.88, 95% CI (5.35, 6.41), SDpos = 1.28, Mneutral = 3.90, 95% CI (3.48, 4.31), SDneutral = 0.98, t(74) = 5.38, p < .001, d = 1.54], and the negative condition felt significantly worse than the neutral group [Mneg = 2.96, 95% CI (2.38, 3.55), SDneg = 1.51, t(74) = 2.60, p = .011, d = 0.72].

times I am extroverted and sociable, while at other times I am introverted and reserved. I love hanging out with my friends, and I consider myself a good friend. I feel like I have settled into uni pretty well. I have learnt my way around campus, the library, etc. If you need any help with Psych 1A or getting around, let me know. It's always nice to make new friends, and you seem like a great person to hang out with! Looking forward to hearing back from you!” Thus, this message included flattery (e.g., “you seem like a great person to hang out with!”), conformity (e.g., “Especially on our values, we are very much alike”), self-promotion (i.e., “I consider myself a good friend”), and a favor (i.e., “If you need any help with Psych 1A or getting around, let me know”). These statements were partly derived from the Barnum/Forer effect study (1949) so as to include vague but plausible self-descriptions that would almost certainly be true because of their generality (e.g., “I too prefer a certain amount of change and variety, and become unhappy when hemmed in by restrictions and limitations”). Processing latencies (time taken to read Olga's message) were also recorded.

5.2.2. Mood effects on responding to ingratiatory messages The two sets of scales measuring social evaluation (α = 0.82) and competence (α = 0.58) of the ingratiator were highly correlated and were averaged for each measure for further analysis. First, a MANOVA was conducted to control for the Type 1 error rate (Cramer & Bock, 1966), indicating a significant overall mood effect on impressions, Pillai's Trace = 0.14, F(4,148) = 2.76, p = .030, ηp2 = 0.07. Follow-up ANOVAs confirmed that mood had a significant influence on impressions on the social evaluation dimension [F(2,74) = 3.37, p = .040, ηp2 = 0.08]. The ingratiator was liked significantly more by participants experiencing a negative compared to a positive [Mneg = 5.11, 95% CI (4.88, 5.33), SDneg = 0.57, Mpos = 4.65, 95% CI (4.34, 4.96), SDpos = 0.75, t(74) = 2.39, p = .019, d = 0.66; see Fig. 4] and neutral mood [Mneutral = 4.72, 95% CI (4.41, 5.04), SDneutral = 0.75, t(74) = 2.01, p = .048, d = 0.56]. There was no significant difference between the positive and neutral conditions [t (74) = 0.35, p = .728, d = 0.10]. Mood also significantly influenced impressions on the competence measure [F(2,74) = 5.17, p = .008, ηp2 = 0.12; Fig. 4]. The ingratiator was percieved as more competent by recipients in a negative versus a positive [Mneg = 4.89, 95% CI (4.70, 5.08), SDneg = 0.49, Mpos = 4.52, 95% CI (4.29, 4.75), SDpos = 0.56, t(74) = 2.81, p = .006, d = 0.77] and neutral mood [Mneutral = 4.53, 95% CI (4.38, 4.68), SDneutral = 0.36, t(74) = 2.70, p = .009, d = 0.75], in line with more accommodative processing of ingratiating information in a negative

5.1.4. Validation of the stimulus message The face validity and realism of these ingratiating messages was established with the same participants used in Experiment 2, who judged it as comparable to the 12 authentic ingratiatory messages used in Experiment 2 on measures of social evaluation and competence. Responses showed that the ingratiating message by “Olga” did not differ significantly from the average of the 12 real participants' messages on measures of social evaluation [MOlga = 4.85, 95% CI (4.62, 5.08), SDOlga = 0.74, Mothers = 4.77, 95% CI (4.64, 4.90), SDothers = 0.42, F(1,41) = 0.72, p = .402, ηp2 = 0.02] and competence [MOlga = 4.25, 95% CI (4.01, 4.49), SDOlga = 0.77, Mothers = 4.40, 95% CI (4.26, 4.54), SDothers = 0.44, F(1,41) = 2.26, p = .140, ηp2 = 0.05], providing empirical support for its realism and plausibility. 5.1.5. Dependent measures Reactions to the ingratiating message and impressions of ‘Olga’ were again assessed on dimensions of social evaluation and competence (E. E. Jones, 1964). Social evaluation was assessed on three 6-point ‘not at all’ to ‘very much’ scales in terms of (1) likeability, (2) friendliness, and (3) honesty. Competence was assessed on 6-point scales measuring perceived (1) competence, (2) self-confidence, (3) strength, and (4) extroversion. Question order was randomized.

