consumer response to interruption features and need for cognitive closure

13
Consumer response to interruption features and need for cognitive closure MIHAI NICULESCU* , COLLIN R. PAYNE and CUAUHTÉMOC LUNA-NEVAREZ Department of Marketing, New Mexico State University, Box 30001, Dept.5280, Las Cruces, NM 88003-8001, USA ABSTRACT Prior research on interruptions focuses entirely on the process being interrupted and assumes interruption homogeneity. Across two studies, we examine how heterogeneous features of interruptions (i.e., timing, frequency, and perceived pleasantness) and consumer individual differences (i.e., need for cognitive closure (NFCC)) impact consumer response toward a product. We nd interruption features have opposing effects on consumer response for consumers high versus low in NFCCdepending upon the perceived valence of the interrup- tion. Specically, individuals with high NFCC respond better to a product when interruptions are perceived to be pleasant and occur late or infrequently. In contrast, those who have low NFCC respond better to a product when interruptions are perceived to be pleasant and occur early or perceived to be unpleasant and occur infrequently. The role of interruption pleasantness is discussed in terms of its predictive power for perceived pleasant but not perceived unpleasant interruptions. Finally, study ndings are placed within marketing contexts that guide managerial implications. Copyright © 2013 John Wiley & Sons, Ltd. Undeniably, interruptions are part of our life. Interruptions come in different shapes and sizes, yet we know little on how such differences affect our behavior generally and con- sumer behavior specically. Imagine, for example, two customers entering a home improvement store to replace their air-conditioning units at home. One customer feels hard pressed to make a decision immediately to replace the defec- tive unit; to this person, any decision seems better than no decision, as long as the task gets done. The other customer is analytical and searches for the best solution that could replace the old unit at home; to this customer, taking time to make the most informed decision is crucial. Store employees interrupt them during early evaluations of the products by engaging in conversation. Will early interrup- tions (versus late interruptions) make it more or less likely that the customers will evaluate positively the particular product? How does each customers urgency to make a decisionor need for cognitive closure(Webster, 1993) impact nal product evaluations? In another setting, a customers child frequently interrupts while the customer is evaluating a product. Will interruption frequency have any distinct impact on the customers evalu- ation of the product? Furthermore, how would perceived pleasantness of the interruption impact product evaluations? In all cases, interruptions or externally generated, randomly occurring, discrete events that break continuity of cognitive focus on a primary task,(Coraggio, 1990) may have unex- pectedpositive or negativeeffects on consumer response (Xia and Sudharshan, 2002; Liu, 2008; Nelson and Meyvis, 2008). This paper examines the overlooked but potentially vital role interruption features play on consumersproduct attitudes and evaluations as a function of their urgency to nd cognitive closure. Past research on interruptions is varied. One research stream suggests interruptions facilitate decision-making in simple tasks but not in complex ones (Baron, 1986). Another research stream shows interruptions facilitate decision- making in both simple and complex tasks but only if non- conscious thoughts are elicited (Dijksterhuis, 2004; Dijksterhuis and Van Olden, 2006). Still, another stream suggests interruptions improve negative experiences but reduce positive experiences (Nelson and Meyvis, 2008). Interestingly, the aforementioned research shares a com- mon denominator; it focuses entirely on the process being interrupted (i.e., engaging in simple versus complex tasks, having conscious versus non-conscious thoughts, or having positive versus negative experiences) and ignores the features of the interruption itself. Although research acknowledges that interruptions are not homogenous (e.g., Speier et al., 1999), it has yet to examine their heterogeneity; there are few known studies examining interruption features. Furthermore, con- sumers are not homogenous either. Important individual differences, such as NFCC, predict consumer judgment and decision-making (Houghton and Grewal, 2000). This research investigates how heterogeneous features of interruptionsin this case their timing, frequency, and perceived pleasantness impact consumer product attitudes and evaluations across different levels of NFCC (Kruglanski and Webster, 1996). The paper is structured as follows. First, we review relevant literature regarding interruptions and interruption features. Next, we explain how the impact of interruptions on product attitudes and evaluations may be inuenced by individual differences in NFCC. Finally, we describe two studies, report their results, and discuss their relevance to marketing contexts. BACKGROUND Interruptions are pervasive in many human-related activities. Yet, research on interruptions is rather scarce in marketing and mostly occurs in nonbusiness contexts. Findings are *Correspondence to: Mihai Niculescu, Assistant Professor of Marketing, Department of Marketing, New Mexico State University, Box 30001, Dept. 5280, Las Cruces, NM 88003-8001, USA. E-mail: [email protected] Associate Professor of Marketing at New Mexico State University Assistant Professor of Marketing at Sacred Heart University Copyright © 2013 John Wiley & Sons, Ltd. Journal of Consumer Behaviour, J. Consumer Behav. 13: 6072 (2014) Published online 28 December 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cb.1459

Upload: cuauhtemoc

Post on 29-Mar-2017

217 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Consumer response to interruption features and need for cognitive closure

Consumer response to interruption features and need for cognitive closure

MIHAI NICULESCU*†, COLLIN R. PAYNE† and CUAUHTÉMOC LUNA-NEVAREZ‡

Department of Marketing, New Mexico State University, Box 30001, Dept.5280, Las Cruces, NM 88003-8001, USA

ABSTRACT

Prior research on interruptions focuses entirely on the process being interrupted and assumes interruption homogeneity. Across two studies,we examine how heterogeneous features of interruptions (i.e., timing, frequency, and perceived pleasantness) and consumer individualdifferences (i.e., need for cognitive closure (NFCC)) impact consumer response toward a product. We find interruption features haveopposing effects on consumer response for consumers high versus low in NFCC—depending upon the perceived valence of the interrup-tion. Specifically, individuals with high NFCC respond better to a product when interruptions are perceived to be pleasant and occur lateor infrequently. In contrast, those who have low NFCC respond better to a product when interruptions are perceived to be pleasant and occurearly or perceived to be unpleasant and occur infrequently. The role of interruption pleasantness is discussed in terms of its predictive powerfor perceived pleasant but not perceived unpleasant interruptions. Finally, study findings are placed within marketing contexts that guidemanagerial implications. Copyright © 2013 John Wiley & Sons, Ltd.

Undeniably, interruptions are part of our life. Interruptionscome in different “shapes and sizes”, yet we know little onhow such differences affect our behavior generally and con-sumer behavior specifically. Imagine, for example, twocustomers entering a home improvement store to replacetheir air-conditioning units at home. One customer feels hardpressed to make a decision immediately to replace the defec-tive unit; to this person, any decision seems better than nodecision, as long as the task gets done. The other customeris analytical and searches for the best solution that couldreplace the old unit at home; to this customer, taking timeto make the most informed decision is crucial. Storeemployees interrupt them during early evaluations of theproducts by engaging in conversation. Will early interrup-tions (versus late interruptions) make it more or less likelythat the customers will evaluate positively the particularproduct? How does each customer’s urgency to make adecision—or “need for cognitive closure” (Webster, 1993)—impact final product evaluations?

In another setting, a customer’s child frequently interruptswhile the customer is evaluating a product. Will interruptionfrequency have any distinct impact on the customer’s evalu-ation of the product? Furthermore, how would perceivedpleasantness of the interruption impact product evaluations?In all cases, interruptions or “externally generated, randomlyoccurring, discrete events that break continuity of cognitivefocus on a primary task,” (Coraggio, 1990) may have unex-pected—positive or negative—effects on consumer response(Xia and Sudharshan, 2002; Liu, 2008; Nelson and Meyvis,2008). This paper examines the overlooked but potentiallyvital role interruption features play on consumers’ productattitudes and evaluations as a function of their urgency tofind cognitive closure.

Past research on interruptions is varied. One researchstream suggests interruptions facilitate decision-making insimple tasks but not in complex ones (Baron, 1986). Anotherresearch stream shows interruptions facilitate decision-making in both simple and complex tasks but only if non-conscious thoughts are elicited (Dijksterhuis, 2004;Dijksterhuis and Van Olden, 2006). Still, another streamsuggests interruptions improve negative experiences butreduce positive experiences (Nelson and Meyvis, 2008).

