Clinical Psychology Review 33 (2013) 1041–1056

Contents lists available at ScienceDirect

Clinical Psychology Review

Systematic information processing style and perseverative worry

Suzanne R Dash ⁎, Frances Meeten, Graham C L DaveyUniversity of Sussex, Brighton, UK

H I G H L I G H T S

• Similar psychological states and appraisals influence worry and systematic processing.• Increased sufficiency thresholds and systematic processing help to understand worry.• Systematic processing may be a transdiagnostic process across perseverative disorders.• Incorporating the HSM in models of worry provides new therapeutic opportunities.

⁎ Corresponding author at: School of Psychology, UniveE-mail address: s.dash@sussex.ac.uk (S.R. Dash).

0272-7358/$ – see front matter © 2013 Elsevier Ltd. All rihttp://dx.doi.org/10.1016/j.cpr.2013.08.007

a b s t r a c t

a r t i c l e i n f o

Article history:Received 19 April 2013Received in revised form 13 August 2013Accepted 19 August 2013Available online 27 August 2013

Keywords:WorryPerseverationSystematic processingSufficiency threshold

This review examines the theoretical rationale for conceiving of systematic information processing as a proximalmechanism for perseverative worry. Systematic processing is characterised by detailed, analytical thought aboutissue-relevant information, and in this way, is similar to the persistent, detailed processing of information thattypifies perseverative worry. We review the key features and determinants of systematic processing, and exam-ine the application of systematic processing to perseverative worry. We argue that systematic processing is amechanism involved in perseverative worry because (1) systematic processing is more likely to be deployedwhen individuals feel that they have not reached a satisfactory level of confidence in their judgement and thisis similar to the worrier's striving to feel adequately prepared, to have considered every possible negativeoutcome/detect all potential danger, and to be sure that they will successfully cope with perceived futureproblems; (2) systematic processing and worry are influenced by similar psychological cognitive states andappraisals; and (3) the functional neuroanatomy underlying systematic processing is located in the same brainregions that are activatedduringworrying. This proposedmechanism is derived from core psychological process-es and offers a number of clinical implications, including the identification of psychological states and appraisalsthat may benefit from therapeutic interventions for worry-based problems.

© 2013 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10422. Dual-process models of information processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10433. The heuristic–systematic model of information processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10434. Processing style summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10445. Application of the HSM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10446. Factors affecting the deployment of systematic processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10447. The sufficiency threshold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10448. Raised sufficiency thresholds in worriers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1045

8.1. Negative mood . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10458.2. Accountability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10478.3. Responsibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10478.4. Desire for control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10478.5. Need for cognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10478.6. Intolerance of uncertainty . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10488.7. Low problem-solving confidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1048

rsity of Sussex, Brighton BN1 9QH, UK. Tel.: +44 1273 876650.

ghts reserved.

1042 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

9. Motivation for systematic processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10499.1. Accuracy motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10499.2. Defence motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10499.3. Impression motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1050

10. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105011. Similarities between worry and systematic processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105012. Worry as a form of systematic processing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105113. Systematic processing and models of worry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1051

13.1. Cognitive models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105113.2. Emotion-based models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105113.3. Integrated affect and cognition models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1052

14. Clinical implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105215. Summary and concluding remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1053

Worry retards reaction and makes clear-cut decisions impossible —

Amelia Earhart

1. Introduction

Worry is defined as “a chain of thoughts and images, negativelyaffect-laden and relatively uncontrollable” (Borkovec, Robinson,Pruzinsky, & DePree, 1983, p. 10). These negative thoughts are aimedat anticipating threats (Mathews, 1990) and solving problems (Davey,1994). Worry thoughts can be catastrophic in nature (Davey & Levy,1998a), and include themes of personal inadequacy (Davey & Levy,1998b). Excessive, uncontrollableworry is the cardinal feature of gener-alized anxiety disorder (GAD) (American Psychiatric Association, 2000,DSM-IV-TR, 4th ed., text rev.), but perseverative worry also plays a rolein the anxious apprehension surrounding social and performance situa-tions seen in social anxiety, in the concerns about panic attack symp-toms in panic disorder, and in the fears about bodily symptoms inhypochondriasis (Purdon & Harrington, 2006). Furthermore, worryhas been implicated in checking and doubting subtypes of obsessive–compulsive disorder (OCD) (Tallis & DeSilva, 1992).Worry also predictstreatment outcomes in psychosis; worry significantly predicted perse-cutory delusions at three-month follow-up, with those scoring higheron measures of worry at baseline showing less symptom improvement(Startup, Freeman, & Garety, 2007). The pervasive nature of persevera-tiveworry across psychopathologies emphasises the need for an under-standing of the mechanisms of worry and effective strategies for theclinical management of worry-based presentations.

In the past 15 years, numerous variables have been shown toincreaseworry perseveration. However, little is known about themech-anisms accounting for the impact of these variables on worry persever-ation. One possiblemechanism is systematic processing, which is definedas “a comprehensive, analytic orientation1 in which perceivers accessand scrutinize all useful information in forming their judgments”(Chaiken, Liberman, & Eagly, 1989, p. 212) — a process that would un-doubtedly result in the worrier2 devoting extended periods of time toconsidering a chain of worry thoughts. This article reviews systematicinformation processing, a processing style that shares many formulistic

1 The use of the word ‘orientation’ suggests a general tendency to use this informationprocessing style.

2 The term ‘worrier’ is used in this paper to refer, predominantly, to individuals who ex-perience worry with a high frequency (i.e., they would score above average on the PennStateWorry Questionnaire developed byMeyer et al. (1990)). However, Ruscio, Borkovecand Ruscio (2001) demonstrated using taxometric analytic methods that normal andpathological worry are best conceptualised as occupying the extreme end of a single con-tinuum, rather than as discrete entities. Consequently, it is anticipated that systematic pro-cessing will play a role in anyone who is worrying, including in the bouts of worrydisplayed by low frequency worriers.

similarities with worry. Firstly, systematic processing is characterised bydetailed, analytical thought about issue-relevant information (Chaikenet al., 1989), and in this way, is similar to the persistent, detailed process-ing of information that characterises perseverative worry. Secondly, bothworry and systematic processing arise through subjectively unconscious/non-deliberative means (Borkovec et al., 1983; Chen & Chaiken, 1999).Thirdly, systematic processing is more likely to be deployed when indi-viduals feel that they have not reached a satisfactory level of confidencein their judgement (Chaiken et al., 1989), and this is similar to theworrier's striving to feel adequately prepared (Borkovec & Roemer,1995), to have considered every possible negative outcome/ to detect allpotential danger (Mathews, 1990), or to be sure that theywill successfullycope with perceived future problems (Davey, Hampton, Farrell, &Davidson, 1992). Fourthly, systematic processing and worry areinfluenced by similar psychological cognitive states and appraisals, in-cluding negative mood, responsibility, desire for control and accountabil-ity (Ambady & Gray, 2002; Bohner, Moskowitz, & Chaiken, 1995; Brainet al., 2008; Johnston & Davey, 1997; Maheswaran & Chaiken, 1991;Roemer & Borkovec, 1993; Startup & Davey, 2003; Tetlock, 1983). Lastly,there is evidence that the functional neuroanatomy underlying systemat-ic processing andworry is the same,with both associatedwith left frontallobe activation (Borkovec, Ray, & Stöber, 1998; Leynes, 2002; Leynes &Phillips, 2008; Nolde, Johnson, & Raye, 1998). This most likely reflectsthat systematic processing and worry are both verbal-based forms of an-alytical thought (Carter, Johnson, & Borkovec, 1986; Evans, 2008). Despitethere being a convincing case for considering the role that systematic pro-cessing plays in worry, it should be noted that systematic processing andworrying are not the same thing. Both represent forms of effortful, analyt-ic thought, but systematic processing is deployed in a broad range of taskswith personal and social significance (e.g., forming attitudes, Martin &Hewstone, 2003). Furthermore, worry is defined by cognitive and emo-tional experiences that are not uniformly encountered in other contextswhere systematic processing occurs. For more detail on how worry andsystematic processing differ, see the section headed ‘Worry as a form ofsystematic processing’). Given the structural similarities of these twophe-nomena, the purpose of this review is to explain the theoretical rationalebehind conceiving of systematic information processing having a role inworry as a proximal mechanism involved in the initiation and mainte-nance of perseverativeworry and its iterative style. Systematic processingdoes not negate the theoretical importance of the variables that have pre-viously been identified as worry promoters; rather it offers a theoreticalframework for understanding how these different variables leadworriersto engage in perseverative worry (see the Systematic processingand Models of Worry Cognitive models section for a description ofempirically-supported models of worry, and a discussion of how theHSM may fit into these models). The key determinants of systematicprocessing will be covered, alongside consideration of the application ofsystematic processing to perseverative worry and existing models ofworry-based pathology.

3 Some authors, e.g., Kruglanski, Thompson and Spiegel (1999), champion single-process models, arguing that forming judgements on the basis of heuristic or systematicprocessing reflects the same underlying process, an ‘if-then’ process. Chen and Chaiken(1999) disagree, arguing that systematic processing is not solely ‘if-then’ processing, andthat it involves integrating many, possibly contradictory, ‘if-then’ associations.

1043S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

2. Dual-process models of information processing

The 1970s and 1980s saw advances in dual-process models ofinformation processing, which were invoked to account for numerousresearch findings in the social cognition literature that appeared contra-dictory to one another (Petty &Wegener, 1999). For example, negativemood states often inhibited attitude change (e.g. Zanna, Kiesler, &Pilkonis, 1970, cited in Petty & Wegener, 1999), but on some occasionsthey facilitated attitude change (e.g. Leventhal, 1970, cited in Petty &Wegener, 1999). A number of different dual-process theories havebeen proposed (Chaiken & Trope, 1999; Evans, 2008) but they sharein common the assumption that human cognition relies on two modesof processing information: (1) a fast associative processing mode thathas low cognitive effort requirements, and (2) a slower mode basedon detailed processingwhich requires high levels of cognitive effort. De-spite this shared assumption, the dual-processmodels differ in regard totheir exact definitions of these two processing modes and the relation-ship between the two processing modes (Chaiken & Trope, 1999).

