Positive affect increases secondary control among causally uncertain individuals

This article was downloaded by: [UQ Library]On: 07 May 2014, At: 04:08Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office:Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Click for updates

Cognition & EmotionPublication details, including instructions for authors and subscriptioninformation:http://www.tandfonline.com/loi/pcem20

Positive affect increases secondary controlamong causally uncertain individualsStephanie J. Tobina & Melanie P. Georgea

a School of Psychology, University of Queensland, Brisbane, QLD 4072,AustraliaPublished online: 06 May 2014.

To cite this article: Stephanie J. Tobin & Melanie P. George (2014): Positive affect increases secondarycontrol among causally uncertain individuals, Cognition & Emotion, DOI: 10.1080/02699931.2014.916253

To link to this article: http://dx.doi.org/10.1080/02699931.2014.916253

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”)contained in the publications on our platform. However, Taylor & Francis, our agents, and ourlicensors make no representations or warranties whatsoever as to the accuracy, completeness, orsuitability for any purpose of the Content. Any opinions and views expressed in this publicationare the opinions and views of the authors, and are not the views of or endorsed by Taylor &Francis. The accuracy of the Content should not be relied upon and should be independentlyverified with primary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilitieswhatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantialor systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, ordistribution in any form to anyone is expressly forbidden. Terms & Conditions of access and usecan be found at http://www.tandfonline.com/page/terms-and-conditions

Positive affect increases secondary control among causallyuncertain individuals

Stephanie J. Tobin and Melanie P. George

School of Psychology, University of Queensland, Brisbane, QLD 4072, Australia

Secondary control (acceptance of and adjustment to negative events) is thought to promote positiveaffect. We examined the opposite path: could positive affect increase secondary control, particularlyamong individuals high in causal uncertainty, who stand to benefit from it the most? In two studies,participants completed a causal uncertainty scale, thought about a problem while listening to affect-inducing music or no music, and then completed items that assessed secondary control. In Study 1,the music induced positive or negative affect. In Study 2, the music induced affect that was high orlow in activation and positive or negative in valence. In both studies, we found that positive affect-inducing music increased secondary control among high causal uncertainty participants. Further-more, trait affect did not account for the effects of causal uncertainty, and music did not influenceprimary control. These findings show that secondary control can fluctuate as a function of stateaffect.

Keywords: Secondary control; Causal uncertainty; Affect.

Sometimes people let the same problem makethem miserable for years when they could just say,So what. That's one of my favorite things to say.So what.

—Andy Warhol, The Philosophy of AndyWarhol

Secondary control

Saying “so what” to one’s problems indicates a degree

of acceptance of existing events and adjustment to

them. This combination of acceptance and adjust-

ment is known as secondary control (Morling &

Evered, 2006). It is called secondary control todistinguish it from the traditional notion of control(i.e., primary control), or the sense that one caninfluence outcomes to suit one’s desires. Past researchsuggests that in the face of unwanted outcomes,secondary control can promote psychological andphysical well-being (Morling & Evered, 2006;Rothbaum, Weisz, & Snyder, 1982). For instance,Rothermund and Brandtstädter (2003) found thatflexibly adjusting one’s goals in the face of constraintshelped to minimise negative affect in a sample ofolder adults. Furthermore, Chipperfield et al. (2012)found that secondary control helps to foster an overallsense of control, which predicts better self-reported

Correspondence should be addressed to: Stephanie J. Tobin, School of Psychology, University of Queensland, St. Lucia, QLD

4072, Australia. E-mail: [email protected]

© 2014 Taylor & Francis 1

COGNITION AND EMOTION, 2014

http://dx.doi.org/10.1080/02699931.2014.916253

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

health, lower hospital admissions, and lower mortal-

ity in a sample of older adults. Conceptually similar

constructs such as goal disengagement (Wrosch,

Miller, Scheier, & Brun de Pontet, 2007) and

“shifting” (accepting stress and adapting oneself to

it; Chen & Miller, 2012) have also been linked with

beneficial health outcomes.1

Researchers have posited that secondary control ismost adaptive when either the actual opportunity toalter events is low or when individuals perceive it tobe low (Brandtstädter & Rothermund, 2002;Heckhausen & Schulz, 1995; Morling & Evered,2006; Rothbaum et al., 1982; Thompson et al.,1998). Consistent with this idea, Thompson et al.(1998) found that secondary control had a protectiveeffect on psychological well-being when perceivedprimary control was low, but not when it was high.Similarly, Bailis, Chipperfield, and Perry (2005)found that positive social comparisons, a secondarycontrol strategy, predicted better health only whenindividuals were low in primary control.

Causal uncertainty

Tobin and Raymundo (2010) found that secondarycontrol is particularly beneficial for individuals whoare high in chronic causal uncertainty. Chroniccausal uncertainty refers to ongoing doubts aboutone’s ability to understand why positive and negativeevents happen to the self and others (Weary &Edwards, 1996; Weary, Tobin, & Edwards, 2010).Chronic causal uncertainty is associated concurrentlywith depression and anxiety and longitudinally witha lack of perceived primary control (Edwards &Weary, 1998; Jacobson, Weary, & Edwards, 1999;Weary & Edwards, 1994). Specifically, longitudinalstudies show that lack of control can bring aboutcausal uncertainty in the first place, and causaluncertainty can further undermine control (Edwards& Weary, 1998; Jacobson et al., 1999).

Discrepancies between current and desired levelsof causal understanding can motivate efforts toimprove one’s understanding and restore a sense ofcontrol and well-being. Consistent with this idea, anumber of studies have found that individuals withhigh levels of chronic causal uncertainty engage inmore thorough information processing strategies(Tobin & Weary, 2008; Weary & Jacobson, 1997;Weary, Jacobson, Edwards, & Tobin, 2001; Weary,Vaughn, Stewart, & Edwards, 2006). However,according to the causal uncertainty model, unsuc-cessful attempts to improve understanding cangenerate negative affect, and when the likelihood ofsuccess is too low, individuals will opt to disengagefrom the goal of improving their understanding(Weary et al., 2010). Particularly in the case ofchronic causal uncertainty, individuals may have ahistory of failed attempts to improve their under-standing which creates a desire to disengage. Con-sistent with the idea that chronic causal uncertaintycan lead to disengagement, there is evidence thatchronic causal uncertainty is associated with the useof disengaged coping strategies in the context of astressful event (Chang, 2000) and with the use ofalcohol to cope with negative thoughts and feelings(Tobin, Loxton, & Neighbors, 2014). Secondarycontrol could be another way of disengaging frompotentially futile attempts to improve their under-standing. If individuals are able to accept and adjustthemselves to their uncertainty and salient negativeevents, it should be easier for them to disengagefrom endless rumination about the causes of thoseevents, and thereby escape the negative affect thatcan stem from unsuccessful uncertainty reductionefforts. Consistent with this idea, Tobin andRaymundo (2010) found that although higher levelsof chronic causal uncertainty predicted increasingnegative affect over time when secondary control waslow, this was not the case when secondary controlwas high.