Interestingly, the aforementioned research shares a com-mon denominator; it focuses entirely on the process beinginterrupted (i.e., engaging in simple versus complex tasks,having conscious versus non-conscious thoughts, or havingpositive versus negative experiences) and ignores the featuresof the interruption itself. Although research acknowledges thatinterruptions are not homogenous (e.g., Speier et al., 1999), ithas yet to examine their heterogeneity; there are few knownstudies examining interruption features. Furthermore, con-sumers are not homogenous either. Important individualdifferences, such as NFCC, predict consumer judgment anddecision-making (Houghton and Grewal, 2000). This researchinvestigates how heterogeneous features of interruptions—inthis case their timing, frequency, and perceived pleasantness—impact consumer product attitudes and evaluations acrossdifferent levels of NFCC (Kruglanski and Webster, 1996).

The paper is structured as follows. First, we reviewrelevant literature regarding interruptions and interruptionfeatures. Next, we explain how the impact of interruptionson product attitudes and evaluations may be influenced byindividual differences in NFCC. Finally, we describe twostudies, report their results, and discuss their relevance tomarketing contexts.

BACKGROUND

Interruptions are pervasive in many human-related activities.Yet, research on interruptions is rather scarce in marketingand mostly occurs in nonbusiness contexts. Findings are

*Correspondence to: Mihai Niculescu, Assistant Professor of Marketing,Department of Marketing, New Mexico State University, Box 30001, Dept.5280, Las Cruces, NM 88003-8001, USA.E-mail: [email protected]†Associate Professor of Marketing at New Mexico State University‡Assistant Professor of Marketing at Sacred Heart University

Copyright © 2013 John Wiley & Sons, Ltd.

Journal of Consumer Behaviour, J. Consumer Behav. 13: 60–72 (2014)Published online 28 December 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/cb.1459

Page 2: Consumer response to interruption features and need for cognitive closure

not consistent, and contrary to popular belief, interruptionscan increase human performance under certain conditions,described by specific cognitive, or affective processes(Baron, 1986).

Early research in psychology has found that individualshave better memory recollection of interrupted unfinished(versus finished) tasks (i.e., the Zeigarnik effect). Zeigarnik(1967) tested this hypothesis using jigsaw-type puzzles andfound that individuals remembered more details when inter-ruptions occurred prior to task closure, whereas memoryperformance dropped significantly once respondents com-pleted the task. The performance drop was due to a transferof information from short to long term memory afterindividuals reached closure, which decreased accessibility todetails and increased response latencies (Marsh et al., 1998).

More recently, the distraction–conflict theory (Baron,1986) suggests that in simple tasks, interruptions trigger higharousal levels and narrow attention that, in turn, allow indi-viduals to dismiss irrelevant cues more easily and improvedecision accuracy. On the contrary, in complex decisions,interruptions excessively narrow attention, resulting inmissing relevant information and leading to performancedeterioration (Payne, 1982; Baron, 1986). Nevertheless,theoretical predictions in complex tasks have been confirmedonly when the presentation format was symbolic (versusspatial) (Speier, 1999).

Contrasting the distraction–conflict theory, the deliberation-without-attention theory predicts that conscious (versus non-conscious) cognitive processing moderates the impact ofinterruptions on decision quality (Dijksterhuis, 2004;Dijksterhuis et al., 2006; Gonzalez-Vallejo et al., 2008).Research suggests that the quality of non-conscious decisionsremains consistently good in both simple and complex situa-tions, whereas the quality of conscious decisions declines ascomplexity increases (Dijksterhuis and Nordgren, 2006).Recent studies question, however, the deliberation-without-attention hypothesis and caution on the generalizability ofthe results (Rey et al., 2009; Waroquier et al., 2010).

Interestingly, interruptions not only interact with but mayalso change cognitive processing styles. For example, Liu(2008) found that when consumers faced feasibility–desirabilitytrade-offs, interruptions caused reversals in information process-ing style, from bottom–up (i.e., data driven) to top–down (i.e.,goal-directed), and led to preference for desirability goals (i.e.,pursuing the value of the end state of an action) over feasibilitygoals (i.e., pursuing the ease of achieving the desired outcome).

Finally, in an equally interesting departure from tradi-tional interruption research, Xia and Sudharshan (2002) andSpeier (1999) suggest that, in addition to primary taskcharacteristics, outcomes may be influenced by cognitive orsocial dimensions (i.e., features) of interruptions, as well asby decision-maker characteristics (e.g., gender, expertise,and coping style; see also Kirmeyer, 1988).

Consider Table 1. Research on interruption features hasassumed cognitive, social, and affective interruption featuresto be independent of each other and has overlooked thepossibility of not only cross-feature interactions but alsofeature interactions with individual differences, such as NFCC.Our research follows this yet untested route and provides early

experimental evidence on the interactive effects of interruptionfeatures and NFCC on consumer response. More specifically,we investigate how timing, frequency, or perceived pleasant-ness of interruptions impact important consumer-relatedvariables, such as product evaluations and attitudes towardthe product across different levels of NFCC.

Similar to the research in Table 1, we build upon a theo-retical framework from an adjacent research field to compen-sate for the absence of a theory on interruptions features. Weuse consumer’s NFCC (Webster, 1993; Kruglanski andWebster, 1996) to derive our working hypotheses and delin-eate boundary conditions for the effects of interruptionfeatures on product attitudes and evaluations (Figure 1).

The need for cognitive closure: A framework to studyinterruption featuresSimilar life situations may trigger different reactions inindividuals. Some individuals prefer quick and definitiveanswers to particular questions. Others will be more cautiousbefore drawing conclusions. Such heterogeneity in an indi-vidual’s motivation toward closure has been measured byresearchers via the NFCC construct (Webster and Kruglanski,1994). Described by a dual stage process of “seizing” onclosure (i.e., the urgency tendency of making a decision orforming an opinion) and “freezing” (i.e., the permanencetendency of keeping the decision or opinion), NFCC ismeasured “along a continuum anchored at one end with astrong need for closure and at the other end with a strong needto avoid closure” (Kruglanski and Webster, 1996).

Need for cognitive closure is considered more than asituational variable—as individual trait differences in thisconstruct are well documented; numerous studies havecaptured individual differences in NFCC equally well throughexperimental manipulation (e.g., Kardes et al., 2004; Cronleyet al., 2005) or psychometric measurement (e.g., Webster andKruglanski, 1994; Houghton and Grewal, 2000). Conse-quently, NFCC has been used extensively as one possiblefactor that interacts with a variety of information processes—such as primacy effects, stereotyping, anchoring (Kruglanskiand Freund, 1983), attribute-based versus alternative-basedsearch (Choi et al., 2008), selective information processing(Kardes et al., 2004), and cognitive biases—such as omissionneglect (Han, 2011) and correspondence bias (Webster, 1993).

Individuals with a high NFCC make decisions withminimal amounts of information and avoid informationincongruent with prior beliefs (Kruglanski, 1990; Kruglanskiet al., 1993; Webster and Kruglanski, 1994). Ambiguityaversion, lower levels of information processing, and limitedfocus on stereotypical (versus diagnostic) information predis-poses high NFCC individuals to attribution biases (e.g.,attributing behavior to personality rather than a situation) andhigher judgmental confidence (Kruglanski and Mayseless,1988). Furthermore, high (versus low) NFCC individuals aremore likely to maintain their initial evaluation of a message orjudgment of a product and less likely to change their perspec-tive (Kardes et al., 2007).

In contrast to high NFCC individuals, those low in NFCCperceive lack of closure as less unusual. They are more com-fortable with ambiguous situations and more likely to engage

Consumer response to interruption features and need for cognitive closure 61

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 3: Consumer response to interruption features and need for cognitive closure

Table

1.Relevantresearch

oninterruptio

nfeatures

Interruptio

ndimension

Theoretical

fram

ework

Primarytask

Decision-maker

characteristics

andsamplesize

Findings/comments

Cognitiv

edimensionsof

interruptio

ns

Gillie

and

Broadbent

(1989)

None

Com

puter-based

adventuregame

four

experiments;n=40;

age18

–45

Mem

oryload

andopportunity

torehearse

maintask

played

norole

indeterm

ining

whether

aninterruptio

nwas

disruptiv

eor

not.Interruptio

nlength

hadno

effect

onrespondents.Interruptio

nssimilar(versus

dissim

ilar)to

themaintask

weremore

disruptiv

e.Speieret

al.