5.2 5.0 Posive 4.8

Neutral Negave

4.6

5.1.6. Recall task After the impression formation judgments, participants were asked to try to remember the details of the message received from Olga. One rater coded all the data for each dependent variable and a second independent rater coded 20% of the data to establish reliability. Two memory measures were calculated for each participant: (1) recall of self-related information (e.g., “she thinks I'd be easy to get along with”), (2) recall of other-related information (e.g., she “considers herself a good friend”).

4.4

Posive

Neutral Negave

5.0 4.9 4.8

Posive

4.7

5.1.7. Mood validation and debriefing As part of the debriefing, participants rated how they felt after the mood induction on 7-point good-bad and happy-sad scales, embedded among several distractor items (e.g., “Have you seen the film before?” “Would you recommend the film to a friend based on the excerpt?”). A comprehensive debriefing concluded the experiment, revealing no knowledge of the mood manipulations or hypotheses.

Neutral

4.6

Negave

4.5 4.4

Posive

5.2. Results

Neutral Negave

Fig. 4. Mood effects on impressions based on ingratiating messages (Experiment 3): Recipients of ingratiation in a negative mood formed more positive impressions in terms of their social evaluation and competence ratings of the ingratiator compared to those in a positive and neutral mood.

5.2.1. Validation of the mood induction Self-ratings on the happy-sad and good-bad scales were highly 192

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

Mood had the opposite effect on recall of other-related information, F (2,74) = 6.34, p = .003, ηp2 = 0.15 (Fig. 5). Participants in a negative mood remembered more other-related information than did positive [Mpos = 5.64, 95% CI (4.78, 6.50), SDpos = 2.08, Mneg = 7.50, 95% CI (6.67, 8.33), SDneg = 2.13, t(74) = 3.39, p = .001, d = 0.93] and neutral mood participants [Mneutral = 6.08, 95% CI (5.36, 6.81), SDneutral = 1.72, t(74) = 2.55, p = .013, d = 0.71]. Greater attention to external information about others is consistent with more accommodative processing (Bless & Fiedler, 2006). There was no significant difference between the positive and neutral conditions, t(74) = 0.78, p = .439, d = 0.22. This pattern of mood effects on recall selectively supports the unique processing predictions proposed by the assimilation-accommodation model (Bless & Fiedler, 2006). Such self/other differences in processing attention are not suggested by theories such as the global-local processing model (Forster & Dannenberg, 2010) or Fredrickson's (2001) broaden-and-build model.

mood (Bless & Fiedler, 2006). There was no difference between the positive and neutral conditions, t(74) = 0.08, p = .935, d = 0.02. 5.2.3. Mood effects on processing latencies Mood had the predicted significant effect on processing latencies, F (2,74) = 5.53, p = .006, ηp2 = 0.13. Participants took longer to process the ingratiating message in negative than in positive [Mpos = 31.89 s, 95% CI (28.96, 34.82), SDpos = 7.10s, Mneg = 43.02 s, 95% CI (35.56, 50.49), SDneg = 19.24 s, t(74) = 3.13, p = .003, d = 0.86] or neutral mood [Mneutral = 34.17 s, 95% CI (31.15, 37.20), SDneutral = 7.17 s, t (74) = 2.46, p = .016, d = 0.68], with no significant difference between the positive and neutral mood conditions, t(74) = 0.62, p = .540, d = 0.18. This finding confirms the predicted slower, more detail oriented and accommodative processing style in a negative mood (Bless & Fiedler, 2006). 5.2.4. Recall The two raters produced significant and high intraclass correlations for both measures of recall [ICCself(16) = 0.86, p < .001, ICCother(16) = 0.90, p < .001], confirming the reliability of the rating procedure. A MANOVA was first conducted to help protect against inflating the Type 1 error rate (Cramer & Bock, 1966), indicating a significant mood effect on recall, Pillai's Trace = 0.24, F(4,148) = 5.07, p = .001, ηp2 = 0.12. Mood had a significant effect on the recall of self-related information, F(2,74) = 5.98, p = .004, ηp2 = 0.14 (Fig. 5). Consistent with assimilative processing (Bless & Fiedler, 2006), participants in a positive mood remembered more self-relevant information compared to those experiencing a negative [Mpos = 0.84, 95% CI (0.43, 1.25), SDpos = 0.99, Mneg = 0.18, 95% CI (0.03, 0.33), SDneg = 0.39, t (74) = 3.40, p = .001, d = 0.93] or neutral mood [Mneutral = 0.38, 95% CI (0.10, 0.65), SDneutral = 0.65, t(74) = 2.30, p = .024, d = 0.65], with no difference between the negative and neutral conditions [t (74) = 1.00, p = .322, d = 0.28].