1 There is some similarity between secondary control and emotion regulation. Reappraisal, an emotion regulation strategythat involves changing one’s thoughts about a situation in order to change one’s emotion (Gross, 2013), is similar to theadjustment component of secondary control, whereby individuals adjust some aspect of the self to bring it in line withexisting events (Morling & Evered, 2006). However, the acceptance component is unique to secondary control.Furthermore, secondary control does not necessarily stem from the goal to regulate emotion, and some researchers positthat secondary control cannot be consciously regulated (Brandtstädter & Rothermund, 2002).

TOBIN AND GEORGE

2 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

Increasing secondary control

Because of its positive effects on psychological andphysical well-being, it seems desirable to developeffective strategies for increasing secondary con-trol. However, much of the existing research onsecondary control examines it as an individualdifference variable (Morling & Evered, 2006), oras a capacity that develops naturally with age(Brandtstädter & Renner, 1990; Heckhausen &Schulz, 1995). There have been only a couple ofattempts to increase secondary control. Theseattempts have involved changing depressogenicthoughts, finding a silver lining in negativeevents, and using relaxation and positive imagery(Weisz, Thurber, Sweeney, Proffitt, & LeGag-noux, 1997), and having students explain howthey were able to learn from or reinterpret a pooracademic outcome in a positive manner (Hall,Perry, Chipperfield, Clifton, & Haynes, 2006).These interventions were effective in reducingdepressive symptomatology (Weisz et al., 1997),increasing secondary control and improving aca-demic outcomes (Hall et al., 2006). However, theywere part of broader interventions that alsoincluded either a primary control element (Weiszet al., 1997) or an attributional retraining element(Hall et al., 2006). Thus, it is unclear that whichcomponent(s) produced the desirable outcomes.We took a more targeted approach in the currentresearch and focused on inducing positive affect.

Prior research has focused primarily on affect asan outcome of secondary control (Chipperfieldet al., 2012; Morling & Evered, 2006; Tobin &Raymundo, 2010). Of interest in the currentresearch is whether affect can influence secondarycontrol. An examination of the affect literaturelends credence to this idea. Affect is known toinfluence cognition in a variety of ways. Forinstance, positive affect can facilitate the retrievalof positive thoughts (Bower, 1981), increasethought flexibility and creativity (Isen & Labroo,2003) and signal that all is well in the environment(Schwarz & Clore, 2007). Folkman and Mosko-witz (2000) argue that positive emotions play animportant role in coping with stress by facilitatingpositive interpretations of events. Positive affect

can also lead to a broader style of thinking(Fredrickson, 2001) that allows people to acceptstressors and the accompanying negative emotions(Larsen, Hemenover, Norris, & Cacioppo, 2003)and find positive meaning in such events (Tugade &Fredrickson, 2002).

In sum, past research supports the idea thatpositive affect could facilitate secondary control.However, valence is not the only dimension ofaffective experience. There is evidence to suggestthat two dimensions underlie core affect: activation/deactivation and pleasure/displeasure (Russell, 2003).High-activation positive states include excited andelated, whereas low-activation positive states includeserene and relaxed. High-activation negative affectivestates include nervous and stressed, whereas low-activation negative states include depressed andlethargic.

Both affective dimensions can influence ourperceptions of ourselves and the world. Comparedto displeasure, pleasure leads people to focus onpositive aspects and interpretations of events(Forgas, 1995). Compared to deactivation, activa-tion signals mobilisation potential or readiness foraction (Russell, 2003). The work of Tsai and hercolleagues sheds some light on possible associa-tions between high and low activation forms ofpositive affect and secondary control. Theseresearchers examined whether cultural differencesin goals to influence others (a primary controlgoal) and adjust to others (a secondary controlgoal) could produce differences in ideal affect.They proposed that high-activation positive affectwould help people influence others, whereas low-activation positive affect would help people adjustto others (Tsai, Miao, Seppala, Fung, & Yeung,2007). In several studies, they found that whenindividuals needed to influence others, theydesired high-activation positive states, whereaswhen they needed to adjust themselves to others,they desired low-activation positive states. In aninterpersonal context, then, low activation positiveaffect may promote secondary control.

This effect may extend to other contexts, how-ever, because activation level may influence theperceived relevance of the affect to the judgement

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 3

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

at hand. Matches between one’s current affectivestate and some aspect of the situation or individualincrease the perceived relevance of the affectivestate, which increases its influence on judgments(Greifeneder, Bless, & Pham, 2011; Tobin &Tidwell, 2013). It seems plausible that low activationpositive affect would be informative for one’sacceptance of and adjustment to a negative event,because secondary control is thought to produce astate of serenity rather than excitement (Morling &Evered, 2006). Thus, people may be more likely toinfer that low relative to high activation positiveaffect indicates their acceptance of and adjustment toa negative event.

Would negative affect have the opposite effect;that is, would it undermine secondary control?There is some evidence that trait negative affectcan make it more difficult for individuals topositively reappraise negative events (Ng &Diener, 2009). However, there is also evidencethat individuals are more likely to search for thesource of negative relative to positive affect, and ifthey think their negative affect stems from some-thing incidental such as the weather, they do notuse it in their judgments (Schwarz & Clore,1983). Thus, if affect as information is part ofthe underlying process, negative affect may exertless influence on judgments related to secondarycontrol.

The current research

Based on existing research and theory, we pre-dicted that positive affect would increase second-ary control, and that low activation positive affectwould be particularly effective. Furthermore, thiseffect should be stronger for individuals with highlevels of chronic causal uncertainty, as secondarycontrol would help them disengage from efforts toimprove their understanding which could other-wise lead to negative affect (Tobin & Raymundo,2010). We tested these hypotheses in two studies.In both studies, we used music to induce differentaffective states while participants thought about aproblem that had been bothering them. We thenassessed secondary control with respect to their

problem: the extent to which they were able toaccept their problem and adjust themselves to it.In Study 1, we focused on the affective states thatshould be the most (low activation positive) andleast (high activation negative) conducive to sec-ondary control. In Study 2, we examined allpossible combinations of affect valence and activa-tion to understand the roles of the two dimen-sions. In both studies, we had a no music controlcondition for comparison purposes. We predictedthat among participants who were high in causaluncertainty, secondary control would be highestwhen they were listening to low activation positiveaffect-inducing music.