(1999)

Inform

ation

overload;

distraction/

confl

icttheory

Sim

pleversus

complex

decision-m

aking;

Productionmanagem

ent

problems(PM)

Twoexperiments;

n=238students

Interruptio

ns(versusno

interruptio

ns)

show

edno

significant

differencesin

task

performance.Interruptio

nwith

dissim

ilar

(versussimilar)contentim

paired

toa

higher

extent

task

performance.Interruption

frequency(4

vs12)decreasedtask

performance

forsimpleandcomplex

tasks.

Czerw

inski

etal.(2000)

None

Workin

MSExcel

Twoexperiments;n=23

MSOffice

users,age26

–56

Interruptio

ntim

ing(interruptionoccurred

viainstantmessageswhile

consum

ers

accessed

afile,its

toolbarmenu,

orthe

contentof

thefile)im

pactsperformance

task.E

arly

interruptio

nsareless

disruptiv

e.Interruptio

nsaremoreharm

fulifthesubject

becomes

fully

immersedin

thetask.

Xia

and

Sudharshan

(2002)

Inform

ationprocessing

—constructiv

echoice

processes(Bettm

anet

al.,1998)

Onlineenvironm

ent

—shopping

task

Fourexperiments;n=240

Interruptio

ns(versusno

interruptio

ns)

increase

thetim

espenton

task.Frequent

(versusinfrequent)interruptio

nsdidnot

change

task

duratio

n.Givingparticipants

thecontrolover

theinterruptio

nalleviated

thenegativ

eeffect

ofinterruptio

nfrequency

onsatisfaction.

Early

interruptio

nsaremore

easily

processedthan

late

interruptio

ns.

Speieret

al.

(2003)

Theoryof

cognitive

fit

Sim

pleversus

complex

decision-m

aking;

PM

problems

One

experiment;

n=136students

Contrasteddifferentpresentatio

nform

ats

(sim

ple-symbolic

tasksversus

complex-

symbolic

tasks)

inthecontextof

simple

versus

complex

decision-m

akingand

interruptio

ns.

Socialdimension

sof

interruptio

nsPerlow

(1999)

None

New

productdevelopm

ent

undertim

epressure;

real

world

setting

Qualitativestudy;

n=45

softwareengineers

Sam

esocial

interactions

may

lead

topositiv

eandnegativ

econsequences,dependingon

synchronizationbetweengroupmem

bers.

Highfrequencynon-synchronized

interruptio

nslead

toloss

ofproductiv

ity.

Synchronizedinterruptio

ns(allgroup

mem

bers

engage

ininteractionat

thesame

time)

lead

topositiv

eperformance.

Bondand

Titu

s(1983)

Theoryof

social

facilitation

Sim

pleversus

complex

Metaanalysis;

n=241studies

Findingssupporttheidea

that

socially

ladeninterruptio

nsim

provespeedof (Contin

ues)

62 M. Niculescu et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 4: Consumer response to interruption features and need for cognitive closure

in extensive search of information prior to a purchase(Kardes et al., 2007). Although decision-making may takemore time, low NFCCs are at lower risk of finding less-than-optimal solutions and are usually more accurate in theprocess (Webster and Kruglanski, 1994). Such differencesbetween individuals in NFCC should influence responses tointerruptions. Next, we document potential interactions ofNFCC with the interruption features of timing, frequency,and perceived pleasantness. Where predictions are possible,we develop working hypotheses.

Interruption timingIn general, consumers are generally receptive and open tonew information in the exploratory (i.e., search) phase ofthe buying process but become selective about informationthey attend to as the evaluation process progresses (Xia andSudharshan, 2002). In later stages, greater attention is allo-cated to decision-making, as cognitive resources becomelimited (Dijksterhuis and Aarts, 2010). Research byCzerwinski et al. (2000) find that early interruptions are lessinvasive than later interruptions. In their study, individualswere interrupted while opening a computer file, goingthrough the toolbar menu, or working with file content. Re-sults suggested that once individuals became immersed inthe primary task, any interruption negatively impacted taskperformance. Nevertheless, individual differences in NFCCmay change these effects as the moment of closure (i) variesacross individuals and (ii) carries a high cognitive load thatreduces available cognitive resources necessary for othertasks. Hence, an interruption during the closure process—independent of the time the interruption occurs—negativelyimpacts the primary task or evaluation process in the caseof our proposed studies (Czerwinski et al., 2000).

Extant literature suggests that high NFCC individualshave a tendency to seek closure—that is, seizing and freezing—in early stages of decision-making. In contrast, low NFCCindividuals tend to postpone closure until late stages of thedecision-making process (Kruglanski et al., 1993). From thisperspective, an early (versus late) interruption should disruptclosure and negatively impact evaluations for individuals withhigh NFCC, and a late (versus early) interruption should dis-rupt closure and lower evaluations for those with low NFCC.

By the same logic, high NFCCs should be less affectedby a late interruption, because by that time, a judgmentwill have already been made. Low NFCC individualsshould be less influenced by early (versus late) interrup-tions, as they seek to make the best decision withoutbeing time constrained and closure does not occur untillate in the decision process. Consequently,

H1a: High NFCC individuals will exhibit lower productratings and attitudes during early (versus late) interruptions.

H1b: Low NFCC individuals will exhibit lower productratings and attitudes during late (versus early) interruptions.

H2a: High (versus low) NFCC individuals will exhibit lowerproduct ratings and attitudes during early interruptions.

H2b: Low (versus high) NFCC individuals will exhibitlower product ratings and attitudes during late interruptions.T

able

1.(Contin

ued)

Interruptio

ndimension

Theoretical

fram

ework

Primarytask

Decision-maker

characteristics

andsamplesize

Findings/comments

completingsimpletasksbutdecrease

speedforcomplex

ones.P

resenceof

others

during

prim

arytask

impacts

respondents’

task

accuracy.

Affectiv

edimensions

ofinterruptio

nsNoknow

nstudies

N/A

N/A

N/A

Alth

ough

theliteraturesuggeststhat

affect

hasastrong

impact

oncognitive

processes

(e.g.,Van

Kleef,2

009),n

oknow

nstudies

have

investigated

theeffect

ofaffective

dimensionsof

interruptio

ns.

Consumer response to interruption features and need for cognitive closure 63

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 5: Consumer response to interruption features and need for cognitive closure

Interruption frequencyDue to their need for accuracy and high perceived account-ability, low NFCC individuals evaluate and select informa-tion over a long period before reaching a conclusion(Kruglanski et al., 1993; Kardes et al., 2007). Any interrup-tions may delay the process and increase perceived durationof the task (Schiffman and Greist-Bousquet, 1992) but willunlikely change the evaluation significantly. The highperceived accountability for their actions, the ability tosearch for diagnostic information, and the lack of perceivedtime pressure to achieve the task will ensure that evaluationbiases are kept under control, and relevant information isintegrated in the evaluation (Kruglanski and Freund, 1983;Webster and Kruglanski, 1994; Houghton and Grewal,2000). Consequently, the number of interruptions shouldbe less relevant for final product evaluations of lowNFCC individuals.

H3a: Low NFCC individuals will exhibit no significantdifference in product ratings and attitudes during lessfrequent (versus more frequent) interruptions.

Conversely, frequent interruptions that force the respondentto postpone closure may frustrate high NFCC individuals(Kruglanski et al., 1993), as interruptions increase the perceivedtask duration and impede goal achievement (i.e., closure)(Schiffman and Greist-Bousquet, 1992). High NFCCs’ verynature to seek closure and to make decisions on the basis ofstereotypical (non-diagnostic) information leads to biasedevaluations (Kruglanski and Mayseless, 1988). The moretime they spend to reach closure, the more they developambiguity aversion (Shiota et al., 2007). In turn, this leadsto frustration and increased biased evaluations after frequent(versus infrequent) interruptions.

H3b: High NFCC individuals will exhibit lower productratings and attitudes during more frequent (versus lessfrequent) interruptions.

Perceived pleasantness of the interruptionThere are significant differences between high and lowNFCC individuals in terms of searching, selecting, and inte-grating information to form an evaluation. On the one hand,low NFCC individuals perform a comprehensive search ofinformation, select diagnostic information and ignore non-diagnostic information, and integrate relevant informationinto a judgment (Kardes et al., 2007). On the other hand,high NFCC individuals rely on cues and stereotypical(non-diagnostic) information as much as on relevant infor-mation to form an evaluation (Kruglanski and Mayseless,1988). For example, Van Kleef et al. (2004) found thatlow NFCC participants in a negotiation task were influencedby their opponents’ feelings (i.e., anger versus happiness)when motivated to consider additional information. On theother hand, high NFCCs ignored the emotional states of theopponents during the task. From this perspective, high NFCCindividuals may ignore the perceived pleasantness of aninterruption and rate products similarly, whereas low NFCCindividuals will somehow integrate the information in theirdecision (Van Kleef et al., 2004). However, these resultsmay not hold in all contexts for the following three reasons.