5.2.5. Mediational analyses The correlations between the two indicators of processing style (response latency and recall performance) and responding to ingratiation were assessed using a mediational analysis relying on a bootstrapping approach testing the indirect effects (Preacher & Hayes, 2004). 5.2.5.1. Response latency as a mediator. For social evaluation judgments, the standardized regression coefficients were significant both between mood (dummy coded) and processing latencies, β = 0.35, 95% CI (0.13, 0.52), p < .001, as well as between latencies and social evaluation β = 0.34, 95% CI (0.15, 0.49), p = .003 (see Fig. 6). The standardized indirect effect, β = 0.35 × 0.34 = 0.12, 95% CI (0.03, 0.23), computed over 2000 bootstrapped samples, indicated that results were consistent with the model predicting that processing latencies should mediate mood effects on social evaluation judgments, p = .004. The effect of mood on social evaluation, β = 0.29, 95% CI (0.08, 0.46), p = .009, became non-significant (Fig. 6) when the mediator was also included in the model, β = 0.17, 95% CI (−0.03, 0.36), p = .132. Thus, processing latencies may fully mediate the mood effect, as predicted by our model, although the observed correlational pattern may also be consistent with other models. (Fiedler et al., 2011) For competence judgments, the standardized regression coefficient between mood and latencies was significant, β = 0.35, 95% CI (0.13, 0.52), p < .001, as was the path from latencies to perceptions of competence, β = 0.22, 95% CI (0.00, 0.45), p = .047 (Fig. 6). The standardized indirect effect, β = 0.35 × 0.22 = 0.08, 95% CI (0.00, 0.21), computed over 2000 bootstrapped samples, indicated that results were consistent with the model predicting that processing latencies significantly mediate mood effects on competence judgments, p = .052. The direct effect of mood on competence, β = 0.35, 95% CI (0.14, 0.52), p = .001, remained significant when the mediator was included in the model, β = 0.27, 95% CI (0.03, 0.48), p = .015 (Fig. 6), suggesting that accommodative processing may partially mediate the effect of mood on the perceived competence of ingratiators.

recall

7.5 7.0

Posive

6.5

Neutral

6.0

Negave

5.5 5.0

Posive Neutral Negave 1.0 0.8 Posive

0.6

Neutral

0.4

5.2.5.2. Recall performance as a mediator. To construct a more sensitive index of assimilative vs. accommodative processing, the difference between self-relevant and other-related recall was used as the processing measure (Bless & Fiedler, 2006). For social evaluation, the standardized regression coefficient between mood and memory was significant, β = 0.42, 95% CI (0.23, 0.60), p < .001, while the path from memory to social evaluation fell short of significance, β = 0.20, 95% CI (−0.01, 0.39), p = .089 (Fig. 6). Nevertheless, the standardized indirect effect, β = 0.42 × 0.20 = 0.08, 95% CI (0.01, 0.20), computed over 2000 bootstrapped samples, was consistent with the predicted model that mood effects on social evaluation ratings may be fully

Negave

0.2 0.0

Posive

Neutral Negave

Fig. 5. Mood effects on the recall of ingratiating messages (Experiment 3): negative mood resulted in the selectively better recall of other-related information (top figure), and positive mood resulted in the better recall of self-related information (bottom figure), consistent with more assimilative processing in positive mood, and more accommodative processing in negative mood.

193

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

SOCIAL EVALUATION

COMPETENCE

Latencies

Latencies

= .35***

= .34**

= .35***

= .29** ( = .17ns)

= .22*

Fig. 6. Convergent correlations consistent with the theorized mediation of mood effects on impressions of the ingratiator: both processing latency (top figures), and recall performance (bottom figures) were positively associated with social evaluation, and latencies also correlated with the perceived competence of the ingratiator. The standardized regression coefficients between mood and the dependent variables, controlling for the mediators, are in parentheses. ns = not significant; * p < .05; ** p < .01; *** p ≤ .001.

= .35*** ( = .27*)

MEDIATION BY PROCESSING LATENCY

SOCIAL EVALUATION

COMPETENCE

Recall

Recall = .20ns

= .42***

= .29** ( = .20ns)

= .15ns

= .42***

= .35*** ( = .29*)

MEDIATION BY RECALL PERFORMANCE cognition and behavior (Bless & Fiedler, 2006; Forgas, 1995; Matovic et al., 2014), insufficient attention has been paid to mood effects on strategic communication, such as the production and interpretation of ingratiating messages. These three experiments provide the first empirical evidence that negative mood may promote both the more effective construction, and the more attentive processing of ingratiating messages. These findings have several intriguing theoretical and practical implications.

mediated by accommodative processing, p = .032 (Fig. 6), βnoMediator = 0.29, 95% CI (0.08, 0.46), p = .009, βwithMediator = 0.20, 95% CI (−0.02, 0.39), p = .088. For competence judgments, the standardized regression coefficient between mood and recall was significant, β = 0.42, 95% CI (0.23, 0.60), p < .001, but the regression coefficient between recall and competence ratings was not, β = 0.15, 95% CI (−0.07, 0.34), p = .187. The standardized indirect effect, β = 0.42 × 0.15 = 0.06, 95% CI (−0.01, 0.17), computed over 2000 bootstrapped samples, was not significant, p = .095, although there was a clear trend in the predicted direction.