STUDY 1

For all studies reported in this manuscript, wedetermined our sample size by aiming for 15–20participants within a given level of causal uncer-tainty and condition. Our participants were under-graduate psychology students who received coursecredit for their participation. All studies wereconducted in a computer laboratory. We report alldata exclusions, manipulations and measures.

Pilot study

We first pilot tested the music to ensure that itelicited the desired affective state. Sixty-one parti-cipants completed the causal uncertainty scale(Weary & Edwards, 1994), followed by a10-item attitude survey that served as a filler task(Tobin & Weary, 2008). The causal uncertaintyscale consists of 14 items (α = .86) expressingdoubt about the causes of positive and negativeevents involving the self and others (e.g., “Whenbad things happen, I generally do not know why”).Participants rated the extent to which they agreedwith each item on a 6-point scale.

Participants then were randomly assigned tolisten to calming, irritating or no music for oneminute. The calming music was a pleasant pianoconcerto (Mozart’s “Piano Concerto no. 21,Andante;” Kreutz, Ott, Teichmann, Osawa, &

TOBIN AND GEORGE

4 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

Vaitl, 2008). The irritating music was a mix ofhigh-pitched buzzing, screeching and digitaltones (Kid 606’s “Sometimes I Thank GodI Can’t Sing Because then Nobody Can Blameme for Anything”). Participants then completeda state affect scale (Tsai, Knutson, & Fung,2006; Watson & Clark, 1994; Watson, Clark, &Tellegen, 1988) while the music continued toplay for those in the calming and irritatingmusic conditions. The affect scale assessedhigh-activation positive affect (e.g., excited, α =.83), high-activation negative affect (e.g., irrit-able, α = .88), sadness (e.g., sad, α = .82) andserenity (e.g., relaxed, α = .83).

A series of hierarchical regression analyses wasconducted to examine the effects of causal uncer-tainty and music condition on state affect. Causaluncertainty scores were centred around the meanand music condition was dummy-coded into twovariables: calm music (1 = calm music, 0 = nomusic) and irritating music (1 = irritating music,0 = no music). To assess the significance of theoverall three-level music variable and its interac-tion with causal uncertainty, we entered the pair ofvariables as a block in a hierarchical regression andexamined change in R2 (Aiken & West, 1991).Specifically, we entered causal uncertainty in block1, the two music variables in block 2 and the twoMusic × Causal Uncertainty interaction terms inblock 3. The change in R2 in block 2 indicates theoverall effect of music condition and the change inR2 in block 3 indicates the overall Music × CausalUncertainty interaction. This approach alsoallowed us to observe any main effects of causaluncertainty in block 1. Because of the dummycoding, the causal uncertainty term in block 3represents the simple effect of causal uncertainty inthe no music condition, rather than the effect ofcausal uncertainty across conditions.

Analyses revealed that higher levels of causaluncertainty were associated with higher levels ofhigh activation negative affect, β = .31, t(59) =2.51, p = .015, and sadness, β = .34, t(59) = 2.82,p = .007. Analyses also revealed a main effect ofmusic condition on serenity, ΔR2 = .42, ΔF(2,57) =21.10, p < .001, and high activation negative affect,ΔR2 = .11, ΔF(2,57) = 3.84, p = .027. An

examination of the individual dummy-coded termsrevealed that relative to the no music condition,serenity was higher in the calm music condition,β = .29, t(55) = 2.42, p = .019, and lower in theirritating music condition, β = −.50, t(55) = −3.94,p < .001. Additionally, compared to the no musiccondition, high-activation negative affect, β = .33,t(55) = 2.20, p = .032, was higher in the irritatingmusic condition. None of the interaction terms weresignificant, ps > .23. Thus, the music had theintended effect on state affect and this effect didnot depend upon level of causal uncertainty.

Method

Participants

One hundred two students participated in themain study. The data from five participants wereexcluded: one participant was an outlier (>3 SD)on one or more measure, one participant entered“nothing” for the problem, and three participantsskipped through one or more screen in less than300 ms. After these exclusions, the sample con-sisted of 75 female and 22 male participants,whose average age was 19.36 (SD = 4.96).

Individual difference measures

Participants began by completing the causal un-certainty scale (α = .78; Weary & Edwards, 1994),along with the trait version of the positive andnegative affect schedule (Watson et al., 1988),which assesses trait levels of high activationpositive and negative affect. Participants rated theextent to which they generally felt 10 positive(α = .83) and 10 negative (α = .84) feelings on a5-point scale. The order of the causal uncertaintyand trait affect scale was randomised, and the lastscale was followed by the attitude filler survey.

Identification of problem

Participants were told that the next part of thestudy involved thinking about an issue and listen-ing to some sounds. They were asked to put onheadphones, think of something that had beenbothering them lately and summarise the problemin a sentence. An examination of the problems

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 5

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

entered indicated that most of the problems wereinterpersonal (e.g., “My girlfriend and I recentlybroke up”; 39%) or academic (e.g., “Inundation ofUniversity work”; 31%), with the remainderincluding various issues such as something aboutthe self, the future, money or political issues.

Music

Depending upon condition, participants were nextasked to spend two minutes thinking about theproblem they identified while listening to calming,irritating or no music. The problem they hadentered earlier appeared at the top of the computerscreen during this time. After two minutes hadelapsed, the music continued to play while parti-cipants completed the dependent measures.

Secondary control

Participants answered eight questions that tappedtheir level of secondary control with respect totheir problem (α = .80; see Appendix). The itemstapped into both components of secondary con-trol: accepting the problem rather than trying tochange it and adjusting themselves to it.

Manipulation checks

A subset of the items from the pilot study wasincluded to assess participants’ current levels ofserenity (5 items, α = .89) and irritation (3 items,α = .71).

Results

Manipulation checks

As in the pilot study, causal uncertainty scoreswere centred around the mean and music condi-tion was dummy-coded and entered in separateblocks to test the overall main and interactioneffect of the three-level variable (Aiken & West,1991). We observed only significant main effectsof music condition on serenity, ΔR2 = .29, ΔF(2,93) = 19.17, p < .001, and irritation, ΔR2 = .25,ΔF(2,93) = 15.59, p < .001. An examination ofthe individual dummy-coded terms revealed thatrelative to the no music condition, serenity washigher in the calm music condition, β = .35,

t(91) = 3.44, p = .001, and lower in the irritatingmusic condition, β = −.27, t(91) = −2.68, p = .009,and irritation was higher in the irritating musiccondition, β = .42, t(91) = 4.03, p < .001.