First, we can question the results obtained for highNFCCs. High NFCC individuals lack motivation to distin-guish between relevant and irrelevant information and inte-grate anything that comes at hand during their productevaluations; in turn, they often have strong beliefs biasingtheir evaluations (Kruglanski and Mayseless, 1988; Websterand Kruglanski, 1994; Houghton and Grewal, 2000). Contraryto Van Kleef et al. (2004), this body of literature suggests thathigh NFCCs should actually be influenced by the perceivedpleasantness of interruptions—possibly non-consciously.

Second, assuming that low NFCCs will be impacted bythe perceived pleasantness of interruptions, the directionalityof the effects on consumer response is less clear. Intuitively,interruptions perceived as relatively pleasant should have apositive impact on consumer evaluations, whereas less

Consumer Response

Product Attitudes

Product Evaluations

HIGH NEED FOR COGNITIVE CLOSURE

Need for Cognitive Closure

Affective Dimension of Interruptions

LOW NEED FOR COGNITIVE CLOSURE

Cognitive Dimensions of Interruptions

INTERRUPTION TIMING (Study 1)

INTERRUPTION FREQUENCY (Study 2)

INTERRUPTION VALENCE(Studies 1&2)

Figure 1. A framework for the study of interruption features and need for cognitive closure.

64 M. Niculescu et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 6: Consumer response to interruption features and need for cognitive closure

pleasant interruptions should negatively influence them (i.e.,positivity/negativity bias) (Herr et al., 1991; Perlow andWeeks, 2002; Chan et al., 2010). However, Ein-Gar et al.(2012) find that negative information not central to the valueproposition may actually lead to more positive evaluations ofthe product. The authors suggest that a negative piece ofinformation “increases the salience of the positive informa-tion” and ultimately “makes the positive information seemmore positive” (i.e., the “blemishing effect”). Both situationsare equally reasonable and well-supported.

Third, when information is relevant, such as the emotionalstate of the opponent in Van Kleef and colleagues’ (2004)negotiation task, the information is expected to be integratedin the evaluation process by low NFCCs. What if, however,the emotional state was irrelevant to the task? Low NFCCindividuals may acknowledge the irrelevancy in the interrup-tion, label it as non-diagnostic, and ignore it during theevaluation process. In this case, the prediction would be thatlow NFCCs would not be affected by interruptions perceivedmore or less pleasant.

Considering the aforementioned reasons, we do not offerconsumer response predictions vis-à-vis the perceived pleas-antness of an interruption, as the literature suggests that itmay have a positive, neutral, or negative effect across bothhigh and low NFCC individuals. However, we do includeperceived pleasantness of an interruption as an independentvariable in both studies 1 and 2 to explore how it interactswith NFCC, interruption timing (study 1), and interruptionfrequency (study 2) on two consumer-related variables:product attitudes and product evaluations. Attitudes areprecursors of evaluations; an evaluation integrates attitudesalong with other information related to the context, stimulus,and goals (Cunningham et al., 2007).

STUDY 1: TIMING AND PERCEIVED PLEASANTNESSOF INTERRUPTIONS

Participants and procedureOne-hundred and twenty students from a large Southwesternuniversity participated in exchange for extra credit. Partici-pants were told they would be taking part in a short consumersurvey by clicking on a link provided at the bottom of arecruitment email. Once they read the consent form, partici-pants were presented a set of 13 product statements—adapted from Gurhan-Canli (2003)—describing an electronicproduct (i.e., HDTV set; see Appendix A). During the task,participants were randomly assigned to be interrupted early(after reading three statements) or late (after reading 10 state-ments) by an anagram solving exercise. Anagrams usingwords unrelated to the task can efficiently break the task athand; each anagram rendered itself to at least 50 solutions(Dijksterhuis et al., 2006). After two minutes of attemptingto solve anagrams, all participants continued reading productstatements and responded to questions relating to attitudestoward and evaluation of the product, respectively. Thesurvey ended with demographic questions, six questionsregarding the NFCC (Kardes et al., 2007), a questionregarding participants’ anagram experience to measure its

perceived pleasantness, and four questions regarding theirknowledge of the particular product category (Appendix B).

Scale reliability. Perceived knowledge of the product cate-gory was measured by averaging four items (α = 0.96; seeAppendix B). Attitude toward the product was measured byaveraging responses to three statements: “The product islikely to possess the stated characteristics,” “I react favorablyto the product,” and “I feel positively toward the product(α = 0.79).” Participants’ evaluation of the product wasmeasured by averaging responses anchored to five 7-pointLikert scales: “negative/positive,” “unfavorable/favorable,”“bad/good,” “dissatisfactory/satisfactory,” and “undesirable/desirable”. An index was created by averaging the ratingsobtained on the five items (α= 0.96).

ResultsTo find balance between internal and external validity we bothmeasure (NFCC and perceived pleasantness of the interruption)and manipulate (interruption timing) key independent variables.A median split was performed on participants’ anagram experi-ence ratings (i.e., perceived pleasantness) and NFCC scores.This resulted in a 2 (timing: early versus late) x 2 (perceivedpleasantness: more versus less) x 2 (NFCC: low versus high)between subjects design. A simple correlation test between allindependent variables (i.e., manipulated andmeasured) providesevidence of their independence (all p-values n.s.), thus reducingthe potential of their confounding. In addition, we tested thetheoretically plausible alternative of a mediating effect ofperceived pleasantness by means of path analysis. The analysisthat treated timing and NFCC as exogenous variables andperceived pleasantness (plausible mediator), product evaluation,and attitude toward the product as endogenous variablessuggested no significant mediation process of perceived pleas-antness and, hence, supported its use as an independent variablein the model. Gender, age, and knowledge of product categorywere used as covariates in the model but were not found to besignificant predictors of the dependent variables.

Product attitudes. A three-way interaction between timing,perceived pleasantness, andNFCCwas found to be statisticallysignificant, F(1,109) = 4.46, p< 0.05 (Figure 2); all maineffects and two-way interactions were nonsignificant. Weassessed our hypotheses by performing planned contrasts onthe conditions of perceived pleasantness of the interruption,separately, as appropriate analyses to follow up the significantthree-way interaction (Meyer, 1991; Montgomery, 2000).

We hypothesized in H1a that high NFCC individuals willexhibit lower product attitudes during early (versus late)interruptions—as early interruptions disrupt the early closureprocess specific to the group. Indeed, we find that for highNFCC individuals, attitudes measured after early interrup-tions were significantly lower than after late interruptions.However, hypothesis H1a only holds for interruptionsthat were perceived to be more pleasant (MEarly = 3.93;MLate = 4.72, t(109) = 1.76, p< 0.051).

1We perform one-tailed t-tests—that is, we test predicted directionality inH1: M1<M2 or M1>M2 but not both.

Consumer response to interruption features and need for cognitive closure 65

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 7: Consumer response to interruption features and need for cognitive closure

In H1b, we hypothesized that low NFCC individuals willexhibit lower attitudes during late (versus early) interrup-tions. We find confirmation of this case for the morepleasantly perceived interruptions only. In our analysis, lowNFCC individuals report worse attitudes for late (versusearly) more pleasant interruptions (MEarly = 4.73; MLate =3.80, t(109) =�2.12, p< 0.05).

However, both high and low NFCC individuals show nostatistically significant change in product attitudes for early(versus late) less pleasantly perceived interruptions (Figure 2).In the case of the interruptions perceived less pleasant, H1a andH1b are not supported.

As hypothesized in H2a and H2b, we find that when theinterruption experience is perceived to be more pleasant, high(versus low) NFCC consumers report significantly lowerproduct attitudes in early interruptions (MLowNFCC= 4.73;MHighNFCC = 3.93, t(109) = 1.82, p< 0.05) and significantlyhigher product attitudes in late interruptions (MLowNFCC= 3.80;MHighNFCC = 4.72, t(109) =�2.05, p< 0.05). However, resultssuggest no significant product attitude differences betweenhigh and low NFCC groups when interruptions wereperceived as less pleasant and occurred early or late. Con-sequently, H2a and H2b are supported for more pleasantlyperceived interruptions but not supported for less pleasantlyperceived interruptions.