The two dominant dual-process models to emerge from the socialcognition literature are the elaboration likelihood model (ELM, Petty &Cacioppo, 1986) and the heuristic–systematic model (HSM; Chaikenet al., 1989). The ELM is based on a continuum of elaboration, which re-fers to the extent that a person will think about issue-relevant argu-ments. At one end of the ELM continuum, the individual will elaborateextensively and generate a lot of issue-relevant thoughts. At the otherend of the continuum, individuals will rely on peripheral cues to maketheir judgement, such as how many points an argument contains, andwill not elaborate on issue-relevant ideas. The ELM contains two pro-cessing routes: the central route (elaboration) and the peripheralroute (minimal elaboration). The central route requires time, effort,and ability, as recipients try to integrate the message arguments withtheir existing knowledge. Conversely, the peripheral route is utilisedwhen motivation is low, or there are constraints on time and ability.The peripheral route involves forming attitudes based upon peripheralcues such as the source of an argument, the number of arguments, orthe length of an argument. In a test of the ELM, Petty and Cacioppo(1984) manipulated argument strength and the number of argumentsthat were presented, as well as participant involvement. They observeda three-way interaction, with low involvement participants formingmore extreme attitudes following a greater number of arguments, re-gardless of the strength of the arguments, whereas the high involve-ment participants showed more extreme attitudes for the largenumber of strong arguments, but crucially, not for the large number ofweak arguments.

Around the time that the ELM was gaining attention, Chaiken andcolleagues published theoretical and empirical papers in support of an-other dual-process model, the heuristic–systematic model (HSM). Thismodel contained one processing route that was essentially identical tothe central route of the ELM.Within theHSM, this was called systematicprocessing. The second processing route was termed heuristic process-ing, and differed slightly from the peripheral route of the ELM in thatpeople were perceived to use heuristics, simple rules of thumb, inorder to assess the validity of messages. Examples of such heuristics in-clude ‘experts can be trusted’, and ‘length [of arguments] impliesstrength’.

The ELM and the HSM have similarities: (1) both include a processrequiring motivation and capacity, and a process that can occur whenthere are low levels of motivation and/or capacity, (2) both incorporatea ‘least-effort’ assumption, namely that humans will use the minimalcognitive resources that they can while still meeting their goal (seeThe sufficiency threshold section), and (3) both acknowledge that mo-tivational (e.g., task importance) and cognitive (e.g., existing knowl-edge) factors can influence processing (Chen & Chaiken, 1999). Thepredominant difference between the ELM and the HSM is the way inwhich the two processing routes are hypothesised to relate to one an-other. In the ELM, the central and peripheral routes are deemed

mutually exclusive. In the HSM, systematic and heuristic processingcan co-occur simultaneously, meaning that they can interact with oneanother, have additive effects, and bias one another.

Furthermore, the HSM allows for multiple motives (accuracy,defence and impression motives; see the Motivation for systematicprocessing section) within its multiple-motive framework, whereasthe ELM accounts solely for accuracy-motivated individuals (Chen &Chaiken, 1999). When considering how dual-process models of infor-mation processing may apply to the worry context, it is preferable toconsider multiple motives. This is because worriers can hold a numberof different goals for their worrying. To illustrate, it has been suggestedthat individuals engage in worry to (1) avoid emotionally-aversivestimuli (Borkovec & Inz, 1990), (2) detect danger (Mathews, 1990),(3) feel prepared for negative outcomes (Borkovec & Roemer, 1995),(4) copewith problems (Davey et al., 1992), and (5) do asmuch as pos-sible aboutwhat is worrying them (Davey, Startup,MacDonald, Jenkins,& Patterson, 2005).While Bohner, Erb, and Siebler (2008) argue that theELM provides a more complete account of the low-intensity processingmode, the HSM allows multiple-motives. Consequently, the HSM pro-vides a dual-process account3 of information processing that is bettersuited to accounting for how information is systematically processedby worriers, and thus the HSM has been chosen as the dual-processmodel of focus.

3. The heuristic–systematic model of information processing

Judgements, such asmaking a decision or forming an attitude, are in-tegral to human cognition. In relation to worry, such judgements mayinclude assessing whether a situation is threatening, generating ‘whatif?’ scenarios, and determiningwhether one is prepared for a potentiallynegative outcome. Social cognition models, such as the HSM, acknowl-edge that individuals are not passive recipients of information whenmaking such judgements, but rather they can process informationwith varying degrees of cognitive effort. As reported above, the HSM in-corporates two processing styles: systematic processing and heuristicprocessing. Systematic processing involves a detailed analysis of infor-mation relevant to an individual's judgement. Chen, Shechter, andChaiken (1996) define systematic processing as “comprehensive analy-sis of judgement-relevant information” (p. 262). Individuals who sys-tematically process pay attention to all relevant information (Chen &Chaiken, 1999), although which information is considered to be rele-vant will vary, and is affected by an individual's motivation and goals.Chaiken et al. (1989) conceptualise systematic processing as the high-end of the data processing continuum, and, as such, they state that it re-quires “more thanmarginal” (p. 212) amounts of cognitive capacity andeffort.

In different situations, people may focus on a subset of available in-formation which enables the use of simple decision rules or cognitiveheuristics to formulate their judgements and decisions. This is heuristicprocessing, and has been contrasted with systematic processing as “amore limited processing mode that demands much less cognitive effortand capacity” (Chaiken et al., 1989, p.213). Heuristic processing is onlyfeasible when heuristic cues are available to the individual. Heuristiccues are defined by Chaiken et al. (1989) as “any variable whosejudgemental impact is hypothesised to be mediated by a simple deci-sion rule” (p.216). One heuristic that has received a great deal of atten-tion is the availability heuristic (Tversky & Kahneman, 1973). In tenstudies, Tversky and Kahneman (1973) built a convincing case thatthe ease of retrieval of information influences estimates of probabilityand frequency. One such example was a study in which participants

1044 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

were asked to estimate whether a letter, e.g. ‘k’, more frequently oc-curred as the first or third letter in words in the English language.According to Tversky and Kahneman (1973), the letter ‘k’ is twice aslikely, in a standard piece of text, to be the third letter in a word as itis to be the first letter. However, Tversky and Kahneman (1973) founda significant bias for participants selecting that the letter wasmore like-ly to occur in the first position than the third position in aword. Tverskyand Kahneman (1973) argued that this is because it is cognitively easierto think of words that begin with ‘k’ than have ‘k’ as the third letter. Nu-merous other heuristics have been identified, including the warm glowheuristic (Monin, 2003), the audience response heuristic (Axsom, Yates,& Chaiken, 1987), and the expertise heuristic (Ratneshwar & Chaiken,1991). However, a list of heuristics does not in itself provide evidencethat heuristics require less cognitive effort than optimal processingstrategies. Consequently, Shah and Oppenheimer (2008) sought to de-fine the processes operating within heuristics that translate into less ef-fort consumption, and they argue in their effort-reduction frameworkthat “heuristics primarily serve the purpose of reducing the effort witha task” (p.207). Shah and Oppenheimer (2008) review evidencesupporting five ways that heuristics serve to reduce the cognitive effortthat a task requires: (1) examining fewer cues, (2) reducing the difficul-ty associated with retrieving and storing cue values, (3) simplifying theweighting principles for cues (i.e., an equal weight is assigned to eachcue, rather than weighing up the validity of each cue), (4) integratingless information, (5) examining fewer alternatives. A thorough exami-nation of heuristic definition and mechanisms of operation is beyondthe scope of this review on the role of systematic processing in worry,but interested readers are referred to a recent review by Gigerenzerand Gaissmaier (2011).

4. Processing style summary

To summarise, systematic processing leads to judgements formed onthe basis of careful examination of arguments, and includes the integra-tion of all useful evidence with information already held by the individ-ual. In contrast, heuristic processing is less effortful, and involves theassimilation of knowledge into existing knowledge structures, such asscripts and schemas. A similar processing style distinction was also ap-plied in Beevers' (2005) dual process model of depression, suggestingthat attention to different forms of processing may be important in un-derstanding a range of psychopathologies.

Apart from their distinctively different modes of operation, psycho-physiological studies also suggest that systematic and heuristic process-ing may be supported by functionally distinct brain processes, withsystematic and heuristic processing associated with activation of theleft and right frontal lobes, respectively (Leynes, 2002; Leynes &Phillips, 2008; Nolde et al., 1998). The processing styles are similar inthat neither is assumed to be wholly objective, and both can enable in-dividuals to achieve their processing goals.

5. Application of the HSM

The HSM was originally derived as a model of persuasion (Chaikenet al., 1989), but has since been applied to risk perception (Trumbo,1999) and decision-making, including decision-making in clinicalhealth contexts (Steginga & Occhipinti, 2004). Although a full reviewof the broader application of the HSM is beyond the scope of this reviewarticle, some examples are provided to demonstrate that the key princi-ples of the HSM apply beyond the persuasion setting. For example, thesufficiency threshold principle (see The sufficiency threshold section formore detail) has been extended to health/risk settings. Kahlor,Dunwoody, Griffin, Neuwirth, and Giese (2003) demonstrated that thenotion that a negative discrepancy between actual and desired confi-dence will promote systematic processing extends to health risk per-ception. They found that the wider the gap between an individual'sunderstanding of risk and the level of understanding that the individual

felt was needed tomake a decision about health risks, the more likely itwas that systematic processing would be deployed. Trumbo (2002) ap-plied the HSM to risk perception and found that the model had goodpredictive power for an individual's risk perception. Systematic process-ingwas found to consistently predict significantly higher risk estimates,whereas heuristic processing led to lower risk estimates. What is more,in a psychopathology context, the notion of heuristic and systematicprocessing has successfully been applied to an understanding of autismspectrum disorders (ASD). For example, De Martino, Harrison, Knafo,Bird, and Dolan (2008) found that individuals with ASD were signifi-cantly less influenced by message framing (whether monetary-basedoutcomes were presented as a loss or a gain), which is a heuristic cue,than individuals without ASD.

The application of the HSM to contexts beyond the persuasion set-ting raises the possibility that applying the HSM to understandingworry perseveration may also be fruitful. Indeed, systematic processinghas already been invoked in explanations of the role of negative mood(Davey, 2006b), and ‘as many as can’ stop rule deployment (Startup &Davey, 2001) in perseverative worry (see ‘Systematic processing andmodels of worry’ section for detail on ‘as many as can’ stop rules).There is evidence that highly anxious people use systematic, ratherthan heuristic processing (Hale, Lemieux, & Mongeau, 1995), whichsuggests that it may be of relevance in understanding other processes,such as worry, that are linked to the anxiety disorders (althoughworry and anxiety are separable constructs, see Davey et al. (1992)).Furthermore, a recent experiment has successfully applied the sufficien-cy threshold concept of the HSM to perseverative worry (Dash & Davey,2012).