Secondary control

Using the same approach, we examined the effectsof causal uncertainty and music condition onsecondary control and found only a significantCausal Uncertainty × Music Condition interac-tion, ΔR2 = .10, ΔF(2,91) = 5.36, p = .006. Anexamination of the individual coefficients revealedthat only the Causal Uncertainty × Calm Musicinteraction was significant, β = .33, t(91) = 2.41,p = .018 (see Figure 1). Simple slopes testsindicated that serenity-inducing music increasedsecondary control relative to no music when causaluncertainty was high (1 SD above the mean),β = .37, t(91) = 2.05, p = .044. When causaluncertainty was low (1 SD below the mean),serenity-inducing music did not increase second-ary control, relative to no music, β = −.31, t(91) =−1.68, p = .096.

Controlling for trait affect

To examine whether our effects were unique tocausal uncertainty or driven by the broader con-struct of trait negative affect, we centred traitnegative affect and included it as a predictor ofsecondary control, along with causal uncertainty,music condition and all interactions. Because ofthe dummy coding of condition, we again used a

123456789

Low (-1 SD) High (+1 SD)

Seco

ndar

y C

ontr

ol

Causal Uncertainty

No MusicCalm MusicIrritating Music

Figure 1. Secondary control as a function of causal uncertainty

and music condition in Study 1.

TOBIN AND GEORGE

6 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

hierarchical approach so that we could look atoverall main effects of our continuous variables inthe first block and at any effects involving thethree-level categorical predictor with the change inR2. This analysis revealed only a significant CausalUncertainty × Music Condition interaction, ΔR2 =.08, ΔF(2,87) = 4.29, p = .017. Using the sameapproach, we examined trait positive affect as apredictor of secondary control. We simplyreplaced trait negative with positive affect in theregression. This analysis revealed a main effect oftrait positive affect, β = .30, t(93) = 2.81, p = .006,such that higher levels of positive affect wereassociated with higher secondary control. How-ever, the Causal Uncertainty × Music Conditioninteraction remained significant, ΔR2 = .10, ΔF(2,87) = 5.22, p = .007.

Discussion

As predicted, we found that serenity-inducingmusic increased problem-related secondary controlwhen participants were high in causal uncertainty.This supports the idea that individuals high incausal uncertainty take advantage of opportunitiesto disengage, and that positive affect can increasesecondary control. We also found a positiveassociation between trait positive affect and sec-ondary control. Interestingly, neither the irritatingmusic nor trait negative affect were associated withsecondary control. Participants may have beenmore likely to identify the actual source of theirnegative affect and refrain from using it asinformation when assessing their level of accept-ance and adjustment (Schwarz & Clore, 1983).

STUDY 2

The primary purpose of Study 2 was to examinewhether the effects observed in Study 1 were due topositive valence, low activation or both. Accordingly,we fully crossed affect valence and activation level.We predicted that low activation positive affect-inducing music would increase secondary control themost among high causal uncertainty individuals,because it should be viewed as most relevant to

judgments about acceptance of and adjustment tonegative events (Greifeneder et al., 2011; Tobin &Tidwell, 2013).

We also included a measure of primary controlwith respect to participants’ problems: the extentto which participants thought they could solvetheir problem. We predicted that any effects ofmusic on primary control would be different tothose observed on secondary control. Whereas lowactivation positive affect should be optimal forsecondary control, we thought that high activationpositive affect might be optimal for primarycontrol, as the high activation component signalsmobilisation potential (Russell, 2003) and may beseen as most relevant to judgments about attaininggoals. However, it is unclear whether high activa-tion positive affect would increase primary controlfor high causal uncertainty individuals, as theytend to have lower levels of perceived primarycontrol (Edwards & Weary, 1998; Jacobson et al.,1999; Weary & Edwards, 1994).

Lastly, in Study 2, we asked participants to ratetheir problems on several dimensions prior to themusic manipulation to examine whether causaluncertainty influenced the type of problems parti-cipants identified.

Pilot study

A pilot study was conducted to test whether theselected songs elicited the desired affective states.The study used a within-subjects design, in which5 male and 13 female participants listened andresponded to four songs, in randomised order.The high activation negative and low activationpositive songs were those used in Study 1. Thehigh activation positive song was an upbeat technopiece (“All I Need” by Paul Van Dyk). The lowactivation negative song was a sad violin piece(“Schindler’s List – Theme” by Ian Sutherland,New World Philharmonic Orchestra and TasminLittle). Participants listened to a song for twominutes and then rated the valence and activationof their affect on 9-point scales (Lang, Bradley, &Cuthbert, 2008) while the song continued to play.They did this for each of the four songs.

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 7

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

A 2 (music valence) × 2 (music activation)

within-subjects ANOVA on activation ratings

revealed only a significant main effect of music

activation, F(1,17) = 62.16, p < .001. Participants

felt more activated when listening to the high

activation music (M = 6.22, SD = 1.34) compared

to the low activation music (M = 3.00, SD = 0.87).

A 2 (music valence) × 2 (music activation)

within-subjects ANOVA on the valence ratings

revealed significant main effects of music valence,

F(1,17) = 140.53, p < .001, and activation,

F(1,17) = 10.03, p = .006, and a Valence ×

Activation interaction, F(1,17) = 8.31, p = .010.

Pairwise comparisons revealed that participants

felt more positive when listening to the high

activation positive music (M = 6.78, SD = 1.35)

compared to the high activation negative music

(M = 2.61, SD = 1.54), F(1,17) = 71.31, p < .001,

and when listening to the low activation positive

music (M = 6.61, SD = 1.58) compared to the low

activation negative music (M = 4.83, SD = 1.82),

F(1,17) = 13.99, p = .002. Thus, the effect of

music valence on state affect valence was signific-

ant and in the predicted direction in both the high

and low activation conditions, albeit stronger in

the high activation condition.

Method

Participants

One hundred sixty-five students participated in

the main study. The data from 21 participants

were excluded: 11 participants were outliers (>3

SD) on one or more measure, 9 participants

skipped through one or more screen in less than

300 ms and 2 participants came close to guessing

our hypothesis. After these exclusions, the sample

consisted of 92 female and 52 male participants,

whose average age was 19.57 (SD = 3.49).

Individual difference measure

Participants began by completing the causal un-

certainty scale (α = .87; Weary & Edwards, 1994),

followed by the attitude filler survey.

Identification of problem

Participants were asked to put on headphones,think of something that had been bothering themlately, and summarise the problem in a sentence.They then rated on 9-point scales how much theproblem had been bothering them, the extent towhich they were able to influence the events inquestion, how well they understood the cause ofthe problem, and how important it was tounderstand the cause of the problem. As in Study1, an examination of the problems entered indi-cated that the most common types of problemswere academic (40%) and interpersonal (31%).