Product evaluation. A three-way interaction between timing,perceived pleasantness, and NFCCwas found to be significant,F(1,109) = 4.11, p< 0.05 (Figure 3). Contrast analyses findsupport for H1a and H1b when the interruptions are perceivedas more pleasant. Specifically, for high NFCC individuals, late(versus early) interruptions improve product evaluationstoward the product by a significant margin (Mearly_pleasant =4.67; Mlate_pleasant = 5.91, t(109) = 1.91, p< 0.05) (Figure 3).In contrast, low NFCC individuals report significantly worseproduct evaluations for late (versus early) interruptions whenthey are perceived to be more pleasant (Mearly_pleasant = 5.18;Mlate_pleasant = 4.00, t(109) = 1.86, p< 0.05). Finally, productevaluations show no statistically significant change for late(versus early) interruptions in both high and low NFCC groupswhen they are perceived to be less pleasant (Figure 3). Hence,H1a and H1b are supported when the interruptions are perceivedto be more pleasant but not supported when they are perceivedto be less pleasant.

Unlike the evaluation of product attitudes, there was nosignificant difference between high and low NFCC productevaluations for early interruptions perceived to be morepleasant—although the results were in the anticipated direc-tion (MLowNFCC_pleasant _early = 5.18;MHighNFCC_pleasant _early =4.67, t(109) = 0.81, n.s.); H2a was not supported.

(a) Pleasant Interruptions

(b) Unpleasant Interruptions

5.18

4.00

4.67

5.91

3.60

4.10

4.60

5.10

5.60

6.10

Early Interruptions(n=34)

Late Interruptions(n=32)

Low NFCC

High NFCC

4.13

5.10

4.83

5.48

3.60

4.10

4.60

5.10

5.60

6.10

Early Interruptions(n=27)

Late Interruptions(n=27)

Low NFCCs

High NFCCs

Figure 3. Impact of interruption timing and need for cognitive clo-sure on product evaluations for (a) pleasant interruptions and (b) un-

pleasant interruptions.

(a) Pleasant Interruptions

(b) Unpleasant Interruptions

4.73

3.80

3.93

4.72

3.60

3.80

4.00

4.20

4.40

4.60

4.80

Early Interruptions(n=34)

Late Interruptions(n=32)

Low NFCC

High NFCC

3.98

4.63

3.90

4.30

3.60

3.80

4.00

4.20

4.40

4.60

4.80

Early Interruptions(n=27)

Late Interruptions(n=27)

Low NFCC

High NFCC

Figure 2. Effect of interruption timing and need for cognitiveclosure on product attitudes for (a) pleasant interruptions and

(b) unpleasant interruptions.

66 M. Niculescu et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 8: Consumer response to interruption features and need for cognitive closure

However, high (versus low) NFCC consumers evaluatedthe product better when the interruptions were late andperceived to be more pleasant (MLowNFCC_pleasant_late = 4.00;MHighNFCC_pleasant_late = 5.91, t(109) =2.94, p< 0.01); theevidence supports H2b. Finally, we find that products are ratedsimilarly by high and low NFCC respondents when theinterruptions were perceived as less pleasant—suggesting nosupport for H2a and H2b (MLowNFCC_unpleasant _early = 4.13 versusMHighNFCC_unpleasant _early = 4.83, t(109) =1.00, n.s., andMLowNFCC_unpleasant _late = 5.10 versusMHighNFCC_unpleasant _early =5.48, t(109)=0.54, n.s.).

DiscussionResults of study 1 suggest the perceived pleasantness ofthe interruption is paramount regarding our proposedhypotheses. For example, in the case of perceived pleasantinterruptions, late interruptions are recommended to earlyinterruptions for high NFCC individuals. Early interruptionsmay occur concomitantly to “seizing and freezing”processes in high NFCC individuals and negatively affectattitudes toward and evaluation of a product. Late interrup-tions occur after the individual has already closed theirevaluation process (i.e., after “seizing” and “freezing”)and, hence, do not pose a similar challenge.

Low NFCCs, however, react more favorably to earlyinterruptions perceived to be more pleasant. Unlike highNFCCs, who seek early closure, low NFCCs search exten-sively for information before reaching a decision (i.e., lateclosure). From this viewpoint, a late (versus early) interrup-tion should negatively affect product evaluations, as it occursduring closure-seeking (i.e., seizing and freezing). This effectexplains why low NFCCs report lower product evaluationsafter late (versus early) interruptions that are perceived tobe more pleasant. However, the same effect has not beenconfirmed in the presence of interruptions perceived to beless pleasant—all respondents evaluated similarly early andlate interruptions in this case. Further research exploringpotential mediating effects may reveal new evidence ofexisting boundary conditions, as these results appear to beconsistent across dependent variables in our study.

Finally, some of the results run counterintuitive to salestraining guidelines that advise against interrupting individ-uals under time pressure and focus on individuals with moretime at hand—as the latter consumers may find interruptionsless disturbing. Because interruptions in the marketplace aresupposedly professional and pleasant, an early (late) inter-ruption strategy may work best in approaching low (high)NFCC individuals.

Although early and late interruptions were found tomoderate the effect of NFCC and perceived pleasantness ofthe interruption on product attitudes and evaluations, itwould be useful to understand additional boundary condi-tions on our dependent variables. Another common featureof interruptions is their frequency. Accordingly, study 2 aimsto understand how interruption frequency—in addition toperceived pleasantness of the interruption and NFCC—impacts consumer response.

STUDY 2: FREQUENCY AND PERCEIVEDPLEASANTNESS OF INTERRUPTIONS

Participants and procedureEighty one students from a large Southwestern universityparticipated in exchange for extra credit. The procedure ofthis study is identical to study 1 with one exception. Insteadof participants being interrupted early or late during exposureto product statements, they were randomly assigned to beinterrupted either once (i.e., low frequency interruption) orthree times (i.e., high frequency interruption) during theevaluation task. As with early and late interruptions, lowand high frequency interruptions involved asking partici-pants to solve anagrams for two minutes. After examiningproduct statements, participants finished the study byanswering the same questions they did in study 1. Similarto study 1, all scales achieved adequate psychometric proper-ties (Appendix B), whereas all independent—measured andmanipulated—variables were not correlated with each other(all p-values n.s.).

ResultsWe manipulated interruption frequency and measured NFCCand perceived pleasantness of the interruption. A follow-uppath analysis removed the theoretically plausible alternativeof perceived pleasantness of the interruption as a mediatingvariable—indicating that it behaved as an independentvariable in this context. Results suggested enough varianceacross the full range of responses measuring the key indepen-dent variables and nonsignificant correlations among them;hence, a median split was performed on participants’ ana-gram experience ratings and NFCC scores. This resulted ina 2 (interruption frequency: low versus high) x 2 (perceivedpleasantness: more versus less) x 2 (NFCC: low versus high)between subjects ANOVA on product attitudes and evaluation.Finally, gender, age, and knowledge of product category wereused as covariates in the model but were not found to besignificant predictors of product attitudes and evaluation.

Product attitudesA three-way interaction between interruption frequency,perceived pleasantness of the interruption, and NFCC wasfound to be statistically significant (F(1,70) = 4.33, p< 0.05;see Figure 4); main effects and lower order interactions werenot significant. We assessed our hypotheses by performingplanned contrasts on perceived pleasantness conditions,separately. Specifically, we hypothesized in H3a that lowNFCC individuals will not be influenced by the frequencyof interruptions and expected evaluations to be flat acrossconditions. Findings are mixed and only partially supportour prediction. Supporting H3a, we find that low NFCC indi-viduals report no significant differences for low (versus high)frequency interruptions perceived to be more pleasant(MLowFreq = 4.63; MHighFreq = 4.09, t(70) = 1.18, n.s.). How-ever, low NFCC participants report significantly worseattitudes for frequent versus infrequent interruptions per-ceived to be less pleasant (MLowFreq = 4.61; MHighFreq = 3.46,t(70) = 1.86, p< 0.05).