6.1. Theoretical implications Three experiments found that negative mood leads to the more extensive and effective use of other-enhancement, conformity and selfenhancement as ingratiation strategies (E. E. Jones, 1964; Exps. 1 & 2), and the more thorough processing of ingratiating messages by recipients resulting in more positive impressions (Exp. 3). This pattern is consistent with negative mood promoting an externally oriented, accommodative processing style (Bless & Fiedler, 2006). The absence of significant mood effects on the fourth ingratiation strategy, rendering favors, suggests that such more costly and demanding behavioral responses may be less open to mood effects in this experimental situation (Forgas, 1995). Thus, participants may have preferred the other three communicative strategies as equally effective but less costly avenues for ingratiation. The correlational patterns obtained in the mediational analyses were consistent with the model predicting process mediation of mood effects on ingratiation, as negative mood resulted in longer and more attentive processing. Response latencies were positively correlated with both producing (Exp. 1) and responding to (Exp. 3) ingratiation, and recall data further confirmed these relations (Exp. 3). Experiment 3 also provided specific evidence that positive mood participants focused more on, and remembered self-related information, consistent with

5.3. Discussion Experiment 3 showed that negative mood significantly enhanced the impact of the ingratiating messages received on impression formation, consistent with recipients' expected more attentive and accommodative processing style (Bless & Fiedler, 2006). Further, the longer processing latencies and better recall of messages indicating more accommodative processing were also positively correlated with more favorable reactions to ingratiation, consistent with the theorized mediational model (Bless & Fiedler, 2006). While causal inferences cannot be made based on correlational mediational analyses due to the possibility that processing latencies and recall may be influenced by variables other than mood (Fiedler et al., 2011), the convergence of results suggests that more accommodative thinking in a negative mood may improve the processing of, and subsequent impact of ingratiation (Bless & Fiedler, 2006). 6. General discussion Despite mounting evidence for the critical role of affect in social 194

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

Experiments 1 and 3 here. As the mood induction film used in Experiment 1 (My Life) featured contents relevant to mortality salience (MS), could Terror Management Theory (TMT) offer a possible alternative explanation? Actually, the “vast majority of TMT studies show no effects on self-report measures of mood”, indicating that there is no direct link between mood states and mortality salience effects (T. Pyszczynski, 2017, personal communication; Greenberg et al., 2003). Further, there is also no plausible mortality salience mechanism that could explain the kind of processing differences found here: “we wouldn't expect MS to lead to more accurate, attentive, or detailed processing’ (T. Pyszczynski, 2017, personal communication). Finally, the pattern of processing effects reported here are reliably found with a variety of thematically different mood induction materials that do not induce mortality salience (e.g., sunny vs. rainy days, pictures, stories, classical music) across the literature (Forgas, 2007, 2011a; Forgas et al., 2009; Forgas & Eich, 2013), so mortality salience cannot be a parsimonious explanation for these effects. Overall, then, negative mood recruiting more attentive processing offers the most plausible explanation of our findings, and the mediational analyses most clearly support this account (Bless & Fiedler, 2006). These experiments extend the growing literature demonstrating the intriguing adaptive benefits of mild negative mood for many cognitive and communication tasks. In social situations that require a more attentive, thorough and externally oriented processing style, negative mood can often improve performance on a variety of interpersonal tasks including eyewitness memory, language processing, self-disclosure and judgmental accuracy (e.g., Beukeboom, 2003; Bless, Mackie, & Schwarz, 1992; Bohner, Crow, Erb, & Schwarz, 1992; Forgas, 2007, 2011a, 2011b; Ruder & Bless, 2003).