6. Factors affecting the deployment of systematic processing

Systematic processing requires greater cognitive effort than heuris-tic processing, and consequently will only occur when an individual ismotivated to expend cognitive resources on processing information.As such, an individual's level of motivation is an important determinantofwhether systematic processing is deployed.What ismore, the explicitand implicit goals that an individual has for processing worry-relatedinformation (e.g., resolving a problem, preparing for a negative out-come, holding concurrent negative and positive beliefs about worry),will influence the point at which individuals feel they have processedthe information thoroughly enough. The next section focuses on thesufficiency threshold principle and motivation, the two key principlesthat determine whether systematic processing is deployed.

7. The sufficiency threshold

A central feature of the HSM is the sufficiency threshold principle,which recognizes the trade-off between cognitive efficiency and taskgoals. While efficiency is important, individuals exert whatever levelof effort is required to attain a sufficient degree of confidence thatthey have satisfactorily accomplished their processing goals (Chaikenet al., 1989). The least effort principle is an essential component of un-derstanding which information processing style individuals will use,and is the basis for the HSM's sufficiency threshold principle. Psycholo-gists such as Allport (1954) argued that individuals constrain their pro-cessing in order to preserve cognitive resources and make their waythrough an information-ladenworld. Related to the principle of least ef-fort, some theorists argue that humans are cognitivemisers, constrainedby limited cognitive capacity, and forced to take shortcuts where possi-ble to maintain efficiency (e.g. Fiske & Taylor, 1991). However, the useof the cognitive miser concept to explainwhy systematic processing re-quires high levels of motivation has been criticised for failing to accountfor how systematic processing can also occur in low-motivation settings(Maheswaran & Chaiken, 1991). Indeed, as highlighted by Hirsch andMathews (2012), among others, “it is unclear why excessive worryabout low probability events [i.e., low-motivation settings] persists

1045S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

when it causes frequentmental distress, with so little apparent benefit”(p. 636). Models of worry need to account for why large amounts oftime and effort are devoted to processing unlikely negative outcomes,situations which would normally be deemed low-motivation settings.The HSM's sufficiency threshold offers an explanation for effortful pro-cessing in low-motivation settings (Chaiken et al., 1989), which willnow be outlined.

The HSM provides a fuller account of the role of motivation in deter-mining processing style by employing the concept of a sufficiencythreshold. The sufficiency threshold principle extends the work ofSimon (1957) who introduced the idea of ‘satisficing’. Individualschoose the first option that addresses most of their needs, rather thanselecting the optimal solution — “finding a course of action that is‘good enough’” (Simon, 1957, p. 205). Individuals must balance theneed to minimise processing demands (maintain cognitive efficiency)with the desire to maximize confidence that they have achieved theirtask goals (e.g. anticipating negative outcomes). While an individualwill rarely be able to achieve complete confidence that they haveachieved their goals for processing information (e.g. anticipated all pos-sible negative outcomes), they can strive for a subjectively desired levelof confidence. Within the HSM, this desired amount of confidence istermed the sufficiency threshold, and it is conceptualised as fallingalong a continuum. Chaiken et al. (1989) state that the exact positioningof the sufficiency threshold (i.e., amount of confidence required) willvary depending on the individual and the situation. Likewise, anindividual's actual level of confidence will vary as a function of individ-ual and situational factors, and may change as new information is en-countered and processed. The discrepancy between actual confidenceand desired confidence is critical when determining which type of pro-cessing will be used: “according to the sufficiency principle, motivationfor systematic processing is a positive function of the negative discrep-ancy that exists between actual confidence and the sufficiency thresh-old” (Maheswaran & Chaiken, 1991, p. 14). Once an individual reachesa level of confidence that matches (or exceeds) their sufficiency thresh-old, cognitive effort should cease. However, while an individual's actualconfidence is below the sufficiency threshold, cognitive effort should beexpended in an attempt to reduce the discrepancy between desired-and actual confidence levels. Consequently, even in situations thatwould typically be considered low-motivation settings, worriers maybemotivated to expend large amounts of cognitive effort if their subjec-tive level of confidence is lower than their desired level of confidence.

Chen and Chaiken (1999) highlight predictions that can be drawnfrom the sufficiency threshold principle. Firstly, systematic processingis more probable as the discrepancy between actual and desired confi-dence increases. Consequently, situations where this gap is widened,such aswhen an individual's sufficiency threshold is increased, or actualconfidence levels are decreased, should result in greater deployment ofsystematic processing. Evidence for this comes from the finding thatsystematic processing is more likely when the task is more important(with the assumption that this is associated with an increase in theamount of confidence that an individual desires) as shown by Darkeet al. (1998) or when actual confidence is reduced, such as was shownby introducing information that contradicted what was indicated bythe heuristic cues (Maheswaran & Chaiken, 1991).

Factors that are understood to increase the deployment of systemat-ic processing, such as feeling accountable for one's decisions (e.g.,Tetlock, 1983), can be understood to operate by raising an individual'ssufficiency threshold and increasing the amount of confidence that theindividual feels is necessary. The sufficiency principle suggests that indi-viduals will engage in greater amounts of systematic processing whenheuristic processing (or small amounts of systematic processing) doesnot confer the desired level of confidence. This is because systematicprocessing is generally thought to provide greater confidence in judge-ments (Bohner et al., 1995; Chaiken et al., 1989; Chen, Duckworth, &Chaiken, 1999; Chen et al., 1996). Alter, Oppenheimer, Epley, and Eyre(2007) demonstrated that when an individual experiences difficulty

processing a message (e.g., the font is degraded or they are instructedto form a furrowed brow implying that they are processing a complexmessage), they show reduced reliance on heuristic cues. The authorsargue that metacognitive experiences of difficulty act as an alarm bellthat activates more analytic forms of processing (i.e., systematicprocessing). See the section on Wells' Metacognitive Model of worry(p. 43) for more information on metacognition. Additionally, a meta-analysis conducted by Kraus (1995) found three studieswhich support-ed the idea that attitudes based on direct experience show greaterattitude–behaviour consistency. Originally, this finding was interpretedby Fazio and colleagues as indicating that attitudes based on direct ex-perience are held with greater confidence (Kraus, 1995). In support ofthis, Pieters and Verplanken (1995) examined intention–behaviourconsistency in a longitudinal study of voting intentions and behaviour.They found that involvement with elections (measured by ‘thecoming elections are very important to me’ and (reverse-scored)‘I don't care who will win the elections in September’) significantly pos-itively predicted an individual's confidence in voting intentions(measured by ‘How confident are you that you actually will vote forthe party you just indicated?’). This suggests that individuals placegreater confidence in judgements made when they have moreinvolvement in the issues, which lends support to the notion thatindividuals will use systematic processing when they require greaterconfidence in their judgements.

8. Raised sufficiency thresholds in worriers

One of the key principles of the HSM is the sufficiency threshold(i.e., the discrepancy between actual and desired confidence in one'sjudgements). Many of the variables that have been implicated in raisedsufficiency thresholds in the social psychology literature (negativemoods, accountability, responsibility and desire for control) are also rel-evant to worriers (see Table 1.). We are now going to discuss the vari-ables that have emerged from the social psychology literature, theevidence that they influence sufficiency thresholds, and the evidencethat these variables play a role in worry processes.Wewill then explorethe ways in which variables that are relevant to perseverative worry(intolerance of uncertainty and low problem-solving confidence), buthave not yet been explicitly investigated in the context of the HSM,are also likely to raise sufficiency thresholds and thus lead to worriersdeploying systematic processing.

8.1. Negative mood

Negative mood, defined as “a dimension of subjective distress andunpleasurable engagement” (Watson, Clark, & Tellegen, 1988, p.1063),has been shown to increase systematic processing. Numerous studieshave demonstrated that negative mood states are associated with ex-tended cognitive processing (Ambady & Gray, 2002; Batra & Stayman,1990; Tiedens & Linton, 2001). For example, Bohner, Bless, Schwarz,and Strack (1988) informed participants that they had either performedwell or poorly on a test. Alongside this, participants were also providedwith information about how likely they were to perform well/poorlybased on the (fictional) percentage scores experienced by previous par-ticipants. Bohner et al. (1988) demonstrated that, after receiving nega-tive information about performance on a test, participants generatedmore causal reasons for the performance than individuals who receivedpositive feedback. This effect occurred regardless of how probable theoutcomewas. Bohner et al. (1988) argue that this study provides clarityon previously confounded studies containing negative events that wereless probable, and they suggest that increased causal reasoningwhen ina negative mood is not the result of a need to explain unexpectedevents. Thus, Bohner et al. (1988) suggested that the findings may bedue to a general systematic approach to information processing whenin a negative mood state and they hypothesised three possible reasonsfor why a negative mood state results in greater causal reasoning.

Table 1Evidence for variables that increase systematic processing (or decrease heuristic processing) and play a role in increased worry levels.

Variable Area of research Authors Key findings

Negative mood Systematicprocessing

Ambady and Gray (2002): Study 4 Sadmood participants were less accurate in their ratings (indicating deliberative processing) comparedto the neutral and sad mood + cognitive load conditions.

Bodenhausen, Sheppard, andKramer (1994): Study 1

Sad mood group were less reliant on stereotype cues (a heuristic cue) than the anger mood group.

Bohner et al. (1988)Participants in the failure condition felt worse than participants in the success condition, and those inthe failure condition generated more items on a thought listing task (indicating systematic processing)regardless of expectancy.

Tiedens and Linton (2001): Study 3 Disgust mood group participants were more likely to use stereotype information (heuristic cue) thanfear mood group participants.

Worry Johnston and Davey (1997) Negative bulletins increased sadness and anxiety, andwere associatedwithmoreworry catastrophisingsteps than positive or neutral bulletins.

Startup and Davey (2001): Study 1 Negative mood group showed significantly higher worry catastrophising than the positive or neutralmood groups.

Accountability Systematicprocessing

Livingston and Sinclair (2008) Participantswho felt accountablewere less likely to demonstrate prejudiced views (less use of heuristiccues) but only in low-prejudice individuals.

Tetlock (1983) Participants who were aware of their accountability before viewing the information were not affectedby the order of information presented, whereas those unaware that they were accountable prior toviewing the information demonstrated a primacy effect.

Worry Brain et al. (2008) Self-accountability (blamed self) was a significant predictor of worry 6 months after screening resultswere disclosed (Time 3).