Consumer response to interruption features and need for cognitive closure 67

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 9: Consumer response to interruption features and need for cognitive closure

For high NFCC individuals, high frequency interruptions(versus low frequency) worsen attitudes toward the productfor the more pleasant interruptions (MLowFreq = 5.50;MHighFreq = 4.34, t(70) =�2.30, p< 0.05), but not for lesspleasant ones (MLow_Freq = 4.27; MHigh_Freq = 4.89, t(70) = 0.91,n.s.; see Figure 4). Thus, these findings support H3b only wheninterruptions are perceived to be more pleasant.

Product evaluationA three-way interaction between interruption frequency,perceived pleasantness of the interruption, and NFCC wasfound to be marginally significant, F(1,70) = 2.82, p< 0.10(Figure 5); no other significant effects were found. Weassessed our hypotheses by performing planned contrastson perceived pleasantness conditions, separately. Plannedcontrasts reveal that for low NFCC individuals, frequent(versus infrequent) interruptions have similar impact onproduct evaluations for both perceived pleasant (MLowFreq =5.50; MHighFreq = 5.15, t(70) = 0.48, n.s.) and less pleasantexperiences (MLowFreq = 5.04; MHighFreq = 4.51, t(70) = 0.54,n.s.). Hence, H3a is fully supported.

We also find partial support for H3b; we predicted thatfor high NFCCs, an increase in interruption frequencywill determine lower product evaluations. Findings suggest

that product evaluations decreased with frequency for inter-ruptions perceived more pleasant only (MLowFreq = 6.36;MHighFreq = 5.03, t(70) = 1.67, p< 0.05); frequent (versusinfrequent) interruptions showed no significant differencein product evaluations for interruptions perceived lesspleasant (MLowFreq = 4.65; MHighFreq = 6.23, t(70) = 1.46, n.s.).

DISCUSSION

In study 1, high and low NFCCs’ reactions produced oppo-site product responses during perceived pleasant (early andlate) interruptions while reporting relatively flat evaluationsduring less pleasant interruptions. Results in study 2 suggestthat frequency interruptions affect high NFCC individualsduring interruptions perceived more pleasant and, surpris-ingly, low NFCC respondents during interruptions perceivedless pleasant.

Overall, interruption frequency has little negativeeffect, if any, on low NFCCs experiencing perceivedpleasant interruptions. A lack of urgency to force closureand nonexistent time constraints may allow low NFCCs tocontinue the search for diagnostic information. Irrelevant

(a) Pleasant Interruptions

(b) Unpleasant Interruptions

5.50

5.15

6.36

5.03

3.60

4.10

4.60

5.10

5.60

6.10

6.60

Low Frequency(n=25)

High Frequency(n=27)

Low NFCC

High NFCC

5.04

4.514.65

6.23

3.60

4.10

4.60

5.10

5.60

6.10

6.60

Low Frequency(n=14)

High Frequency(n=15)

Low NFCCs

High NFCCs

Figure 5. Impact of interruption frequency and need for cognitiveclosure on product evaluations for (a) pleasant interruptions and

(b) unpleasant interruptions.

(a) Pleasant Interruptions

(b) Unpleasant Interruptions

4.63

4.09

5.50

4.34

3.00

3.50

4.00

4.50

5.00

5.50

6.00

Low Frequency(n=25)

High Frequency(n=27)

Low NFCC

High NFCC

4.61

3.46

4.27

4.89

3.00

3.50

4.00

4.50

5.00

5.50

6.00

Low Frequency(n=14)

High Frequency(n=15)

Low NFCCs

High NFCCs

Figure 4. Impact of interruption frequency and need for cogni-tive closure on product attitudes for (a) pleasant interruptions

and (b) unpleasant interruptions.

68 M. Niculescu et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 10: Consumer response to interruption features and need for cognitive closure

non-task related interruptions may be discounted, andfrequent breaks during the evaluation task do not showsignificant effects on the evaluation task. In contrast, inter-ruption frequency has a strong negative effect on high NFCCsexperiencing perceived pleasant interruptions—as it increasesinterruption salience, postpones decision closure, and, in turn,distracts the high NFCC group when performing the evaluationprocess (Schiffman and Greist-Bousquet, 1992; Kruglanskiet al., 1993).

GENERAL DISCUSSION AND MANAGERIALIMPLICATIONS

Across two studies, we examined the impact of interruptionfeatures (i.e., timing, frequency, and perceived pleasantness)and a consumer individual difference variable (i.e., NFCC)on product evaluations and attitudes. In general, we findinterruption features have opposing effects on product evalu-ations and attitudes for consumers high versus low in NFCC.The effect is contingent upon the perceived pleasantness ofthe interruption. For example, those who have high NFCCrate products respond better when interruptions are perceivedto be more pleasant and occur: (i) late or (ii) infrequently. Incontrast, those who have low NFCC respond better to aproduct when interruptions are (i) perceived to be morepleasant and occur early or (ii) perceived to be less pleasantand occur infrequently.

In study 1, findings on interruption timing are consistentwith NFCC predictions when interruptions are perceived tobe more pleasant (Webster and Kruglanski, 1994; Kruglanskiand Webster, 1996); product ratings and attitudes toward theproduct worsen when interruptions disrupt closure—that is,early interruptions negatively affect product judgments ofhigh NFCC individuals, and late interruptions negativelyimpact product evaluations of low NFCC individuals.Similarly, in study 2, findings on interruption frequencysupport predictions for high and low NFCCs for perceivedpleasant interruptions but not for less pleasant ones. Theseresults suggest that the NFCC framework proposed in thispaper performs well in predicting and explaining outcomesof interruptions perceived more pleasant. However, itsexplanatory power is limited for interruptions perceivedless pleasant. In the latter case, findings may be relevantfor predictive purposes only.

Theory implicationsThese findings have interesting implications for theoryand practice. First, for theory, previous research hasmostly assumed interruption homogeneity in terms oftheir effect on consumer response. This research sug-gests that heterogeneity in features of interruptions mayresult in predictable outcomes. Theory development maydelineate both a taxonomy of interruption features and theirlikely interactions resulting in particular types of consumerattitudes and behavior (Xia and Sudharshan, 2002). Futureesearch may examine the interaction of other interruptionfeatures on consumer response. This would calibrate more

precisely the likely effect of multiple interruption featuresoperating simultaneously.

Second, in this paper, we opted for a more controlledenvironment to isolate the effects of interruptions andused anagrams to interrupt the main evaluation task.Although research has shown that anagram tasks are“laboratory equivalents to ‘real-life’ problem solving”(Ammons and Ammons, 1959), they may have somelimitations in sales contexts—anagrams are representativeto only a specific set of interruptions—that is, interrup-tions that do not match the main task. Our findings shouldbe extrapolated with caution to other situations wherestore employees interrupt customers with information rel-evant to the consumer evaluation process. Future researchshould expand the set of tools used to interrupt consumerprocesses and increase the external validity of our findings.For example, anagrams could be eplaced with more realisticinterruptions—such as phone calls, pop-up advertisementson computer displays, store announcements, or storeemployees—to determine the boundaries of interruptioneffects on consumer evaluative processes. New studiescould also use bigger sample sizes and respondents recruitedfrom other populations to increase the external validity ofthe results.

Third, research could investigate mediational processesto explain effects of perceived pleasantness of an interrup-tion on consumers. Extensive research has shown thatpositively and negatively valenced information is processedby distinct brain areas (Cacioppo and Berntson, 1994;Sutton et al., 1997; Cunningham, Raye, and Johnson,2004). Yet, it is unclear whether one or another determinesfaster processing times or more extreme evaluations(Cunningham and Johnson, 2007, p.232)—that is, in casethey are considered relevant enough to be integrated in theevaluation process.

Fourth, it is important to determine contexts under whichperceived pleasantness of interruptions might behave as aresult of, rather than independent of, other exogenousvariables—such as the locus of attribution of the affectiveresponses to interruptions (self versus others) (Heider, 1958),interruption difficulty or relevance (e.g., Speier, 1999), andperceived match between the interruption task and main task(e.g., Gillie and Broadbent, 1989). Although not found in ourresearch, this alternative role of perceived pleasantness isplausible from a theoretical standpoint as the interplay betweencognition and emotion can be complex and bidirectional(Power and Dalgleish, 2008).

Fifth, future studies could manipulate (instead of measure)perceived pleasantness and NFCC to increase contextualgeneralizability; similarly, other relevant contexts, suchas employee-related factors (e.g., job satisfaction or per-ceived self-efficacy) or organization-related factors (e.g.,role conflict or role ambiguity), could be examinedregarding how they shape perceptions of more or lesspleasant interruption experiences.