their more internally oriented assimilative processing style. Those in a negative mood remembered other-relevant details better, consistent with externally oriented, accommodative processing (Bless & Fiedler, 2006). This is the first time such a dichotomy in recall has been demonstrated between internally oriented/self-related information and externally oriented/other related information. The assimilation-accommodation model offers the most general and parsimonious explanation for such mood effects. Although recall effects are also broadly consistent with other affect-cognition models such as the motivated mood-repair strategy (Clark & Isen, 1982), the globallocal process model (Forster & Dannenberg, 2010), and Fredrickson's (2001) broaden-and-build theory, these models do not predict more internally oriented information processing and enhanced self-related recall in a positive mood and externally focused thinking and enhanced other-related recall in a negative mood. Indeed, the more global focus in a positive mood suggested by Forster and Dannenberg (2010) and Fredrickson (2001) should imply better recall of all information (both self-related and other-related) in a positive mood. Motivated mood repair explanations (Clark & Isen, 1982) may imply an opposite prediction, that negative mood participants should focus on and remember more self-related details in ingratiating messages as an effective mood repair strategy. The assimilation-accommodation model (Bless & Fiedler, 2006) offers the most general and parsimonious account of the pattern reported here. It appears that paying a little less attention to the external environment due to a more internal and assimilative processing style in a positive mood (Bless & Fiedler, 2006) only reduces, but does not eliminate the production or interpretation of ingratiation. Mood-induced shifts in assimilative vs. accommodative processing occur along a subtle continuum (Bless & Fiedler, 2006); in the case of ingratiation, the processing benefits of negative mood were most marked, while there were no significant differences between positive and neutral mood groups in ingratiation effectiveness. Could it be that producing more positive ingratiating messages in Experiment 1, and accepting more positive ingratiatory messages in Experiment 3 when in a negative mood may be due to some kind of contrast effect (Kenrick, Montello, Gutierres, & Trost, 1993; Sedikides, 1995)? There are several reasons why the contrast account seems unlikely here. First, the affect-cognition literature overwhelming supports the existence of mood congruence rather than mood contrast effects in similar situations (Bower, 1981; Forgas & Eich, 2013). In addition, contrast effect explanations (Kenrick et al., 1993) have difficulty explaining the pattern of longer processing latencies and more externally oriented recall in a negative mood found here and elsewhere (Forgas & Eich, 2013). Interestingly, negative mood in Experiment 3 failed to produce greater suspicion and skepticism about messages, as found in some studies previously (Forgas & East, 2008). However, in prior studies negative mood increased skepticism about clearly dubious and uncertain information. As the ingratiating messages used here were plausible, realistic and credible, skepticism was less likely, allowing negative mood participants to focus more on the content of the messages. Indeed, the plausibility of the message used here was empirically confirmed in a validation study. Thus, negative mood can improve participants' attention to credible and realistic ingratiating messages and increase their impact, but may increase skepticism if the message is suspicious or not credible (Forgas & East, 2008). A further possibility is that participants in a negative mood were simply more compliant and were better at following instructions (a kind of accentuated demand effect). However, this account is inconsistent with the overwhelming weight of evidence showing that mood effects on social cognition and communication reliably occur even in the absence of specific instructions to behave in a particular way (Forgas, 2011a; Forgas et al., 2009; Matovic et al., 2014). Further, such a demand explanation also cannot account for the overall pattern of moodinduced processing differences and process mediation demonstrated in

6.2. Practical implications These experiments also have important and somewhat counterintuitive practical implications for our understanding of strategic interpersonal behavior, and for professional domains where impression management is of importance. Professionals working in advertising, political, clinical, legal and organizational settings may well benefit from training programs highlighting the subtle and often subconscious links between affective states and production of, and reactions to ingratiation. These results are also potentially relevant to the effective management of personal relationships, and online dating situations in particular (Cacioppo, Cacioppo, Gonzaga, Ogburn, & VanderWeele, 2013), suggesting that ingratiating strategies such as other-enhancement, conformity and self-enhancement (E. E. Jones, 1964) re likely to be more effectively produced, and more favorably received in a mild negative mood. 6.3. Limitations and future prospects These mood-induced processing effects are subject to various boundary conditions (Forgas & Eich, 2013; Gable & Harmon-Jones, 2008, 2010). For example, external demands, or strong arousal and motivation can reduce and override mood effects (Forgas, 1995; Forgas & Eich, 2013), suggesting that the more personally relevant and demanding the strategic communication task, the more likely that motivated processing may interfere with automatic and subconscious mood effects on information processing and ingratiation (Forgas, 1995; Forgas & Fiedler, 1996). As the motivation to ingratiate is often an implicit component of social interaction (Leary & Kowalski, 1990), it would be important to explore in future studies how variations in intrinsic and extrinsic motivation may influence mood effects on ingratiation. We used a common and ecologically valid real-life context for studying ingratiation here, online communication. However, it would be of considerable interest to replicate and extend these findings in 195

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

further realistic situations such as various face-to-face encounters. Future work might also look at exploring the boundary conditions of these effects, by manipulating the complexity and other characteristics of the ingratiating situation, such as the status, power and gender balance of the communicators. As the results across these three experiments appear to be conceptually consistent and reliable, producing theoretically coherent findings across different raters and ingratiation tasks, it is reasonable to expect that similar results may also be obtained in other settings. It would also be of interest to evaluate the influence of more intense and concrete emotional states such as anger, joy, pride, disgust, fear, depression and anxiety on the production and reception of ingratiation (Forgas, 2013; Matovic et al., 2014). The present finding that mild negative mood enhances both the construction and the reception of ingratiation suggests that other emotions may also produce similarly counterintuitive results. As emotions are perhaps the primary dimension of interpersonal behavior (Zajonc, 1981), investigating the influence of various emotions on strategic interpersonal behavior remains an important task. In conclusion, these experiments confirmed for the first time that negative mood tends to produce more extensive and more effective ingratiating strategies such as other-enhancement, conformity and selfenhancement. Further, recipients in a negative mood were also found to be more influenced by credible ingratiating messages. Correlations were consistent with our theoretically established mediational model predicting that processing differences mediate mood effects on ingratiation (Bless & Fiedler, 2006). These three experiments extend affect-cognition research to the new domain of ingratiation, highlighting the importance of further research exploring the subtle ways that mood states may influence strategic interpersonal behaviors.