Sweet and Pelton-Sweet (2008) Accountabilitywas associatedwith anxiety, e.g. “Please don't bemad if it'swrong”; “I'mnervous” (p.32).Responsibility Systematic

processingBohner et al. (1995)Uleman (1989)

Responsibility has been included in theoretical accounts of effortful, systematic processing, e.g. Bohneret al. (1995) and Uleman (1989), but empirical studies where responsibility has been measured/manipulated need to be conducted in future.

Worry Startup and Davey (2003) Experimentally increasing responsibility resulted in greater worry levels.Need for cognition Systematic

processingBatra and Stayman (1990) Positive mood reduced elaboration of thoughts but this was moderated by need for cognition, with the

effect stronger for those low in need for cognition.Cacioppo et al. (1983): Study 1 The effect of argument quality on message evaluation had a greater impact in high (vs. low) need for

cognition participants. High (vs. low) need for cognition participants reported expending morecognitive effort and had greater recall of the message arguments.

Worry Davey, Tallis, and Capuzzo (1996):Study 2

Whenusing an ‘asmany as can’ stop rule (‘Imust do asmuch as possible’), highworriers showedgreaterperseveration than low worriers and experienced greater negative mood.

Startup and Davey (2001): Study 2 Whenusing an ‘asmany as can’ stop rule (‘Imust do asmuch as possible’), highworriers showedgreaterperseveration than low worriers and experienced greater negative mood.

Desire for control Systematicprocessing

Swann, Stephenson, and Pittman(1981)

Individualswhounderwent control deprivation soughtmore information than thosewho felt in control.Information utility played a role, with thosewho believed that the information would be useful seekingmore than those who did not.

Worry Borkovec (1994) Five themes emerged following interviews with worriers about what function their worry served:“attempts to control the environment” (p.16) (‘superstitious avoidance of catastrophe’; ‘actualavoidance of catastrophe’) and “attempts to control oneself (emotions and/or behaviours)” (p. 16)(‘avoidance of deeper emotional topics’; ‘coping preparation’; ‘motivating device’).

1046 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

Firstly, enhanced causal reasoning may serve a protective function;searching for causal reasons makes it more likely that the individualwill identify external and self-irrelevant causes. Secondly, it may enableindividuals to avoid similar states in the future. Lastly, Bohner et al.(1988) suggest that causal thinking reduces the intensity of emotions,and can function as a mood control strategy. These theorised reasonsfor increased processing when in a negative mood bear a remarkablesimilarity to the theory that worry is deployed as a coping strategy foravoiding negative outcomes (Borkovec, 1994) and negative emotionalexperiences (Borkovec & Inz, 1990), as well as fitting with beliefs thatworriers hold, such as ‘worrying helps me to avoid problems in thefuture’ (Wells & Cartwright-Hatton, 2004).

Pathological worriers experience negative mood frequently(e.g. Meyer, Miller, Metzger, & Borkovec, 1990). The association be-tween negative mood and worry may emerge from the negative cogni-tive processes associated with chronic worrying such as negativeoutcome anticipation (Szabó & Lovibond, 2002), processing negativethoughts (Borkovec et al., 1998), catastrophising (Davey, 2006a), andpersonal inadequacy appraisals (Davey & Levy, 1998b). However, anumber of studies support the view that, rather than simply being aconsequence of worrying, negative mood is a causal contributor.Johnston and Davey (1997) found that inducing a negative moodresulted in significantly more catastrophising steps (each step is akin toa catastrophic worry thought) being generated during a catastrophisinginterview than positive or neutral moods. This and similar studies(e.g., Startup & Davey, 2001) demonstrate that negative mood hasa causal effect on worry perseveration, and is not simply an outcome of

worrying. The endemic negative mood experienced by high worriers islikely to trigger systematic processing.

There are a number of outcomes of experiencing persistent negativemoods that make it more likely that worriers will be motivated to usesystematic processing. Firstly negative mood states are associated withincreased performance standards (Cervone, Kopp, Schaumann, &Scott, 1994; Scott & Cervone, 2002). Negativemoods are also associatedwith an increased desire on the part of the individual experiencing thenegative mood to try to understand and repair their mood (Schwarz &Clore, 1983). Furthermore, the negativemoods experienced byworriers(e.g. sadness and anxiety) are related to uncertainty, and this can resultin systematic processing (Tiedens & Linton, 2001), see the sections onIntolerance of uncertainty and Cognitive models for more informationon the role that uncertainty plays in worry.

Dash and Davey (2012) examined the role that mood plays insufficiency thresholds in a worry context. Participants were inducedinto either a negative or neutral mood state and then indicated the ex-tent that they would feel accountable, responsible, want control, andhave a high need for cognition within a worry context. An individual'ssufficiency threshold (estimated by summing the scores for appraisalsof accountability, responsibility, desire for control, and need for cognition)indicates the likelihood that they will engage in systematic processing,with a higher sufficiency threshold indicating a greater likelihood. Indi-viduals induced into a negative mood state had a significantly higher suf-ficiency threshold than individuals induced into a neutralmood state, andthe sufficiency threshold was found to mediate the association betweennegative mood and self-reported worry.

1047S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

8.2. Accountability

Accountability is defined as the pressure to justify one's opinions toothers. It promotes effortful information processing in an attempt toavoid judgements that could cause embarrassment (e.g., Livingston &Sinclair, 2008). Experimental manipulations of accountability involveinforming the high accountability group that they will need to justifytheir ideas to others (e.g. Erb, Pierro, Mannetti, Spiegel, & Kruglanski,2007). Tetlock (1983) argued that if the participant is aware of theviews of the individual that they are accountable to, then they canadopt a heuristic-style approach and simply select the socially accept-able viewpoint. However, if the participant is unaware of the views ofthe person they are accountable to then they may wish to engage inmore effortful processing so that they can successfully justify theirviews. According to the accountability hypothesis proposed by Tetlock(1983), effortful processing should occur in the group of individualswho are justifying their views to an individual whose views are un-known, but not when they are justifying their ideas to an individual de-scribed as liberal or conservative. In the latter groups, Tetlock predictedthat a tactical attitudinal shift would occur, with the participant movingthe attitudes that they have to justify closer to that of the liberal or con-servative person. Tetlock carried out such an experiment; participantswere either told that they would have to justify their views to a groupwhose views were known (liberal or conservative) or unknown, orthey were not told that they would have to justify their ideas. The pat-tern of results obtained by Tetlock (1983) supported the hypothesesoutlined above, and suggests that accountability only motivates moreeffortful information processing when the heuristic of agreeing withthe other person's views is not available.

The notion of individuals striving to find socially-acceptableviewpoints fits with the existence of socially-motivated worries (seeImpression motivation, p.35), whereby worriers endorse items such as‘I worry about saying or doing the wrong thing among strangers’.What ismore, accountability has been shown to influenceworry; feelingsof accountability were found to predict the worry levels of women un-dergoing cancer screening (Brain et al., 2008). Consequently, feelings ofaccountability inworriers as identified by Brain et al. (2008)will lead toworriers having raised sufficiency thresholds.

8.3. Responsibility

Responsibility is closely related to task importance and, as such, isanother factor that is associated with systematic processing. Salkovskis,Rachman, Ladouceur and Freeston (1992, cited in Salkovskis, Richards,& Forrester, 1995) define responsibility as “the belief that one haspower which is pivotal to bring about or prevent subjectively crucialnegative outcomes” (p. 285). Manipulations of responsibility includeinforming participants that their responses will have a large influence(e.g., “[your responses] may influence the budget received by such stu-dents”) versus no influence (e.g., “your responses are of no real impor-tance beyond the purpose of this experiment”) (e.g., Startup & Davey,2003). Bohner et al. (1995) have suggested that responsibility increasessystematic processing by increasing task importance and raising suffi-ciency thresholds. Some authors have linked responsibility to Tetlock's(1983) conception of accountability. For example, individuals may en-gage in “responsible” processing when they are consciously concernedabout their reputation as a thinker (Uleman, 1989). Uleman (1989) sug-gests that systematic processing can form part of responsible process-ing, in which individuals are motivated to scrutinise information withgreater intentional control to ensure that it meets accountability stan-dards (for others and themselves). Worriers appear to hold worry inhigh regard, as a method of responsible processing, endorsing itemssuch as ‘worrying is a sign of a good person’ and ‘I would be selfish if Inever worried’ (Cartwright-Hatton & Wells, 1997). Interestingly, per-ceptions of responsibility are affected by mood states; individuals in-duced into a negative mood reported feeling more responsible than

participants induced into a neutral mood (Tiedens & Linton, 2001).Chaiken et al. (1989) reported that participants given sole responsibilityversus shared responsibility for evaluating messages were more moti-vated to engage in systematic processing.

Responsibility consistently emerges in factor analyses of worry con-tent, appraisals and beliefs (Cartwright-Hatton &Wells, 1997; Langlois,Freeston, & Ladouceur, 2000; Sugiura, 2007), and experimentally in-creasing responsibility increases worrying (Startup & Davey, 2003).What ismore,worriers report feeling responsible for negative outcomes(Davey et al., 1992), whichmay cause theworrier to systematically pro-cess any information that they perceive to be related to potential nega-tive outcomes.

8.4. Desire for control

Desire for control is themotive to control events in one's life and sys-tematic processingmay be deployed in an attempt to maximize control(Burger & Cooper, 1979). Maheswaran and Chaiken (1991) report thatsituational factors, such as control deprivation, and dispositional factors,like desire for control, increase the extensiveness ofmessage processing,and they postulate that this is due to raised sufficiency thresholds, withthe consequence thatmore systematic processing is required in order tomeet the sufficiency threshold. Kofta and Sedek (1998) suggest that aloss of control motivates an individual to deploy cognitive resources inan attempt to try to regain control, and that a loss of control may infact free up cognitive resources for deployment in effortful processing.

Desire for control is also related to worry; clinical descriptions high-light worriers' desire to control unwanted thoughts (e.g. Roemer &Borkovec, 1993;Wells, 1995). Borkovec (1994) explored the functional-ity of worry by asking worriers what purposes their worry had. Fivethemes emerged, which Borkovec (1994) categorised as representing“attempts to control the environment” (p.16) (‘superstitious avoidanceof catastrophe’; ‘actual avoidance of catastrophe’) and “attempts to con-trol oneself (emotions and/or behaviours)” (p. 16) (‘avoidance ofdeeper emotional topics’; ‘coping preparation’; ‘motivating device’). Ev-idently worriers strive to control the external world and their reactionsto it, and this high desire for control will result in raised sufficiencythresholds, making it more likely that systematic processing of worry-related information will be deployed.