Finally, this paper investigated the effects of interruptionfeatures on cognitive consumer response. Nevertheless,research suggests that interruptions also produce emotional(e.g., job stress, Kirmeyer, 1988) and physiological responses,

Consumer response to interruption features and need for cognitive closure 69

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 11: Consumer response to interruption features and need for cognitive closure

such as an increase in blood pressure and heart rate, and adecrease in pulse amplitude (Ray et al., 1984). These aspectscould be analyzed in consumer contexts, as well.

Practice implicationsFor practice, our findings suggest that salespeople andmarketing managers may benefit from understanding theeffects of interruption timing and frequency on consumers’product evaluations. This could help overcome features ofthe environment (e.g., noise) or consumer that could unnec-essarily bias product evaluation. Given that a consumer islikely to be interrupted sometime during their shoppingexperience, knowing when to approach customers withnon-relevant information is essential (e.g., greetings, suchas “How are you doing today?” or “Do you find everythingok today?”). Customer approach could occur in retailcontexts, such as brick and click, and service contexts, suchas health care and car repair.

When a potential consumer evaluates a product and lookspressed for time (a proxy for high NFCC), this researchsuggests never interrupting them early in their experience,especially when the conversation is not relevant to theproduct being evaluated. If a salesperson must interruptconsumers who are pressed for time (e.g., company policyrequires engagement), then an attempt should be made tocreate pleasant interruptions late or infrequently in theshopping experience. Also, when it is hard to determinecustomer’s urgency to make a decision, it is best to keepinterruptions at a minimum. Although frequent interruptionsmay not be extremely harmful to individuals with lowNFCC, they may negatively impact product evaluations ofhigh NFCC customers—even when interruptions are per-ceived to be pleasant. In addition, we caution that otherfactors not considered may play a significant role (e.g.,what is the impact on the store’s reputation or the effectof customer’s mood), and hence, an extensive analysisof the context is advisable prior to implementing thecustomer engagement strategy.

Online contexts may also benefit from this research.Online advertising is rampant, yet the effects of advertisinginterruptions (e.g., banner, pop-up, and rich media ads) onconsumer webpage experience are poorly understood. Abusiness, for example, whose target market is primarilyrepresented by low NFCCs (e.g., photography enthusiasts)could place pleasing advertising interruptions at thebeginning of their webpage experience and worry less abouthow pleasant advertising interruptions are as a function ofwebpage viewing depth.

Finally, there is a critical need to understand better how todeliver necessary unpleasant interruptions in service contexts—such as health care and product repair. For example, thisresearch suggests it may be better to approach a high NFCCcustomer multiple times while they are waiting to get theircar fixed if it becomes apparent that the car will needexpensive repairs. In contrast, low NFCC customers maybe approached with an unpleasant interruption of costlyrepairs at any time, as long as the number of interruptionsis kept at a minimum.

APPENDIX APRODUCT STATEMENTS

(ADAPTED FROM GURHAN-CANLI, 2003)INSTRUCTIONS.In the following pages, you will be reading a series of statements abouta real HDTV (High Definition Television) set. Later, you will be askedsome questions about the information you have read. The informationprovided has been compiled on the basis of Consumer Reports 2010.

STATEMENTS:

1. The HDTV set is available in two standard sizes.2. Due to its state-of-the-art technology that prevents interference

and noise, it delivers high quality sound with sonic purity andclarity.

3. It comes with a well-functioning universal remote that can alsooperate other electronic equipment.

4. It has an aesthetically pleasing finish and an award-winningdesign.

5. The HDTV has average weight.6. It is highly energy efficient.7. It comes with a comprehensive user’s manual that is easy tounderstand.

8. Based on Consumer Reports’ 2010 annual questionnaire, theHDTV set was rated poor in reliability due to the number ofrepairs needed.

9. It is difficult to use because it does not have advancedautomatic features.

10. It does not have additional input/output jacks. Additional jacksprovide versatility by letting you connect other equipment.

11. It has a cord wrap and hooks.12. It does not have a power back up feature. Settings need to be

reset after a brief power outage.13. It does not have automatic volume correction. Volume

correction circuitry automatically lowers the volume of loudsounds and music.

APPENDIX BVARIABLES AND MEASURES

Product Evaluation (average of five items; α= 0.96, study 1;α= 0.98, study 2)After reading the previous statements, my overall evaluation ofthe described HDTV set is (on a 1–9 scale):1. Very negative to very positive2. Very unfavorable to very favorable3. Very bad to very good4. Very dissatisfactory to very satisfactory5. Very undesirable to very desirable

Attitudes Toward the Product (average of three items;α= 0.79, study 1; α= 0.86, study 2)Please evaluate your attitude toward the described HDTV set(on a 1 = strongly disagree to 7 = strongly agree scale):1. The product is likely to possess the stated characteristics.2. I react favorably to the product.3. I feel positively towards the product.

Product Knowledge (average of four items; α= 0.96, study 1;α= 0.95, study 2)Please rate your level of knowledge of TV sets using the scalebelow (on a 1 = strongly disagree to 7 = strongly agree scale,unless otherwise indicated):1. I consider myself knowledgeable on TV sets.2. My knowledge of TV sets helps me to understand very

technical information about this product.3. I use my knowledge on TV sets to verify that advertising

claims are in fact true.

(Continues)

70 M. Niculescu et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 12: Consumer response to interruption features and need for cognitive closure

BIOGRAPHICAL NOTES

Mihai Niculescu, is Assistant Professor of Marketing at NewMexico State University.

Collin R. Payne, is Associate Professor of Marketing at NewMexico State University.

Cuauhtémoc Luna-Nevarez, is Assistant Professor of Marketing atSacred Heart University.

REFERENCES

Ammons RB, Ammons CH. 1959. Rational evaluation of the‘standard anagram task’ as a laboratory analogue of ‘real-life’problem solving. Psychological Reports 5: 718–720.

Baron RS. 1986. Distraction-conflict theory: progress and problems.Advances in Experimental Social Psychology 19: 1–39.

Bettman JR, Luce MF, Payne JW. 1998. Constructive con-sumer choice processes. Journal of Consumer Research25(3): 187–217.

Bond CF, Titus LJ. 1983. Social facilitation: a meta-analysis of 241studies. Psychological Bulletin 94(2): 265–292.

Cacioppo JT, Berntson GG. 1994. Relationship between attitudesand evaluative space: A critical review, with emphasis on theseparability of positive and negative substrates. Psychologicalbulletin 115(3): 401–423.

Chan J, Jiang Z, Tan B. 2010. Understanding online interruption-based advertising: impacts of exposure timing, advertisingintent, and brand image. IEEE Transactions on EngineeringManagement 57(3): 365–379.

Choi JA, Koo M, Choi I, Auh S. 2008. Need for cognitiveclosure and information search strategy. Psychology & Mar-keting 25(11): 1027–1042.

Coraggio L. 1990. Deleterious effects of intermittent interruptionson the task performance of knowledge workers: a laboratoryinvestigation. Dissertation, University of Arizona.

Cronley ML, Posavac SS, Meyer T, Kardes FR, Kellaris JJ. 2005. Aselective hypothesis testing perspective on price-quality infer-ence and inference-based choice. Journal of Consumer Psychol-ogy 15(2): 159–169.

Cunningham WA, Raye CL, Johnson MK. 2004. Implicit andexplicit evaluation: fMRI correlates of valence, emotionalintensity, and control in the processing of attitudes. Journal ofCognitive Neuroscience 16(10): 1717–1729.

Cunningham WA, Johnson MK. 2007. Attitudes and evaluation:Toward a component process framework. In Harmon-Jones E,Winkielman P (eds). Social neuroscience: Integrating biologicaland psychological explanations of social behavior. The GuilfordPress: NY; 227–245.

Cunningham WA, Zelazo PD, Packer DJ, Van Bavel JJ. 2007. Theiterative reprocessing model: A multilevel framework for atti-tudes and evaluation. Social Cognition 25(5): 736–760.

Czerwinski M, Cuttrell E, Horvitz E. 2000. Instant messaging:effects of relevance and time. In Turner S, Turner P (eds). Peo-ple and Computers XIV: Proceedings of HCI 2000 (vol. 2).British Computer Society: London, UK; 71–76.

Dijksterhuis A. 2004. Think different: the merits of unconsciousthought in preference development and decision making. Jour-nal of Personality and Social Psychology 87(5): 586–598.