affective consequences of everyday social interactions: The hidden benefits of putting one's best face forward. Journal of Personality and Social Psychology, 92, 990–1005. http://dx.doi.org/10.1037/0022-3514.92.6.990. Fiedler, K. (2001). Affective influences on social information processing. In J. P. Forgas (Ed.). Handbook of affect and social cognition (pp. 163–185). Mahwah, New Jersey: Lawrence Erlbaum Associates. Fiedler, K., Schott, M., & Meiser, T. (2011). What mediation analysis can (not) do. Journal of Experimental Social Psychology, 47, 1231–1236. http://dx.doi.org/10.1016/j.jesp. 2011.05.007. Forgas, J. P. (1995). Mood and judgment: The affect infusion model (AIM). Psychological Bulletin, 117, 39–66. http://dx.doi.org/10.1037/0033-2909/95/. Forgas, J. P. (1999). On feeling good and being rude: Affective influences on language use and request formulations. Journal of Personality and Social Psychology, 76, 928–939. http://dx.doi.org/10.1177/0146167299025007007. Forgas, J. P. (2007). When sad is better than happy: Negative affect can improve the quality and effectiveness of persuasive messages and social influence strategies. Journal of Experimental Social Psychology, 43, 513–528. http://dx.doi.org/10.1016/j. jesp.2006.05.006. Forgas, J. P. (2011a). Affective influences on self-disclosure: Mood effects on the intimacy and reciprocity of disclosing personal information. Journal of Personality and Social Psychology, 100, 449–461. http://dx.doi.org/10.1037/a0021129. Forgas, J. P. (2011b). Can negative affect eliminate the power of first impressions? Affective influences on primacy and recency effects in impression formation. Journal of Experimental Social Psychology, 47, 425–429. http://dx.doi.org/10.1016/j.jesp. 2010.11.005. Forgas, J. P. (2011c). She just doesn't look like a philosopher…? Affective influences on the halo effect in impression formation. European Journal of Social Psychology, 41, 812–817. http://dx.doi.org/10.1002/ejsp.842. Forgas, J. P. (2013). Don't worry, be sad! On the cognitive, motivational, and interpersonal benefits of negative mood. Current Directions in Psychological Science, 22, 225–232. http://dx.doi.org/10.1177/0963721412474458. Forgas, J. P., & East, R. (2008). On being happy and gullible: Mood effects on skepticism and the detection of deception. Journal of Experimental Social Psychology, 44, 1362–1367. http://dx.doi.org/10.1016/j.jesp.2008.04.010. Forgas, J. P., & Eich, E. (2013). Affective influences on cognition: Mood congruence, mood dependence, and mood effects on processing strategies. In A. F. Healy, & R. W. Proctor (Vol. Eds.), Handbook of psychology: Experimental psychology. Vol. 4. Handbook of psychology: Experimental psychology (pp. 61–82). NJ, US: Wiley. Forgas, J. P., & Fiedler, K. (1996). Us and them: Mood effects on intergroup discrimination. Journal of Personality and Social Psychology, 70, 28–40. http://dx.doi. org/10.1037/0022-3514.70.1.28. Forgas, J. P., Goldenberg, L., & Unkelbach, C. (2009). Can bad weather improve your memory? An unobtrusive field study of natural mood effects on real-life memory. Journal of Experimental Social Psychology, 45, 254–257. http://dx.doi.org/10.1016/j. jesp.2008.08.014. Forgas, J. P., Laham, S. M., & Vargas, P. T. (2005). Mood effects on eyewitness memory: Affective influences on susceptibility to misinformation. Journal of Experimental Social Psychology, 41, 574–588. http://dx.doi.org/10.1016/j.jesp.2004.11.005. Forster, J., & Dannenberg, L. (2010). GLOMOsys: A systems account of global versus local processing. Psychological Inquiry, 21, 175–197. http://dx.doi.org/10.1080/ 1047840X.2010.487849. Fredrickson, B. L. (2001). The role of positive emotions in positive psychology: The broaden-and-build theory of positive emotions. American Psychologist, 56, 218–226. http://dx.doi.org/10.1037/0003-066X.56.3.218. Gable, P. A., & Harmon-Jones, E. (2008). Approach-motivated positive affect reduces breadth of attention. Psychological Science, 19, 476–482. http://dx.doi.org/10.1111/j. 1467-9280.2008.02112.x. Gable, P. A., & Harmon-Jones, E. (2010). The blues broaden, but the nasty narrows: Attentional consequences of negative affects low and high in motivational intensity. Psychological Science, 21, 211–215. http://dx.doi.org/10.1177/0956797609359622. Gosling, S. D., Rentfrow, P. J., & Swann, W. B. (2003). A very brief measure of the big-five personality domains. Journal of Research in Personality, 37, 504–528. http://dx.doi. org/10.1016/s0092-6566(03)00046-1. Greenberg, J., Martens, A., Jonas, E., Eisenstadt, D., Pyszczynski, T., & Solomon, S. (2003). Psychological defense in anticipation of anxiety: Eliminating the potential for anxiety eliminates the effect of mortality salience on worldview defense. Psychological Science, 14, 516–519. http://dx.doi.org/10.1111/1467-9280.03454. Isen, A. M., Daubman, K. A., & Nowicki, G. P. (1987). Positive affect facilitates creative problem solving. Journal of Personality and Social Psychology, 52, 1122–1131. http:// dx.doi.org/10.1037/0022-3514.52.6.1122. Jones, E. E. (1964). Ingratiation. New York: Appleton-Century-Crofts. Jones, E. E., Gergen, K. J., & Jones, R. G. (1963). Tactics of ingratiation among leaders and subordinates in a status hierarchy. Psychological Monographs: General and Applied, 77, 1–20. http://dx.doi.org/10.1037/h0093832. Jones, R. G., & Jones, E. E. (1964). Optimum conformity as an ingratiation tactic. Journal of Personality, 32, 436–458. http://dx.doi.org/10.1111/j.1467-6494.1964.tb01351.x. Kenrick, D. T., Montello, D. R., Gutierres, S. E., & Trost, M. R. (1993). Effects of physical attractiveness on affect and perceptual judgments: When social comparison overrides social reinforcement. Personality and Social Psychology Bulletin, 19, 195–199. http:// dx.doi.org/10.1177/0146167293192008. Koch, A. S., Forgas, J. P., & Matovic, D. (2013). Can negative mood improve your conversation? Affective influences on conforming to Grice's communication norms. European Journal of Social Psychology, 43, 326–334. http://dx.doi.org/10.1002/ejsp. 1950. Leary, M. R., & Kowalski, R. M. (1990). Impression management: A literature review and two-component model. Psychological Bulletin, 107, 34–47. http://dx.doi.org/10.