Music

Depending upon the condition to which they hadbeen randomly assigned, participants were nextasked to spend two minutes thinking about theirproblem while listening to one of the four musicselections or no music. The problem they hadentered earlier appeared at the top of the computerscreen during this time. After two minutes hadelapsed, the music continued to play while parti-cipants completed the dependent measures. Themusic conditions formed a 2 (valence: positive,negative) × 2 (activation: high, low) between-subjects factorial design.

Secondary control

As in Study 1, participants answered eight ques-tions that tapped their level of secondary controlwith respect to their problem (α = .79).

Primary control

Five additional items assessed the extent to whichparticipants thought they could solve their prob-lem (α = .88; see Appendix). The order of theprimary and secondary control sets of items wasrandomised.

Manipulation checks

A subset of the items from the first pilot studywas included to assess participants’ current levelsof serenity (5 items, α = .87), sadness (5 items,

TOBIN AND GEORGE

8 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

α = .86), excitement (3 items, α = .80) andirritation (3 items, α = .49).2

Results

Correlations between causal uncertainty and per-ceptions of problems before the music manipula-tion were examined. Higher causal uncertainty wasassociated with less understanding of the cause ofthe problem, r = −.28, p = .001, as we wouldexpect. However, causal uncertainty was not sig-nificantly associated with the extent to whichparticipants were bothered by their problem, r =.01, p = .877, thought they could influence theirproblem, r = −.09, p = .281, or thought it wasimportant to understand the cause of their prob-lem, r = −.03, p = .705. Thus, high and low causaluncertainty participants identified similar types ofproblems, but higher causal uncertainty was asso-ciated with less understanding of the causes.

Manipulation checks

As in Study 1, we centred causal uncertainty scoresand dummy coded music condition to compareeach type of music to the no music controlcondition. We observed only significant maineffects of music condition on serenity, ΔR2 =.17, ΔF(4, 138) = 7.36, p < .001, excitement,ΔR2 = .08, ΔF(4, 138) = 3.16, p = .016, andirritation, ΔR2 = .10, ΔF(4, 138) = 3.77, p = .006.The main effect of music condition on sadnesswas marginally significant, ΔR2 = .06, ΔF(4, 138)= 2.11, p = .083. An examination of the individualdummy coded terms revealed that relative to theno music condition, serenity was higher in thecalm music condition, β = .23, t(134) = 2.36,p = .020, and lower in the irritating musiccondition, β = −.28, t(134) = −2.81, p = .006;excitement was higher in the exciting musiccondition, β = .21, t(134) = 2.00, p = .047; sadnesswas higher in the sad music condition, β = .20,t(134) = 1.94, p = .054; and irritation was higher

in the irritating music condition, β = .29, t(134) =2.82, p = .006.

Secondary control

For the subsequent analyses, we focused onparticipants who listened to music. We did thisso that we could test our 2 (music valence) × 2(music activation) factorial design. Music valence(−1 = negative, +1 = positive) and activation (−1 =low activation, +1 = high activation) were effects-coded and causal uncertainty scores were centredaround the mean. Specific comparisons with theno music control condition were later examined toshed additional light on the observed effects.

We regressed secondary control scores oncentred causal uncertainty, music valence, musicactivation and all interactions. This analysisrevealed only a Causal Uncertainty × MusicValence interaction, β = .23, t(104) = 2.49,p = .015 (see Figure 2). Simple slopes analysesrevealed a significant effect of music valence onsecondary control for high causal uncertaintyparticipants, β = .40, t(104) = 2.87, p = .005, butnot for low causal uncertainty participants,β = −.08, t(104) = −0.65, p = .518.

In order to compare the different types ofmusic to the no music control at high and low

1

2

3

4

5

6

7

8

9

Low (-1 SD) High (+1 SD)

Seco

ndar

y C

ontr

ol

Causal Uncertainty

Positive Valence

Negative Valence

Figure 2. Secondary control as a function of causal uncertainty

and music valence in Study 2. The dotted line indicates the level of

secondary control in the no music condition (M = 5.20, SD =

1.57). Causal uncertainty was not correlated with secondary control

in the no music condition, r = −.09, p = .632.

2 The reliability of the irritation scale was poor in Study 2 (α = .49), so we also examined the items individually. Weobtained the predicted effect when we examined the item “irritable” but not “hostile” or “jittery”. However, in Study 1, thereliability of the irritation scale was good (α = .71) and we obtained the predicted effect on all three items.

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 9

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

levels of causal uncertainty, we examined the effectof our dummy-coded music condition variables onsecondary control at 1 SD above and 1 SD belowthe mean on causal uncertainty. These analysesrevealed that among high causal uncertaintyparticipants, calm, β = .32, t(134) = 2.33,p = .021 and exciting, β = .30, t(134) = 2.53,p = .013, music increased secondary controlrelative to no music. However, among low causaluncertainty participants, neither calm, β = −.27,t(134) = −1.87, p = .064, nor exciting, β = .11,t(134) = 0.67, p = .501, music significantlyincreased secondary control. Regardless of causaluncertainty level, neither sad nor irritating musicinfluenced secondary control, ps > .58.

Primary control

To examine whether our effects were specific tosecondary control, we next regressed primary controlscores on centred causal uncertainty, music valence,music activation and all interactions. There were nosignificant main or interaction effects of causaluncertainty, music valence or music activation, ps >.21. We should also note that primary and secondarycontrol were not correlated with each other, r = .11,p = .196 (across all conditions).

Discussion

The results indicated that positive affect-inducingmusic increased problem-related secondary controlamong high causal uncertainty individuals, relative tonegative-affect inducing music and no music. As inStudy 1, negative affect-inducing music did notinfluence secondary control. Furthermore, Study 2revealed that the activation level of the positiveaffect-inducing music did not seem to matter. Bothexciting and calming music increased secondarycontrol among individuals with high levels of causaluncertainty. Thus, it does not seem to be the casethat the affective state thought to stem fromsecondary control (i.e., serenity) is the most likelyto increase levels of problem-related secondarycontrol. Rather, positive affect of any activation levelseems to enhance problem-related secondary control.Furthermore, as in Study 1, it was the individuals

high in causal uncertainty, who stood to benefit themost from secondary control, who showed the effect.Lastly, Study 2 revealed that our affect induction didnot influence primary control, showing that oureffects were specific to secondary control

GENERAL DISCUSSION

Summary of findings

The current research reveals that positive affect-inducing music increases secondary control for highcausal uncertainty individuals when they thinkabout a negative event. Specifically, in Study 1, wefound that serenity-inducing music increased prob-lem-related secondary control for high causal un-certainty individuals, relative to no music, and thatthis effect was not due to trait affect. In Study 2, wefound that positive affect-inducing music (high orlow activation) increased problem-related secondarycontrol for high causal uncertainty individuals,relative to negative affect-inducing music and tono music. When we examined primary controlperceptions in Study 2, we found no significanteffects of causal uncertainty or music, indicatingthat the effects observed on secondary control werespecific to secondary control. Furthermore, anexamination of pre-music perceptions of the pro-blems in Study 2 revealed that high and low causaluncertainty participants were bothered by theirproblem to a similar extent, perceived a similarlevel of influence over their problem, and thought itwas equally important to understand the cause oftheir problem. However, as we would expect, highcausal uncertainty participants understood thecauses of their problems to a lesser extent.