Dijksterhuis A, Aarts H. 2010. Goals, attention, and (un)conscious-ness. Annual Review of Psychology 61: 467–490.

Dijksterhuis A, Nordgren LF. 2006. A theory of unconsciousthought. Perspectives on Psychological Science 1(2): 95–109.

Dijksterhuis A, van Olden Z. 2006. On the benefits of thinkingunconsciously: unconscious thought can increase post-choicesatisfaction. Journal of Experimental Social Psychology 42:627–631.

Dijksterhuis A, Bos MW, Nordgren LF, van Baaren RB. 2006. Onmaking the right choice: the deliberation-without-attentioneffect. Science 311: 1005–1007.

Ein-Gar D, Shiv B, Tormala ZL. 2012. When blemishing leads toblossoming: the positive effect of negative information. Journalof Consumer Research 38(February): 846–859.

Gillie T, Broadbent D. 1989. What makes interruptions disruptive?A study of length, similarity, and complexity. PsychologicalResearch 50(4): 243–250.

Gonzalez-Vallejo C, Lassiter GD, Bellezza FS, Lindberg MJ. 2008.“Save angels perhaps”: a critical examination of unconsciousthought theory and the deliberation-without-attention effect.Review of General Psychology 12(3): 282–296.

Gurhan-Canli Z. 2003. The effect of expected variability of productquality and attribute uniqueness on family brand associations.Journal of Consumer Research 30(1): 105–114.

Han X. 2011. Omission neglect and the bias blind spot: effects ofthe self-other asymmetry in susceptibility to bias and responsive-ness to debiasing. Dissertation, University of Cincinnati.

Heider F. 1958. The Psychology of Interpersonal Relations.Lawrence Erlbaum: Hillsdale, New Jersey.

Herr PM, Kardes FR, Kim J. 1991. Effects of word-of-mouth andproduct-attribute information on persuasion: an accessibility-diagnosticity perspective. Journal of Consumer Research 17(March): 454–462.

Houghton DC, Grewal R. 2000. Please, let’s get an answer—anyanswer: Need for consumer cognitive closure. Psychology &Marketing 17(11): 911–934.

Kardes FR, Cronley ML, Kellaris JJ, Posavac SS. 2004. The role ofselective information processing in price-quality inference.Journal of Consumer Research 31(2): 368–374.

Kardes FR, Fennis BM, Hirt ER, Tormala ZL, Bullington B. 2007.The role of the need for cognitive closure in the effectiveness ofthe disrupt-then-reframe influence technique. Journal ofConsumer Research 34(October): 377–385.

Kirmeyer SL. 1988. Coping with competing demands: interruption andthe type a pattern. Journal of Applied Psychology 73(4): 621–629.

Kruglanski AW. 1990. Lay epistemics theory in social cognitivepsychology. Psychological Inquiry 1: 181–197.

Kruglanski AW, Freund T. 1983. The freezing and unfreezing oflay-inferences: effects on impressional primacy, ethnicstereotyping, and numerical anchoring. Journal of ExperimentalSocial Psychology 19(5): 448–468.

4. Overall, my knowledge of TV sets is: (1 = very poor to7 = very good)

Perceived Pleasantness of the Interruption (median split)Solving the anagram was a pleasant activity: (on a 1 = stronglydisagree to 9 = strongly agree scale)

Need for Cognitive Closure (median split of sum;α= 0.84, study 1; α= 0.89, study 2)Please indicate how much you agree or disagree with eachstatement below using the following scale (1 = strongly disagreeto 6 = strongly agree):1. I find that a well ordered life with regular hours suits my

temperament.2. I don’t like to be with people who are capable of

unexpected actions.

3. I find that establishing a consistent routine enables me toenjoy life more.

4. I enjoy having a clear and structured mode of life.5. I like to have a place for everything and everything in

its place.6. I dislike unpredictable situations.

(Appendix b continued)

Consumer response to interruption features and need for cognitive closure 71

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb

Page 13: Consumer response to interruption features and need for cognitive closure

Kruglanski AW, Mayseless O. 1988. Contextual effects in hypoth-esis testing: the role of competing alternatives and epistemicmotivations. Social Cognition 6: 1–21.

Kruglanski AW, Webster DM. 1996. Motivated closing of the mind:‘seizing’ and ‘freezing’. Psychological Review 103(2): 263–283.

Kruglanski AW,Webster DM, Klem A. 1993. Motivated resistanceand openness to persuasion in the presence or absence of priorinformation. Journal of Personality and Social Psychology65: 861–876.

Liu W. 2008. Focusing on desirability: the effect of decision inter-ruption and suspension of preferences. Journal of ConsumerResearch 35(December): 640–652.

Marsh RL, Hicks JL, Bink ML. 1998. Activation of completed,uncompleted, and partially completed intentions. Journal ofExperimental Psychology: Learning, Memory, and Cognition24(2): 350–361.

Meyer DL. 1991. Misinterpretation of interaction effects: a replyto Rosnow and Rosenthal. Psychological Bulletin 110(3):571–573.

Montgomery DC. 2000. Design and Analysis of Experiments(5th edn). John Wiley & Sons: New York, NY.

Nelson LD, Meyvis T. 2008. Interrupted consumption: disruptingadaptation to hedonic experiences. Journal of MarketingResearch 45(December): 654–664.

Payne JW. 1982. Contingent decision behavior. PsychologicalBulletin 92(2): 382–402.

Perlow LA. 1999. The time famine: Toward a sociology of worktime. Administrative Science Quarterly 44(1): 57–81.

Perlow L, Weeks J. 2002. Who’s helping whom? Layers of cultureand workplace behavior. Journal of Organizational Behavior23: 345–361.

Power M, Dalgleish T. 2008. Cognition and Emotion. From Orderto Disorder. Psychology Press: New York, NY.

Ray RL, Brady JV, Emurian HH. 1984. Cardiovascular effects ofnoise during complex task performance. International Journalof Psychophysiology 1(4): 335–340.

Rey A, Goldstein RM, Perruchet P. 2009. Does unconsciousthought improve complex decision making? PsychologicalResearch 73: 372–379.

Schiffman N, Greist-Bousquet S. 1992. The effect of task inter-ruption and closure on perceived duration. Bulletin of thePsychonomic Society 30(1): 9–11.

Shiota MN, Keltner D, Mossman A. 2007. The nature of awe:elicitors, appraisals, and effects on self-concept. Cognition andEmotion 21(5): 944–963.

Speier C. 1999. The effects of interruptions, task complexity, andinformation presentation on computer-supported decision-making performance. Decision Sciences 34(4): 771–797.

Speier C, Vessey I, Valacich JS. 2003. The effects of interruptions,task complexity, and information presentation on computer-supported decision-making performance. Decision Sciences 34(4):771–797.

Sutton SK, Davidson RJ, Donzella B, Irwin W, Dottl DA. 1997.Manipulating affective state using extended picture presenta-tions. Psychophysiology 34(20): 217–223.

Van Kleef GA, De Dreu CK, Manstead AS. 2004. The interpersonaleffects of emotions in negotiations: a motivated information pro-cessing approach. Journal of Personality and Social Psychology87(4): 510–528.

Van Kleef GA. 2009. How emotions regulate social life the emo-tions as social information (EASI) model. Current Directionsin Psychological Science 18(3): 184–188.

Waroquier L, Marchiori D, Klein O, Cleeremans A. 2010. Is itbetter to think unconsciously or to trust your first impression?A reassessment of unconscious thought theory. Social Psycho-logical and Personality Science 1(2): 111–118.

Webster DM. 1993. Motivated augmentation and reduction ofthe overattribution bias. Journal of Personality and SocialPsychology 67: 1049–1062.

Webster DM, Kruglanski AW. 1994. Individual differences in needfor cognitive closure. Journal of Personality and Social Psychology67: 261–271.

Xia L, Sudharshan D. 2002. Effects of interruptions on consumeronline decision processes. Journal of Consumer Psychology12(3): 265–280.

Zeigarnik BV. 1967. On finished and unfinished tasks. In Ellis WD(ed). A Sourcebook of Gestalt Psychology. Humanities Press:New York.

72 M. Niculescu et al.

Copyright © 2013 John Wiley & Sons, Ltd. J. Consumer Behav. 13: 60–72 (2014)

DOI: 10.1002/cb