References Baumeister, R. F. (1982). A self-presentational view of social phenomena. Psychological Bulletin, 91, 3–26. http://dx.doi.org/10.1037/0033-2909.91.1.3. Beukeboom, C. J. (2003). How mood turns on language (Doctoral dissertation)Free University of Amsterdam. Bless, H., Clore, G. L., Schwarz, N., Golisano, V., Rabe, C., & Wolk, M. (1996). Mood and the use of scripts: Does a happy mood really lead to mindlessness? Journal of Personality and Social Psychology, 71, 665–679. http://dx.doi.org/10.1037/00223514.71.4.665. Bless, H., & Fiedler, K. (1995). Affective states and the influence of activated general knowledge. Personality and Social Psychology Bulletin, 21, 766–778. http://dx.doi.org/ 10.1177/0146167295217010. Bless, H., & Fiedler, K. (2006). Mood and the regulation of information processing and behavior. In J. P. Forgas (Ed.). Affect in social thinking and behavior (pp. 65–84). New York: Psychology Press. Bless, H., Mackie, D. M., & Schwarz, N. (1992). Mood effects on attitude judgments: Independent effects of mood before and after message elaboration. Journal of Personality and Social Psychology, 63, 585–595. http://dx.doi.org/10.1037/00223514.63.4.585. Bodenhausen, G. V., Kramer, G. P., & Susser, K. (1994). Happiness and stereotypic thinking in social judgment. Journal of Personality and Social Psychology, 66, 621–632. http://dx.doi.org/10.1037/0022-3514.66.4.621. Bohner, G., Crow, K., Erb, H.-P., & Schwarz, N. (1992). Affect and persuasion: Mood effects on the processing of message content and context cues and on subsequent behavior. European Journal of Social Psychology, 22, 511–530. http://dx.doi.org/10. 1002/ejsp.2420220602. Bower, G. H. (1981). Mood and memory. American Psychologist, 36, 129–148. http://dx. doi.org/10.1037/0003-066X.36.2.129. Cacioppo, J. T., Cacioppo, S., Gonzaga, G. C., Ogburn, E. L., & VanderWeele, T. J. (2013). Marital satisfaction and break-ups differ across on-line and off-line meeting venues. Proceedings of the National Academy of Sciences, 110, 10135–10140. http://dx.doi. org/10.1073/pnas.1222447110. Clark, M. S., & Isen, A. M. (1982). Toward understanding the relationship between feeling states and social behavior. In A. H. Hastorf, & A. M. Isen (Eds.). Cognitive social psychology (pp. 73–108). New York: Elsevier/North-Holland. Collins, N. L., & Miller, L. C. (1994). Self-disclosure and liking: A meta-analytic review. Psychological Bulletin, 116, 457–475. http://dx.doi.org/10.1037/0033-2909.116.3. 457. Cramer, E. M., & Bock, R. D. (1966). Multivariate analysis. Review of Educational Research, 36, 604–617. http://dx.doi.org/10.3102/00346543036005604. Dunbar, R. (1996). Grooming, gossip, and the evolution of language. Cambridge, MA: Harvard University Press. Dunn, E. W., Biesanz, J. C., Human, L. J., & Finn, S. (2007). Misunderstanding the