Affective influences on secondary control

How does positive affect increase secondary con-trol among causally uncertain people when theyconsider a personal problem? Positive affect couldincrease secondary control through various pro-cesses: by influencing the content of people’sthoughts when they considered their problem,influencing the amount of thought that occurred,or by serving as information when people

TOBIN AND GEORGE

10 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

considered the extent to which they could acceptand adapt themselves to the problem. Becausenegative affect had no influence on secondarycontrol, we suspect that affect as informationmay be the process (Schwarz & Clore, 1983).However, it would be useful to examine theprocess directly in future research.

Interestingly, we found that high activationpositive affect-inducing music worked just as wellas low activation affect-inducing music in terms ofincreasing problem-related secondary control amongcausally uncertain individuals. This was contrary toour prediction that low activation affect would betterfacilitate secondary control. We thought that lowactivation positive affect would seem more relevant(Greifeneder et al., 2011) to being able to accept andadjust oneself to the problem, as serenity is thoughtto be closely associated with secondary control(Morling & Evered, 2006). Furthermore, lowactivation would signal low mobilisation potential(Russell, 2003), which should facilitate accepting theproblem as it is rather than trying to change it.However, only the valence of the music seemed tomatter; as long as it induced positive affect, problem-related secondary control increased among thosehigh in causal uncertainty.

In both studies, we found that negative affect-inducing music did not reduce levels of problem-related secondary control relative to the no musiccondition. Interestingly, we observed a similar effectwith trait negative affect. Although trait positiveaffect was associated with higher levels of secondarycontrol, trait negative affect was not associated withsecondary control. It is possible that this outcomeoccurred because the participants were more likelyto identify the source of negative affect as incidental(e.g., due to the music) and discount its utility formaking judgments about their acceptance of andadjustment to their problem. Indeed, Schwarz andClore (1983) found that people were more likely toavoid the influence of negative relative to positiveaffect on their judgments when the source of theiraffect was made salient.

Lastly, when we examined primary control as anoutcome variable, we found no significant effectsof affect-inducing music or causal uncertainty.Although we had expected that any effects would

be different to those observed on secondary control,it was surprising to find no significant effects at all. Itseemed plausible that high activation positive affectwould increase perceptions of primary control.Indeed, Haase, Poulin, and Heckhausen (2012)found that (high activation) positive affect predictedincreases in primary control striving for career andeducational goals over time. It may be the case thatour methodology was optimised for detecting effectson secondary control rather than primary control.That is, by having participants focus on a problemthat had been bothering them rather than askingthem about goals they were pursuing, we may haveset the stage for detecting affective influences onsecondary control rather than primary control.Another possibility is that the primary controljudgments were less malleable than the secondarycontrol judgments, rendering them less susceptibleto affective influence (Greifeneder et al., 2011).

Moderation by causal uncertainty

As predicted, the effect of positive affect-inducingmusic on problem-related secondary control wasstronger for high causal uncertainty participants.Indeed, positive affect-inducing music did notincrease problem-related secondary control amonglow causal uncertainty participants. Our primaryreason for anticipating a stronger effect amonghigh causal uncertainty individuals was due totheir chronic struggles to understand and thetemptation to disengage. Without secondary con-trol, struggles to understand can lead to escalatingnegative affect over time (Tobin & Raymundo,2010). Secondary control should allow high causaluncertainty individuals to disengage from futileattempts to improve their understanding and feelbetter as a result. Thus, secondary control shouldbe more appealing to high causal uncertaintyindividuals. When conditions facilitate secondarycontrol, they should embrace it.

The effects of causal uncertainty seemed to bespecific to uncertainty, rather than the negativeaffect or loss of control that often accompanies it.In Study 1, we found that the Causal Uncertainty ×Music interaction remained significant when wecontrolled for trait affect. Furthermore, in Study 2,

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 11

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

we found that causal uncertainty was not asso-ciated with initial perceptions of influence overtheir problem, or with perceptions of primarycontrol after they thought about their problem.However, it is possible that low perceived primarycontrol would have effects similar to thoseobserved with high causal uncertainty if it wereexamined as a predictor. Secondary controlresearchers typically view secondary control asbeing of greater utility when actual or per-ceived primary control is low (Bailis et al.,2005; Brandtstädter & Rothermund, 2002;Heckhausen & Schulz, 1995; Morling &Evered, 2006; Rothbaum et al., 1982; Thompsonet al., 1998).

Conclusions

Broadly speaking, our findings are consistent withrecent research that highlights the benefits ofpositive affect (Lyubomirsky, King, & Diener,2005). Although early conceptualizations of pos-itive affect pointed towards its downsides (e.g.,heuristic thinking), more recent conceptualizationsview it as a resource (Aspinwall, 1998). It has beenlinked with a variety of beneficial outcomesincluding greater motivation to pursue one’s goals(Haase et al., 2012), as well as better health,greater sociability, greater altruism, better conflictresolution and original thinking (Lyubomirskyet al., 2005). Similarly, our research shows thatpositive affect-inducing music can help causallyuncertainty people achieve a sense of secondarycontrol over a bothersome event, which shouldpromote psychological and physical well-being.We also found that higher levels of trait positiveaffect were associated with higher levels of sec-ondary control. Our findings highlight theimportance of measuring and controlling forpositive affect when examining secondary controlin correlational research. Although secondarycontrol researchers have typically viewed secondarycontrol as a relatively stable variable that produceschanges in well-being, the current research showsthat the causal path between secondary controland positive affect can run the other way, withpositive affect facilitating secondary control.

The current research shows that secondary con-

trol can be conceptualised as a state variable that

fluctuates as a function of one’s environment. This

approach stands in contrast to the traditional

approach of viewing secondary control as an indi-

vidual difference or developmental variable (Brandt-

städter & Renner, 1990; Heckhausen & Schulz,

1995; Morling & Evered, 2006). Viewing secondary

control as a state variable helps us understand why

secondary control might be easier to achieve in

certain types of situations. We identified positive

affect-inducing music as one situational variable that

can enhance problem-related secondary control, but

any factor that increases positive affect should have

the same effect. Understanding the antecedents of

secondary control helps us predict when people will

be able to say “so what” to their problems and

experience improved psychological and physical

health.