8.5. Need for cognition

Need for cognition (NFC) is defined as “a need to structure relevantsituations in meaningful, integrated ways” (Cohen, Stotland, & Wolfe,1955, p. 291), and individuals who possess high levels of NFC are likelyto find systematic processing less aversive in terms of its cognitive ex-penditure. Theways inwhich an individualwishes to structure informa-tion to make it ‘meaningful’ and ‘integrated’ vary as a function ofindividual and situational factors (Cohen et al., 1955). Batra andStayman (1990) found that high NFC individuals demonstrated greaterelaboration of ideas relating to a message. Cacioppo, Petty, and Morris(1983) provided individuals with messages varying in argument quali-ty. Participants who were high NFC showed a greater impact of argu-ment quality on their post-message attitudes, indicating that they haddeployed systematic processing to a greater extent than their low NFCcounterparts. Cohen et al. (1955) suggest that situations in which anindividual is unable to achieve the amount of cognition that they desire,such as situations that are ambiguous, create frustration. What is more,Cohen et al. (1955) demonstrated that individuals who are high/medi-um NFC more readily identify situations as being ambiguous than low-NFC individuals. This raised perception of ambiguity is likely to increasean individual's desire to systematically process information.

Meyer et al. (1990) did not find evidence to suggest that worriershave a high dispositional level of NFC generally, but there are indicationsthat, when it comes to thinking about worry-relevant information, per-severative worriers prefer to think about the information and find

1048 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

meaning. For example, high worriers are more likely to endorse ‘asmany as can’ stop rules, which specify that a task should be completedfully and properly (e.g., Startup & Davey, 2001). For more informationon ‘as many as can’ stop rules, see Integrated affect and cognitionmodels section. Worriers also endorse beliefs that worry helps ‘to planthe future more effectively’ (Cartwright-Hatton & Wells, 1997), andprovides ‘the opportunity to analyse situations and work out the prosand cons’ (Davey, Tallis, & Capuzzo, 1996). The desire that worriershave to think through issues related to worries, in part because of thebelief that worrying is useful, will raise their sufficiency threshold andincrease the amount of effortful thinking that is required.

8.6. Intolerance of uncertainty

As noted in the introduction to this section, it is not just the variablesidentified in the social psychology literature (negative mood, account-ability, responsibility, desire for control, and need for cognition) thataffect an individual's sufficiency threshold; variables linked to persever-ative worry may also affect sufficiency thresholds. One such variable isintolerance of uncertainty (IU) (Ladouceur, Talbot, & Dugas, 1997),which was defined by Ladouceur, Blais, Freeston, and Dugas (1998) as“the way in which an individual perceives information in uncertainsituations and responds to this information with a set of cognitive,emotional and behavioral reactions” (p. 141). More recently, IUwas de-fined by Koerner and Dugas (2008) as “a dispositional characteristicthat arises from a set of negative beliefs about uncertainty and itsconnotations and consequences” (p. 631). IU appears to be underpinnedby two factors (for a review, see Birrell, Meares, Wilkinson, & Freeston,2011). These have been interpreted as representing prospective IU(anxiety about encountering uncertainty) and inhibitory IU (paralysisin the face of uncertainty) (Carleton, Norton, & Asmundson, 2007;McEvoy & Mahoney, 2011). Prospective IU appears to be associatedwith worry and compulsions, whereas inhibitory IU is more associatedwith social anxiety, panic and agoraphobia, and depression (McEvoy &Mahoney, 2011). This highlights that IU is a transdiagnostic construct,but focusing on specific aspects of IU may be fruitful in understandingspecific symptoms, such as worry.

Worriers frequently couch their worries in the form of ‘what if?’questions and because most worries are based on negative things thatmight happen in the future, it is difficult for the individual to knowwhether they will or won't occur. This ‘not knowing’ is experienced asan unpleasant state and gives rise to IU. A number of studies have dem-onstrated that IU correlates with trait worry4 (e.g., Buhr & Dugas, 2006).In Dugas, Gagnon, Ladouceur, and Freeston's (1998) cognitive model ofGAD, IU plays a central role, and is posited to drive ‘what if…?’questioning (see the Cognitive models section on the IU model ofworry). High IU individuals hold processing biases that result in the per-ception of ambiguous or uncertain events as threatening and intolerable(Butler & Mathews, 1983, 1987), despite the low probability of suchevents actually occurring (Ladouceur et al., 1997). Significantly, giventhese processing biases, high IU individuals also perceive negativeevents as more probable in uncertain situations than low IU individuals(MacLeod,Williams, & Bekerian, 1991). In addition, perhaps as a strate-gy for identifying possible negative outcomes, high IU individuals seekmore information before making decisions that they are confidentwith (Carleton, Sharpe, & Asmundson, 2007).

Chaiken et al. (1989) state that seeking out and integrating allrelevant information is characteristic of systematic processing, and asreviewed above, information-seeking is characteristic of high IU individ-uals. ‘Searching for more information’ has been used as an indicator ofsystematic processing in a number of studies (e.g. Griffin et al., 2008),which suggests that high IU is associatedwith the deployment of system-atic processing. Given that high IU individuals require more information

4 De Bruin et al. (2007) found that, while IU predicted traitworry, therewas little, if anyassociation between IU and idiosyncratic worry bouts.

than their low IU counterparts in order to be confident in their decisionmaking (Carleton, Sharpe, & Asmundson, 2007; Ladouceur et al., 1997;Tallis, Eysenck, &Mathews, 1991), high IU individualsmay have high suf-ficiency thresholds (i.e., lower confidence levels than they desire). Onestrategy that individuals may activate in an attempt to minimise this dis-crepancy is to seek more information, which may result in the use of‘what if?’ thinking in an attempt to consider all possible negative out-comes of a given ambiguous situation.

IU is also implicated in raised sufficiency thresholds through its asso-ciation with desire for control. There is evidence suggesting a link be-tween IU and desire for control, in the context of psychopathology.Freeston, Rhéaume, Letarte, Dugas, and Ladouceur (1994) emphasisethe role of control processes in their description of the manifestationof IU as “behavioural attempts to control the future and avoid uncertain-ty” (p.799). This is reflected in the dominant measure of IU, the Intoler-ance of Uncertainty Scale (Freeston et al., 1994), which includes itemsthat assess attempts to control future events. Sookman and Pinard(2002) demonstrated an association between IU in OCD and a needfor control. Furthermore, Moulding and Kyrios (2007) found that IUcould be predicted by an individual's control cognitions, and that desirefor control was correlated with the obsessive–compulsive sub-domainof IU and perfectionism. Evidence of an association between IU and astrong desire for control has also been found in the context of anorexianervosa (Sternheim, Konstantellou, Startup, & Schmidt, 2011). It hasbeen suggested that the association between a desire for control andIU might arise because of a form of IU described as a ‘need to knowwhat will happen’ (Niceli & Castelfranchi, 2001), and Mushtaq, Bland,and Schaefer (2011) suggest that “the perception and estimation of un-certainty might play a key role in monitoring processes and the evalua-tion of the ‘need for control’” (p.249). Thus, IU may raise sufficiencythresholds due to its effects on desire for control, and subsequently re-sult inmore systematic processing. Linked to this, worriers also have el-evated evidence requirements (Tallis et al., 1991), whichwill raise theirdesired level of confidence in their judgements. As such, oneway that IUmight be related to extended worry bouts is through raised sufficiencythresholds, which lead the individual to process worry-relevant infor-mation in a detailed manner.

8.7. Low problem-solving confidence

Another variable associated with perseverative worry that also hasimplications for an individual's sufficiency threshold is problem-solving confidence. Problem-solving confidence is defined as “anindividual's self-assurance in awide range of problem-solving activities,a belief and trust in one's problem-solving abilities (general problem-solving self-efficacy) and coping effectiveness” (Heppner & Lee, 2002,p. 347). Problem-solving is pivotal to the way in which some authorsdefine worrying. Davey (1994) describes worrying as “a constructiveand appropriate task-oriented process that contributes to the solvingof problems and the reduction of anxiety” (p.327). Problem-solvinghas been postulated as the adaptive form of worry (Borkovec et al.,1983; Mathews, 1990) and worriers endorse the belief that worryingis a useful process for solving problems. For example, worriers aremore likely to report believing that “worrying helps me to solve prob-lems” (Cartwright-Hatton & Wells, 1997) and “worrying gives me theopportunity to analyze situations and work out the pros and cons”(Davey, Tallis, & Capuzzo, 1996), than non-worriers. Davey et al.(1992) suggest that worry is deployed as a coping strategy aimed atsolving problems, but argue that it is when this adaptive, problem-solving worry is disrupted that pathological worry occurs. Davey et al.(1992) suggest twoways inwhich theproblem-solvingprocessmaybe-come disrupted in chronic worriers. Firstly, they suggest that worriersmay adopt an emotion-focused coping strategy, rather than copingstrategies aimed at dealing with the cause of the stress. This is reflectedin Robichaud and Dugas' (2005) assertion that worriers have a negativeproblem-orientation, which they define as “a set of dysfunctional

1049S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

attitudes toward social problem-solving” (p.391). Individuals who pos-sess negative problem-orientations are more likely to experience highlevels of worry, and emotional problem-orientation has been identifiedas a stronger predictor ofworry than cognitive or behavioural indicators(Dugas, Freeston, & Ladouceur, 1997). Secondly, Davey et al. (1992)suggest that worriers may have low problem-solving confidence (PSC),which prevents successful closure of the worry process.

Worriers may not feel confident that they have met the goals fortheir worrying (i.e., they may have a raised sufficiency threshold) dueto the relationship betweenworry and lowproblem-solving confidence.Within theHSM, Chaiken et al. (1989) emphasise thatwhen individualshave not met their sufficiency threshold, they are more likely to deploysystematic information processing. The amount of confidence that indi-viduals have in their judgement can be represented by a judgementalconfidence continuum. At one point on this continuum is the amountof confidence that an individual desires in the judgement that they aremaking, such as whether they have worried as much as possible(i.e., the sufficiency threshold). Another important point on this contin-uum is the amount of confidence that individuals currently have in theirjudgement (i.e., their actual amount of confidence). When there is anegative discrepancy between these two points, with actual confidencebeing lower than desired confidence, individuals are motivated to try tominimise this discrepancy. Worriers with low PSC are likely to have alarge discrepancy between their desired level of confidence and theiractual level of confidence. The sufficiency principle suggests that indi-vidualswill engage in greater amounts of systematic processing (i.e., de-tailed, integrative processing) when heuristic processing (or smallamounts of systematic processing) does not confer the desired level ofconfidence (Chaiken et al., 1989). This is because systematic processingis generally thought to provide greater confidence in our judgements(Bohner et al., 1995; Chaiken et al., 1989; Chen et al., 1996; Chenet al., 1999).