196

Journal of Experimental Social Psychology 76 (2018) 186–197

D. Matovic, J.P. Forgas

affective states. In E. T. Higgins, & R. M. Sorrentino (Vol. Eds.), Handbook of motivation and cognition: Foundations of social behavior. Vol. 2. Handbook of motivation and cognition: Foundations of social behavior (pp. 527–561). New York: Guilford. Sedikides, C. (1995). Central and peripheral self-conceptions are differentially influenced by mood: Tests of the differential sensitivity hypothesis. Journal of Personality and Social Psychology, 69, 759–777. http://dx.doi.org/10.1037/0022-3514.69.4.759. Seiter, J. S. (2007). Ingratiation and gratuity: The effect of complimenting customers on tipping behavior in restaurants. Journal of Applied Social Psychology, 37, 478–485. http://dx.doi.org/10.1111/j.1559-1816.2007.00169.x. Storbeck, J., & Clore, G. L. (2005). With sadness conies accuracy; with happiness, false memory: Mood and the false memory effect. Psychological Science, 16, 785–791. http://dx.doi.org/10.1111/j.1467-9280.2005.01615.x. Tice, D. M., Butler, J. L., Muraven, M. B., & Stillwell, A. M. (1995). When modesty prevails: Differential favorability of self-presentation to friends and strangers. Journal of Personality and Social Psychology, 69, 1120–1138. http://dx.doi.org/10.1037/00223514.69.6.1120. Vonk, R. (2002). Self-serving interpretations of flattery: Why ingratiation works. Journal of Personality and Social Psychology, 82, 515–526. http://dx.doi.org/10.1037//00223514.82.4.515. Westermann, R., Spies, K., Stahl, G. K., & Hesse, F. W. (1996). Relative effectiveness and validity of mood induction procedures: A meta-analysis. Journal of Social Psychology, 26, 557–580. http://dx.doi.org/10.1002/(SICI)1099-0992(199607)26:4<557::AIDEJSP769>3.0.CO;2-4. Zajonc, R. B. (1981). A one-factor mind about mind and emotion. American Psychologist, 36, 102–103. http://dx.doi.org/10.1037/0003-066X.36.1.102.

1037/0033-2909.107.1.34. Lindeman, M., & Verkasalo, M. (2005). Measuring values with the short Schwartz's value survey. Journal of Personality Assessment, 85, 170–178. http://dx.doi.org/10.1207/ s15327752jpa8502_09. Machiavelli, N. (1961). The prince. G. Bull, Trans.London: Penguin. Matovic, D., Koch, A. S., & Forgas, J. P. (2014). Can negative mood improve language understanding? Affective influences on the ability to detect ambiguous communication. Journal of Experimental Social Psychology, 52, 44–49. http://dx.doi.org/10.1016/ j.jesp.2013.12.003. Preacher, K. J., & Hayes, A. F. (2004). SPSS and SAS procedures for estimating indirect effects in simple mediation models. Behavior Research Methods, Instruments, & Computers, 36, 717–731. http://dx.doi.org/10.3758/BF03206553. Proost, K., Schreurs, B., De Witte, K., & Derous, E. (2010). Ingratiation and self-promotion in the selection interview: The effects of using single tactics or a combination of tactics on interviewer judgments. Journal of Applied Social Psychology, 40, 2155–2169. http://dx.doi.org/10.1111/j.1559-1816.2010.00654.x. Ruder, M., & Bless, H. (2003). Mood and the reliance on the ease of retrieval heuristic. Journal of Personality and Social Psychology, 85, 20–32. http://dx.doi.org/10.1037/ 0022-3514.85.1.20. Schlenker, B. R., & Leary, M. R. (1982). Audiences' reactions to self-enhancing, self-denigrating, and accurate self-presentations. Journal of Experimental Social Psychology, 18, 89–104. http://dx.doi.org/10.1016/0022-1031%2882%2990083-X. Schutz, A. (1997). Self-presentational tactics of talk-show guests: A comparison of politicians, experts, and entertainers. Journal of Applied Social Psychology, 27, 1941–1952. http://dx.doi.org/10.1111/j.1559-1816.1997.tb01633.x. Schwarz, N. (1990). Feelings as information: Informational and motivational functions of

197