Manuscript received 5 November 2013

Revised manuscript received 14 April 2014

Manuscript accepted 15 April 2014

First published online 6 May 2014

REFERENCES

Aiken, L. S., & West, S. G. (1991). Multiple regression:

Testing and interpreting interactions. Newbury Park,CA: Sage.

Aspinwall, L. G. (1998). Rethinking the role of positiveaffect in self-regulation. Motivation and Emotion, 22,1–32. doi:10.1023/A:1023080224401

Bailis, D. S., Chipperfield, J. G., & Perry, R. P. (2005).Optimistic social comparisons of older adults low inprimary control: A prospective analysis of hospitaliza-tion and mortality. Health Psychology, 24, 393–401.doi:10.1037/0278-6133.24.4.393

Bower, G. H. (1981). Mood and memory. American

Psychologist, 36, 129–148. doi:10.1037/0003-066X.36.2.129

Brandtstädter, J., & Renner, G. (1990). Tenacious goalpursuit and flexible goal adjustment: Explication andage-related analysis of assimilative and accommodativestrategies of coping. Psychology and Aging, 5, 58–67.doi:10.1037/0882-7974.5.1.58

Brandtstädter, J., & Rothermund, G. (2002). The life-course dynamics of goal pursuit and goal adjustment:

TOBIN AND GEORGE

12 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

A two-process framework. Developmental Review,22, 117–150. doi:10.1006/drev.2001.0539

Chang, E. C. (2000). Causal uncertainty and depressivesymptoms: Appraisals and coping as mediatingvariables. Journal of Social and Clinical Psychology,19, 420–436. doi:10.1521/jscp.2000.19.3.420

Chen, E., & Miller, G. E. (2012). “Shift-and-persist”strategies: Why low socioeconomic status isn’t alwaysbad for health. Perspectives on Psychological Science, 7,135–158. doi:10.1177/1745691612436694

Chipperfield, J. G., Newall, N. E., Perry, R. P., Stewart,T. L., Bailis, D. S., & Ruthig, J. C. (2012). Sense ofcontrol in late life: Health and survival implications.Personality and Social Psychology Bulletin, 38, 1081–1092. doi:10.1177/0146167212444758

Edwards, J. A., & Weary, G. (1998). Antecedents ofcausal uncertainty and perceived control: A prospect-ive study. European Journal of Personality, 12,135–148. doi:10.1002/(SICI)1099-0984(199803/04)12:2<135::AID-PER301>3.0.CO;2-Y

Folkman, S., & Moskowitz, J. T. (2000). Stress,positive emotion, and coping. Current Directions in

Psychological Science, 9, 115–118. doi:10.1111/1467-8721.00073

Forgas, J. P. (1995). Mood and judgment: The affectinfusion model (AIM). Psychological Bulletin, 117,39–66. doi:10.1037/0033-2909.117.1.39

Fredrickson, B. L. (2001). The role of positive emotionsin positive psychology: The broaden-and-build the-ory of positive emotions. American Psychologist, 56,218–226. doi:10.1037/0003-066X.56.3.218

Greifeneder, R., Bless, H., & Pham, M. T. (2011).When do people rely on affective and cognitivefeelings in judgment? A review. Personality and SocialPsychology Review, 15, 107–141. doi:10.1177/1088868310367640

Gross, J. J. (2013). Emotion regulation: Taking stockand moving forward. Emotion, 13, 359–365.doi:10.1037/a0032135

Haase, C. M., Poulin, M. J., & Heckhausen, J. (2012).Happiness as a motivator: Positive affectpredicts primary control striving for career andeducational goals. Personality and Social Psychology

Bulletin, 38, 1093–1104. doi:10.1177/0146167212444906

Hall, N. C., Perry, R. P., Chipperfield, J. G., Clifton,R. A., & Haynes, T. L. (2006). Enhancing primaryand secondary control in achievement settings throughwriting-based attributional retraining. Journal of Socialand Clinical Psychology, 25, 361–391. doi:10.1521/jscp.2006.25.4.361

Heckhausen, J., & Schulz, R. (1995). A life-span theoryof control. Psychological Review, 102, 284–304.doi:10.1037/0033-295X.102.2.284

Isen, A. M., & Labroo, A. A. (2003). Some ways inwhich positive affect facilitates decision making andjudgment. In S. L. Schneider & J. Shanteau (Eds.),Emerging perspectives on judgment and decision research

(pp. 365–393). New York, NY: Cambridge UniversityPress.

Jacobson, J. A., Weary, G., & Edwards, J. A. (1999).Certainty-related beliefs and depressive symptomato-logy: Concurrent and longitudinal relationships. SocialCognition, 17, 19–45. doi:10.1521/soco.1999.17.1.19

Kreutz, G., Ott, U., Teichmann, D., Osawa, P., &Vaitl, D. (2008). Using music to induce emotions:Influences of musical preference and absorption.Psychology of Music, 36(1), 101–126. doi:10.1177/0305735607082623

Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2008).International affective picture system (IAPS): Affective

ratings of pictures and instruction manual (TechnicalReport A-8). Gainesville: University of Florida.

Larsen, J. T., Hemenover, S. H., Norris, C. J., &Cacioppo, J. T. (2003). Turning adversity to advant-age: On the virtues of the coactivation of positive andnegative emotions. In L. G. Aspinwall & U. M.Staudinger (Eds.), A psychology of human strengths:

Fundamental questions and future directions for a

positive psychology (pp. 211–225). Washington, DC:American Psychological Association. doi:10.1037/10566-015

Lyubomirsky, S., King, L., & Diener, E. (2005). Thebenefits of frequent positive affect: Does happinesslead to success? Psychological Bulletin, 131, 803–855.doi:10.1037/0033-2909.131.6.803

Morling, B., & Evered, S. (2006). Secondary controlreviewed and defined. Psychological Bulletin, 132,269–296. doi:10.1037/0033-2909.132.2.269

Ng, W., & Diener, E. (2009). Feeling bad? The “Power”of positive thinking may not apply to everyone. Journalof Research in Personality, 43, 455–463. doi:10.1016/j.jrp.2009.01.020

Rothbaum, F., Weisz, J. R., & Snyder, S. S. (1982). Chan-ging the world and changing the self: A two-processmodel of perceived control. Journal of Personality and

Social Psychology, 42, 5–37. doi:10.1037/0022-3514.42.1.5

Rothermund, K., & Brandtstädter, J. (2003). Depres-sion in later life: Cross-sequential patterns andpossible determinants. Psychology and Aging, 18,80–90. doi:10.1037/0882-7974.18.1.80

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 13

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

Russell, J. A. (2003). Core affect and the psychologicalconstruction of emotion. Psychological Review, 110,145–172. doi:10.1037/0033-295X.110.1.145

Schwarz, N., & Clore, G. L. (1983). Mood, misattribu-tion, and judgments of well-being: Informative anddirective functions of affective states. Journal of

Personality and Social Psychology, 45, 513–523.doi:10.1037/0022-3514.45.3.513

Schwarz, N., & Clore, G. L. (2007). Feelings andphenomenal experiences. In E. T. Higgins & A. W.Kruglanski (Eds.), Social psychology: Handbook of basic

principles (2nd ed., pp. 385–407). New York, NY:Guilford.