Petty, Briñol, and Tormala (2002) suggest that metacognitions –

individuals' thoughts and awareness of their own thoughts and thoughtprocesses, such as how confident they are in their thoughts – play animportant role in forming judgements. They found that when individ-uals were confident in the validity of their thoughts, valenced cognitiveresponses were more predictive of attitudes. When problem-solvingconfidence is low, individuals have greater feelings of uncertainty(Tormala, Rucher, & Seger, 2008). This uncertainty means that individ-uals with low problem-solving confidence are more likely to feel thatthey have not met their sufficiency threshold (i.e., their actual confi-dence in their judgement is lower than their desired confidence),mean-ing that low problem-solving confidence should result in a greaterdeployment of systematic processing. A number of studies have indicat-ed that when individuals feel uncertain, they are more likely to deploydetailed processing (e.g., Weary & Jacobson, 1997). This is probablybecause when individuals feel certain, they feel confident that theyhave sufficient knowledge, whereas when they feel uncertain, they arenot confident that they have sufficient knowledge, and so seekmore in-formation. In the case of the worrier, this may lead to reassurance seek-ing, searching for more information, and/or extended bouts of thoughtabout worry-relevant issues.

There are a number of problem-solving based reasons for anticipat-ing that worriers will deploy systematic processing. Firstly, worriers ex-hibit low problem-solving confidence (Belzer & Maydeu-O., 2002;Davey, 1994; Davey & Levy, 1998a; Davey et al., 1992), and this in-creases the likelihood that their judgementswill be deemed inadequate.Secondly, worriers frequently couch their worries in terms of personalinadequacy (Davey & Levy, 1998a, 1998b), which is likely to give riseto the belief that their judgements may not be accurate or effective.Lastly, worriers have been identified as having negative problem-orientation (Dugas et al., 1997; Robichaud & Dugas, 2005). A negativeproblem-orientation is defined as “perceived threat of problems towell-being, self-inefficacy or doubt over one's problem-solving ability,the tendency to be pessimistic about the outcome, and low frustration

tolerance” (Robichaud & Dugas, 2005, p. 391), and therefore is a charac-teristic that will reduce an individual's actual confidence.

9. Motivation for systematic processing

The sufficiency threshold is just one of the factors influencingwhether individuals deploy systematic information processing. The sec-ond key factor is the individual's level of motivation. Individuals areeconomy-minded, and as long as valid heuristic cues are present inthe environment, systematic processingwill only occurwhen the individ-ual is sufficiently motivated to justify the extra cognitive expenditure.A number of motivational factors have been identified (accuracy-,defence- and impression-motivation) and were incorporated into themultiple-motive framework of the HSM (Chaiken et al., 1989). Theseare typically linked to the goals that an individual has for processingthe information. It should be acknowledged that the HSM assumesthat both systematic and heuristic processing can enable an individualto satisfy their motivational goals (Chaiken et al., 1989), however, thisreview will be limited to occasions when systematic processing isdeployed. Individuals are likely to hold more than one motive, and thedescriptions below should be considered to refer to a primary motiva-tion, rather than a sole motivating force (Chen & Chaiken, 1999).

9.1. Accuracy motivation

Chaiken et al. (1989) originally formulated the HSM around the goalof assessing the validity of persuasive messages. They describe accuracymotivation as the “desire to form or to hold valid, accurate attitudes”(p.214). Inworriers, thismay relate to the assessment of all threat infor-mation when assessing whether a negative outcome could occur.

Characteristics of worriers increase the likelihood that they aremotivated to ensure that their appraisal/judgement of a situation isaccurate and valid (i.e., they possess high levels of accuracy motiva-tion). For example, worriers believe that worrying is a necessary pro-cess to engage in to avoid future threats and problems (e.g. Davey,Tallis, & Capuzzo, 1996). Furthermore, worriers believe that it is im-portant to think about the issues asmuch as possible whenworrying.One theoretical perspective that has shed light on worriers' motiva-tion to do as much as possible is the mood-as-input hypothesis(Davey, 2006b; Meeten & Davey, 2011) (Please see the Integrated af-fect and cognition models section for more detail on this model.), asevidenced by their endorsement of the use of the ‘as many as can’ stoprule for worry. The ‘as many as can’ stop rule is adopted by worrierswhen they engage in an open-endedworry bout; it is a goal thatworriershold, either implicitly or explicitly, to continue worrying until they havedone asmuch as possible to deal with or resolve theworry. For example,an individual whose goal is to feel prepared would continue worryinguntil they felt they had done all that they could to feel prepared. This de-ployment of ‘asmany as can’ goal-directed stop rules supports the notionthatworrierswill bemotivated to attain accurate judgements ofwhetheran issue has been fully resolved (Davey et al., 2005).

9.2. Defence motivation

Individuals can bemotivated by a “desire to formor to defendpartic-ular attitudinal positions” (p. 234, Chaiken et al., 1989). This is known asdefence motivation. Individuals who are motivated to defend particularattitudinal positions have the goal of confirming the validity of theseparticular attitudinal positions and disconfirming the validity of anyother attitudinal positions. Individuals who are defence-motivated andutilise systematic processing will do so selectively. Defence-motivatedsystematic processing involves focusing on attitude-relevant informa-tion that backs the attitudinal position that the individual wishes topreserve. Alternatively, it may include focusing on information that op-poses the unsupported attitudinal position, and lead to less favourableinterpretations of information that opposes the favoured position or

1050 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

supports alternative positions. Giner-Sorolila and Chaiken (1997) foundthat students relied on heuristic cues when their viewpoints were sup-ported by such cues, but used systematic processing when hostile heu-ristic cues were presented.

Defence motivation could come into play when worriers seek to de-fend the beliefs that they hold about worrying. Worry is an unpleasantexperience, and chronic worriers experience increasing distress as aworry bout progresses (Vasey & Borkovec, 1992). Cognitive dissonancetheory would predict that worriers would hold some positive beliefsabout why they worry in order to integrate the negative experience ofworrywith the compulsion toworry, and thusmaintain a sense of inter-nal consistency (Festinger, 1957). Indeed, a number of authors haveidentified that chronic worriers hold positive beliefs about deployingworry as a coping strategy, including beliefs such as ‘worrying keepsme in control’, ‘in order to get something done I have to worry aboutit’, ‘worrying gives me the opportunity to analyse situations and workout the pros and cons’, and ‘by worrying, I reorganise and plan mytime better’ (e.g. Borkovec, Hazlett-Stevens, & Diaz, 1999; Cartwright-Hatton & Wells, 1997; Davey, Tallis, & Capuzzo, 1996; Wells & Carter,1999). Thus, worriers may process information about their worries ina way that allows them to continue to hold these positive beliefsabout the utility of worry. Additionally, worriers hold a number ofstrong beliefs about the negative consequences of worry, such as ‘wor-rying increases my anxiety and so decreases my performance’, ‘worry-ing makes me irrational’, and ‘worrying distorts the problem I haveand so I am unable to solve it’ (Davey, Tallis, & Capuzzo, 1996). Alteringthe beliefs that worriers hold about worrying forms a central tenet ofmetacognitive therapy for GAD (e.g. Wells, 1999) (see the CognitiveModels section for more information on the metacognitive model ofworry). Thus worriers may process information about their worries ina way that allows them to maintain both positive and negative beliefsabout worrying.

9.3. Impression motivation

The third processing goal identified in Chaiken et al.'s (1989)multiple-motive framework is impression motivation. Impression-motivated individuals have a “desire to express attitudes that will besocially acceptable to potential evaluators, both real and imagined”(Chaiken et al., 1989, p. 234). Thus when individuals are impression-motivated their processing goal is to assess whether a particular attitu-dinal position is more socially-acceptable than alternatives. What isconsidered socially-acceptable will vary from situation to situation.The desire to hold a socially-acceptable attitude is typically heightenedwhen individuals believe that they may have to share it with others orjustify it to others. Furthermore, Chaiken et al. (1989) suggest that im-pression motivation is more likely when individuals are made awareof interpersonal relationships, or when the identities or groupmember-ship of significant others is emphasised. Chaiken et al. (1989) suggestthat high levels of impression-motivation can manifest in both system-atic and heuristic processing. If an individual's attitudinal position isknown, an individual can use a simple agreement heuristic, such as‘agreement facilitates liking’. But similarly, they may apply systematicprocessing to ensure that their ideas are socially acceptable.

Impression-motivated attitudes are strategic, and although they canbecome internalised, they more frequently reflect concessions to situa-tional pressures and usually revert to original positions when the situa-tional pressures are removed. When processing systematically toachieve the goal of assessing whether an attitudinal position is sociallyacceptable, the individual gives more extensive consideration to thedetail of the attitudinal messages, and whether they are socially-acceptable. Chen et al. (1996) found that when participants wereinstructed to have a pleasant interaction with another individual,impression-motivated individuals expressed attitudes that were similarto those of their partner, reflecting the use of the ‘go along to get along’

heuristic. It was also found that use of this heuristic biased subsequentsystematic processing in impression-motivated participants.

Impression motivation also appears to be a feature of chronicworriers. The Anxious Thoughts Inventory (Wells, 1994), which wasdesigned as a multidimensional measure of worry, includes a socialworry subscale. The social worry subscale consists of items such as ‘Iworry about saying or doing the wrong thing among strangers’ and ‘Iworry that people don't like me’. Individuals with GAD were found tobe comparable to individuals with social anxiety on the social worrysubscale, highlighting the importance of social worries in GAD (Wells& Carter, 2001). Furthermore, Purdon and Harrington (2006) highlightthe role of worry-type processes in social anxiety, which they claimhas anticipatory anxiety as a hallmark feature.

10. Summary

There is evidence to suggest that worriers have raised sufficiencythresholds as they (1) experience negative mood, (2) are intolerant ofuncertainty, (3) have a discrepancy between their desired confidenceand actual confidence in their judgements, (4) feel accountable and re-sponsible, (5) and desire control of, and have a high need for cognitionabout worry-relevant information. What is more, (6) worriers are moti-vated to ensure that judgements that they make about worries are accu-rate,fit with their understanding of theworld, and are socially acceptable.Consequently worriers have a high propensity to deploy systematic pro-cessing during worry bouts.

11. Similarities between worry and systematic processing

There are many similarities between systematic processing as an ef-fortful form of information processing and chronic worrying. For exam-ple as outlined above, systematic processing is associated with raisedsufficiency thresholds. Individuals will exert increasing levels of analyt-ical effort to attain a degree of confidence that they have achieved theirprocessing goals and appraisals relevant to responsibility and account-ability contributing to raising this ‘sufficiency threshold’. Similar pro-cesses appear to underlie perseverative worry. First, chronic worrierspossess personality characteristics that will raise the threshold for feel-ing confident that goals associated with judgements and decisions havebeen successfully met (e.g., perfectionism, IU, inflated concerns overmistakes) (Dugas et al., 1997; Frost, Marten, Lahart, & Rosenblate,1990; Pratt, Tallis, & Eysenck, 1997; Stöber & Joormann, 2001).Worriersalso exhibit cognitive appraisal processes known to facilitate systematicprocessing, such as feelings of responsibility and accountability for out-comes (Wells & Carter, 2001; Wells & Papageorgiou, 1998).

Secondly, metacognitive approaches to understanding pathologicalworry argue that chronic worriers possess metacognitive beliefs thatworrying is a necessary process to engage in to successfully avoid futurethreats (Borkovec et al., 1999; Davey, Tallis, & Capuzzo, 1996; Wells,1995), and such beliefs will inevitably raise sufficiency thresholds to en-sure that whatever level of effort is required is deployed to successfullyengage with the worry and achieve the necessary levels of confidencethat the worry has been dealt with.

In addition to these processing similarities, systematic processingand worrying appear to share similar functional brain characteristics.Systematic processing appears to be supported by a functionally distinctbrain process located in the left frontal lobes (Leynes, 2002; Leynes &Phillips, 2008; Nolde et al., 1998), and studies have reported evidencethat increases in worrying are associated with increased left hemi-sphere frontal activation (Borkovec et al., 1998). Furthermore, suc-cessful treatments for worry-related diagnoses such as GAD lead tosignificant reductions in left hemisphere activation (Hoehn-Saric,Schlund, & Wong, 2004). This is consistent with systematic processingand worrisome thought both being predominantly left hemisphere ac-tivities (Carter et al., 1986) involved in the systematic, verbal processingof information. What is more, high worriers have been shown to have

1051S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

restricted working memory capacity during worry (Hayes, Hirsch, &Mathews, 2008; Leigh & Hirsch, 2011) and this may reflect the factthat perseverative worriers are employing a more effortful cognitivestrategy than lowworriers, and thus have less available working mem-ory capacity.

12. Worry as a form of systematic processing

While systematic processing and worrying share many factors incommon, they are not the same thing. Both represent forms of effortful,analytic thought, but systematic processing is deployed in a broad rangeof judgement tasks having both personal and social significance. For ex-ample, systematic processing can be deployed in decision making, suchas treatment decisions in prostate cancer (Steginga & Occhipinti, 2004),and when forming attitudes (Martin & Hewstone, 2003). Systematicprocessing also plays a role when processing risk information (Kahloret al., 2003) and when deciding who to vote for (Forehand, Gastil, &Smith, 2004). Systematic processing can be an adaptive strategy toovercome biases in our cognitive processing and can provide greater ac-curacy when making judgements that can be misled by heuristics(Tversky & Kahneman, 1974).

Systematic processingmight be viewed as a form of analytic thoughtthat many individuals might call worrying, but more properly repre-sents constructive problem-solving in which activities such as logicalanalysis, problem-solving, information-seeking, and active behaviouralcoping all play a central role (Davey, 1994). It is onlywhen this adaptive,analytical process is ‘thwarted’ that the characteristics of pathologicalworrying begin to emerge (e.g., increasing emotional discomfort,catastrophising, perceived lack of control over theworry process). If sys-tematic processing is the effortful, analytic, verbally-based thought thatmany researchers have also used to define worrying (Borkovec & Inz,1990; Freeston, Dugas, & Ladouceur, 1996), then pathological worryingis additionally defined by those processes that ‘thwart’ the effectivenessof systematic processing. Such processes may involve low PSC (Davey,1994; Davey, Jubb, & Cameron, 1996), metacognitions about the nega-tive consequences of worrying (Wells, 1995), feelings of personal inad-equacy (Davey & Levy, 1998b), IU (Dugas, Buhr, & Ladouceur, 2004) andan avoidance coping style (see Berenbaum, 2010, for some further ex-amples of processes that may ‘thwart’ closure of the worry process).

Exploring systematic processing as a mechanism of perseverativeworry has the potential to bridge the gap between theories of worryat the level of clinical constructs, such as IU, and explanations of proxi-mal mechanisms that underlie worry perseveration. Indeed, becauseof its theoretical purity as a functionally-distinct effortful form of analyt-ical thought, systematic processing may be a truly transdiagnostic pro-cess that is involved in a range of perseverative, pathological thoughtprocesses, including rumination, worrying, brooding, doubting, andhyper-vigilant scanning to name but a few. A better understanding ofthe proximal mechanisms involved in worry perseveration offers thepromise of effective, theoretically-informed and targeted interventionsfor worry-based mental health problems.

13. Systematic processing and models of worry

In order to appreciate the role that conceiving of systematic informa-tion processing may have in our understanding of perseverative worry,it is important to consider current theoretical conceptualisations. Fiveprominent models of perseverative worry were identified in a reviewby Behar, DiMarco, Hekler, and Staples (2009): the cognitive avoidancemodel (Borkovec, 1994), the IU model (Dugas et al., 2004), themetacognitive model (Wells, 1999), the emotion dysregulation model(Mennin, Heimberg, Turk, & Fresco, 2002), and the acceptance-basedmodel (Roemer & Orsillo, 2005). Two additional models are themood-as-input model (Davey, 2006b), and Hirsch and Mathews'(2012) cognitive model of worry. These models can be broadly distin-guished into models with a primary cognitive focus (the IU model, the

metacognitive model, and the cognitive model of worry), models witha primary emotional/affective focus (the emotion dysregulation andthe acceptance basedmodels) and integrativemodels placing equal em-phasis on cognition and emotion (the avoidance model, and the mood-as-input model) (Behar et al., 2009).

13.1. Cognitive models

The cognitive models of worry appear to be consistent with the ideathat worriers hold raised sufficiency thresholds (greater discrepanciesbetween desired and actual confidence levels). The IU model of worry(Dugas et al., 1997; Dugas et al., 2004) proposes that individuals withhigh IU find uncertain or ambiguous situations very distressing, andthat this triggers ‘what if…?’ thinking, as well as a negative problem-orientation and heightened cognitive avoidance, which maintainsworry. Recent examinations of the IU construct have shown that IUhas two underlying latent components: prospective IU and inhibitoryIU (Carleton, Norton, & Asmundson, 2007; McEvoy & Mahoney, 2011).Prospective anxiety is anxiety in anticipation of uncertainty, whereasinhibitory anxiety is inaction in the face of uncertainty. Given thatsystematic processing requires considerable cognitive effort (i.e., it ischaracterised by action rather than inaction) it is interesting that pro-spective IU is a significant predictor of GAD, whereas inhibitory IU isnot (McEvoy & Mahoney, 2011).

The distinction between systematic and heuristic processing sharessimilarities with the top-down and bottom-up processes described byHirsch and Mathews (2012) in their cognitive model of worry. In thismodel, worry is posited to arise from an interaction between involun-tary bottom-up and voluntary top-down processes. Based on evidencethat worriers show biases in attention to threat cues and threat inter-pretation biases, Hirsch and Mathews incorporate involuntarybottom-up processes, which they describe as “pre-existing biases andhabitual thought patterns” (p. 638). Biases allow us to process informa-tion with less cognitive effort; if a feature consistent with a threat isidentified, the worrier will identify the situation as a threat, withoutanalysing and integrating all information about the situation. This isconsistent with Shah and Oppenheimer's (2008) definition of heuristicsas processes that “allow decisionmakers to process information in a lesseffortful manner than one would expect from an optimal decision rule”(p. 207). Hirsch and Mathews (2012) describe that role that top-downprocesses can have once intrusions are experienced: “Intrusions per-ceived as problems to be resolved tend to provoke efforts to deal withthem taking verbal form, as if searching for possible answers to ques-tions in the form of ‘What if…?’” (p. 639). This verbal, analytical think-ing style in response to intrusions shares similarities with Chaikenet al.'s (1989) definition of systematic processing as detailed, analyticalthought about issue-relevant information.

The metacognitive model of worry (Wells, 1995, 2006) acknowl-edges the importance of two kinds of worry. Type 1 worry occurs in re-sponse to anxiety-provoking stimuli and is associated with positivebeliefs aboutworry, such as believing thatworrywill help the individualto cope. Type 2worry is activated during Type 1worry, and is associatedwith negative beliefs about the danger and uncontrollability of worry.Type 2 worry is then suggested to result in a range of unhelpful strate-gies deployed in an attempt to control worry, including reassuranceseeking, checking and thought suppression.While it is harder to accom-modate negative beliefs within the systematic information processingmodel, it may be that the associated distress causes a heightenednegative mood. As reported above, Bohner et al. (1988) suggest thatcausal thinking reduces the intensity of emotions, and can function asa mood control strategy.

13.2. Emotion-based models

The utility of detailed, analytical thinking as a vehicle for regulatingemotions (Bohner et al., 1988) also suggests a way that systematic

1052 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

processing may fit with the emotional theories of worry, i.e., it may bedeployed by individuals in an attempt to understand, regulate or repairthe endemic negative mood experienced by worriers. The emotiondysregulation model of worry (Mennin, Heimberg, Turk, & Fresco,2005) incorporates the idea that worry as a form of cognitive avoidanceis driven by an individual's response to, and regulation of, emotional ex-perience. Individuals with GAD experiencemore intense negative emo-tions, poorer understanding of emotions, greater negative reactivity toemotional experiences, and are less effective at self-soothing (Menninet al., 2005; Salters-Pedneault, Roemer, Tull, Rucker, & Mennin, 2006).It is hypothesised that the strategies used by individuals with emotiondysregulation may actually lead to them feeling worse — and this in-creased negative mood may drive systematic processing (e.g., Ambady& Gray, 2002; Batra & Stayman, 1990; Tiedens & Linton, 2001). Theacceptance-based model of worry (Roemer & Orsillo, 2005) has fourcomponents including aversive negative reactions to internal experi-ences, problematic fusion between internal experiences, avoidanceand behavioural restriction. Like the emotional dysregulation model,this theory is in its infancy and greater research, including longitudinaldesigns and experimental manipulation of components of the model,is required to be confident in the relationship between the componentsof the proposed acceptance-based model and perseverative worry(Behar et al., 2009). Nonetheless, this model also highlights the impor-tance of negative emotional states in worry, and such states have beenshown to raise sufficiency thresholds in an experimental study (Dash& Davey, 2012).