Thompson, S. C., Thomas, C., Rickabaugh, C. A.,Tantamjarik, P., Otsuki, T., Pan, D., … Sinar, E.(1998). Primary and secondary control over age-related changes in physical appearance. Journal of

Personality, 66, 583–605. doi:10.1111/1467-6494.00025

Tobin, S. J., Loxton, N. J., & Neighbors, C. (2014).Coping with causal uncertainty through alcohol use.Addictive Behaviors, 39, 580–585. doi:10.1016/j.addbeh.2013.11.009

Tobin, S. J., & Raymundo, M. M. (2010). Causaluncertainty and psychological well-being: The mod-erating role of accommodation (secondary control).Personality and Social Psychology Bulletin, 36,371–383. doi:10.1177/0146167209359701

Tobin, S. J., & Tidwell, J. (2013). The role of taskdifficulty and affect activation level in the use ofaffect as information. Journal of Experimental SocialPsychology, 49, 250–253. doi:10.1016/j.jesp.2012.11.011

Tobin, S. J., & Weary, G. (2008). The effects of causaluncertainty, causal importance, and initial attitude onattention to causal persuasive arguments. Social

Cognition, 26, 44–65. doi:10.1521/soco.2008.26.1.44Tsai, J. L., Knutson, B., & Fung, H. H. (2006).

Cultural variation in affect valuation. Journal of

Personality and Social Psychology, 90, 288–307.doi:10.1037/0022-3514.90.2.288

Tsai, J. L., Miao, F. F., Seppala, E., Fung, H. H., &Yeung, D. Y. (2007). Influence and adjustment goals:Sources of cultural differences in ideal affect. Journal ofPersonality and Social Psychology, 92, 1102–1117.doi:10.1037/0022-3514.92.6.1102

Tugade, M. M., & Fredrickson, B. L. (2002).Positive emotions and emotional intelligence. InL. Feldman Barrett, & P. Salovey (Eds.), The

wisdom in feeling: Psychological processes in emotional

intelligence (pp. 319–340). New York, NY: Guil-ford Press.

Watson, D., & Clark, L. A. (1994). The PANAS-X:

Manual for the positive and negative affect schedule -

expanded form. Iowa City: The University of Iowa.Watson, D., Clark, L. A., & Tellegen, A. (1988).

Development and validation of brief measures ofpositive and negative affect: The PANASscales. Journal of Personality and Social Psychology, 54,1063–1070. doi:10.1037/0022-3514.54.6.1063

Weary, G., & Edwards, J. A. (1994). Individualdifferences in causal uncertainty. Journal of Personal-ity and Social Psychology, 67, 308–318. doi:10.1037/0022-3514.67.2.308

Weary, G., & Edwards, J. A. (1996). Causal uncertaintyand related goal structures. In R. Sorrentino &E. T. Higgins (Eds.), The handbook of motivation

and cognition (Vol. 3): The interpersonal context

(pp. 148–181). New York, NY: Guilford.Weary, G., & Jacobson, J. A. (1997). Causal uncertainty

beliefs and diagnostic information seeking. Journal ofPersonality and Social Psychology, 73, 839–848.doi:10.1037/0022-3514.73.4.839

Weary, G., Jacobson, J. A., Edwards, J. A., & Tobin, S.J. (2001). Chronic and temporarily activated causaluncertainty beliefs and stereotype usage. Journal ofPersonality and Social Psychology, 81, 206–219.doi:10.1037/0022-3514.81.2.206

Weary, G., Tobin, S. J., & Edwards, J. E. (2010). Thecausal uncertainty model revisited. In R. M. Arkin,K. C. Oleson, & P. J. Carroll (Eds.), Handbook of the

uncertain self (pp. 78–100). New York, NY: Psycho-logy Press.

Weary, G., Vaughn, L. A., Stewart, B. D., & Edwards, J.A. (2006). Adjusting for the correspondence bias:Effects of causal uncertainty, cognitive busyness, andcausal strength of situational information. Journal of

Experimental Social Psychology, 42, 87–94. doi:10.1016/j.jesp.2005.01.003

Weisz, J. R., Thurber, C. A., Sweeney, L., Proffitt, V.D., & LeGagnoux, G. L. (1997). Brief treatment ofmild-to-moderate child depression using primaryand secondary control enhancement training. Journalof Consulting and Clinical Psychology, 65, 703–707.doi:10.1037/0022-006X.65.4.703

Wrosch, C., Miller, G. E., Scheier, M. F., & Brun dePontet, S. (2007). Giving up on unattainable goals:Benefits for health? Personality and Social Psychology

Bulletin, 33, 251–265. doi:10.1177/0146167206294905

TOBIN AND GEORGE

14 COGNITION AND EMOTION, 2014

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014

APPENDIX

Secondary control items

1. To what extent can you learn to live with this problem

even if the situation does not improve? 1= not at all,

9 = to a great extent.

2. How likely is it that something good will come out of

the problem? 1 = not at all likely, 9 = very likely.

3. To what extent can you accept this problem as a part of

life? 1 = not at all, 9 = to a great extent.

4. To what extent can you adapt to this problem? 1 = not

at all, 9 = to a great extent.

5. This problem is not as bad as it seemed. 1 = strongly

disagree, 9 = strongly agree.

6. Even if this problem persists, I will be okay. 1 = strongly

disagree, 9 = strongly agree.

7. Periods of good and bad luck even out in the end.

1 = strongly disagree, 9 = strongly agree.

8. Other people can help me deal with this problem.

1 = strongly disagree, 9 = strongly agree.

Primary control items

1. How likely is it that you will resolve this problem? 1 =

not at all likely, 9 = very likely.

2. To what extent can you fix this problem? 1 = not at all,

9 = to a great extent.

3. To what extent will you spend time trying to solve this

problem? 1 = not at all, 9 = to a great extent.

4. How much effort will you put into trying to solve this

problem? 1 = no effort at all, 9 = a great deal of effort.

5. I can do something about this problem. 1 = strongly

disagree, 9 = strongly agree.

POSITIVE AFFECT AND SECONDARY CONTROL

COGNITION AND EMOTION, 2014 15

Dow

nloa

ded

by [

UQ

Lib

rary

] at

04:

08 0

7 M

ay 2

014