13.3. Integrated affect and cognition models

There are also models that have integrated cognitive and affectiveaspects. The avoidance model of worry (Borkovec, 1994) is based onthe idea that worry is a verbal linguistic activity that individuals engagein to avoid processing more distressing images and the somatic andemotional activation that would accompany experiencing such images.Worry is then negatively reinforced because the avoidance of unpleas-ant imagery leads to positive beliefs in the utility of worry. The linguisticnature of systematic processing may make it an appealing cognitiveavoidance activity for the worrier. Indeed, the neurobiological overlapseen in studies that have identified left frontal lobe activation in worryand systematic processing (Borkovec et al., 1998; Leynes, 2002;Leynes & Phillips, 2008; Nolde et al., 1998) most likely reflects thatboth are verbal-based forms of analytical thought (Carter et al., 1986;Evans, 2008). Verkuil, Brosschot, Borkovec, and Thayer (2009) mea-sured heart rate and heart rate variability during three conditions:worry, problem-solving about a personally irrelevant issue, and relaxa-tion. Verkuil et al. (2009) found that heart ratewashigher andheart ratevariability was lower during the worry and problem-solving conditionsthan the relaxation condition. This difference was not accounted for bythe participants' emotional responses (assessed using visual analoguescales). Consequently, Verkuil et al. (2009) claim that “mere mentalload is responsible for – at least a part of – the physiological effects ofworry” (p. 439). This is consistent with the notion that worriers areusing effortful information processing but does not necessarily supportthe notion that worriers do this in order to avoid more negative emo-tional experiences.

Worry is typically open-ended; there is no definite end-point. Con-sequently, an individual mustmake a decision aboutwhether to contin-ue with their worry bout, or terminate it. The mood-as-input model ofworry (Davey, 2006b; Meeten & Davey, 2011) suggests that individualsuse theirmood to decide if they havemet theirworry goals According tothe mood-as-input theory, mood provides information about whetherthe stop rules have been met (Davey, 2006b). Stop rules are criteriafor deciding when to stop a task, and as such, are conceptualised as de-cision rules (e.g., Vaughn, Malik, Schwartz, Petkova, & Trudeau, 2006).For example, if an individual is worrying, they may apply either an ‘asmany as can’ (AMA) stop rule, whereby they believe that they must

do as much worrying as possible, or a ‘feel like continuing’ (FLC) stoprule, which involves stopping when the worry process is no longersomething that they feel like doing (Davey et al., 2005). If individualsare in a negativemood and using an AMA rule, their negativemood sug-gests that the task has not been completed asmuch as possible (positivemood is usually experiencedwhen a task is completed), and individualsare likely to persevere at the task. The mood-as-input model is consis-tent with the idea that worriers have goals which they wish to be con-fident that they have met (i.e., whether they have done as much asthey can), and is similar to the emotion-based models in that it high-lights the importance of negative affect (which can trigger systematicprocessing).

The existingmodels have provided uswith a better understanding ofthe cognitive and affective components ofworry, and theway that affectand cognition might interact to bring about, or at least maintain, boutsof worry. However, it is not always clear from these models exactlyhow the constructs identified in thesemodels (e.g., intolerance of uncer-tainty, positive beliefs about worry) lead to an individual generating achain of worry thoughts. One of us (GCLD) has argued before that “thedetails of how such causal relationships are mediated will be found inproximal explanations of the activity of worrying, which determine theonset, duration and sequential properties of worry bouts” (Davey,2006b, p. 218). Such proximal explanations need not compete withexistingmodels of worry, rather they provide an explanation at a differ-ent level of elaborative detail. Conceiving a role for systematic process-ing in perseverative worry offers a way of unifying these differentmodels through a couple of mechanisms. First, the sufficiency thresholdprinciple holds that the greater the negative discrepancy between de-sired confidence and actual confidence (i.e., when desired confidenceis lower than actual confidence), the more likely it is that an individualwill deploy effortful chains of thought in an attempt to minimise thisdiscrepancy. Consequently, factors in the models of worry that accountfor either increased desired confidence and/or lowered actual confi-dence can be understood to increase the discrepancy. Factors from themodels reviewed above that would be expected to raise the desiredamount of confidence include intolerance of uncertainty surroundingnegative outcomes (Dugas et al., 2004), holding positive metacognitivebeliefs about the usefulness ofworry formanagingproblems,while con-currently holding beliefs that worry is dangerous (Wells, 1999), andstriving to do as much as possible in the form of the deployment of ‘asmany as can’ stop rules for worrying (Davey, 2006b). Secondly, manyof these models of worry emphasise the role of negative mood as eithera direct or indirect driving process for perseverative worrying, and it iswell known that negative mood is a trigger for perseverative, effortfulthought in the form of systematic information processing (e.g., Ambady& Gray, 2002; Batra & Stayman, 1990; Tiedens & Linton, 2001). Inaddition, negative mood facilitates the endorsement of cognitive ap-praisals known to increase systematic as opposed to heuristic pro-cessing (Dash & Davey, 2012). As such, the activation of systematicprocessing by negative mood states offers a mechanism to explainhow negativemoodmay activate worryingwithin a number of theoriesof pathological worrying.

Apart from these direct links to theories of worrying, systematicinformation processing is a form of decision-making that is wellresearched, offers a proximal explanation for the way in which worry-ing becomes a perseverative and effortful activity, and posits some set-ting conditions (e.g. raised sufficiency thresholds, negative mood) thattrigger worrying. It spans many of the constructs found in contempo-rary models of worrying, including intolerance of uncertainty, perfec-tionism, low problem-solving confidence, positive beliefs about theutility of worrying, and attempts to avoid negative mood states.

14. Clinical implications

As outlined in this review, the heuristic–systematic model (HSM)has important implications for our theoretical views of perseverative

1053S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

worrying, and we introduce some of these implications for future re-search here. Firstly, the HSM does not specify what combination of fac-tors that promote or thwart systematic processing manifests in thedeployment of systematic processing. Future work should aim to deter-mine whether the raised sufficiency thresholds of worriers result in thedeployment of systematic processing. Thiswill require thedevelopmentof procedures that provide a more direct measurement of systematicprocessing (c.f. Dash & Davey, 2012) and can be applied to the worrycontext. Secondly, while this review article has focused on systematicprocessing, the HSM also includes heuristic processing, and the heuris-tics that are relevant to the worry process should be examined. Suchheuristic cues could include selective engagement with threat cues, asan example of threat processing biases that worriers possess (Hirschet al., 2011), but could also include behaviours such as reassurance seek-ing (Rector, Kamkar, Cassin, Ayearst, & Laposa, 2011), which fits wellwith the notion of the consensus heuristic (O'Keefe, 2002). It is likelythat heuristic processing of threat cues acts to bias the individual's sub-sequent systematic processing, as shown by Chaiken and Maheswaran(1994). Lastly, sufficiency thresholds should be examined in clinicalpopulations, with the aim of guiding therapeutic interventions basedon a greater understanding of the variables that drive prolongedworry bouts.

When considering the therapeutic implications of conceiving of sys-tematic processing as a processing style that occurs in worrying, it isfirst necessary to acknowledge that there will be conditions in whichthis is a useful, adaptive processing style. Systematic processing ismore resistant to the sorts of biases and errors that can occur when re-lying upon heuristic processing (e.g., Tversky&Kahneman, 1974). How-ever, while avoiding errors in thinking is desirable when makingimportant decisions, such as which house to buy, it is less criticalwhen thinking about less important issues, such as which sandwich tobuy for lunch. High worriers report more worries about minor thingsthan low worriers (Roemer, Molina, & Borkovec, 1997), suggestingthat they apply detailed thought to issues that others do not deem nec-essary. One possible useful application of the HSM to theGAD therapeu-tic context is that socialisation to two processing modes, an effortfulmode and a less-effortful mode, allows therapists to engage in a dia-logue with clients about when effortful processing is warranted andwhen it is not.

The application of theHSM toperseverativeworrying highlights thatattention should be paid to the appraisals that worriers hold about theimportance of their task, such as feeling accountable (e.g., Brain et al.,2008) and desiring control (e.g., Freeston et al., 1994), as these ap-praisals raise sufficiency thresholds (e.g., Chaiken et al., 1989), increas-ing the likelihood that an individual will feel the need to considerworry thoughts in detail. Thus, conceiving of systematic processing asa mechanism involved in perseverative worry provides a theoreticalframework for the use of cognitive-behavioural therapy (Covin,Ouimet, Seeds, & Dozois, 2008) and metacognitive therapy for GAD(Wells & King, 2006).

Furthermore, Hirsch andMathews (2012) suggest that the existenceof top-down and bottom-up processes involved in worry perseverationindicate that the most effective treatments for worry are likely to bethose that tackle both types of processes. Treatments which focus oncognitive bias modification (Hakamata et al., 2010) may alter bottom-up/heuristic-type processing, while those that focus on, for example,modifying metacognitive beliefs about the usefulness/danger of worry(Wells & King, 2006) may alter an individual's sufficiency threshold,and subsequent likelihood of deploying systematic processing.

Finally, given that negative mood has been shown to increase suffi-ciency thresholds, and that this was associated with higher levels ofworry (Dash & Davey, 2012), there is a need for treatments that addressthe role of lowmood and the treatment of comorbid depression in highworriers. Therapeutic techniques such as behavioural activation mayserve to improve mood (Dobson et al., 2008) and has recently been ap-plied to the treatment of worry in a pilot trial (Chen, Liu, Rapee, & Pillay,

2013). Improvedmood, as shown by reduced depression scores in Chenet al.'s (2013) study, may prevent worriers from entering into a persev-erative worry bout.

15. Summary and concluding remarks

Worry can be conceived as a multifaceted phenomenon, consistingof identifying problems, threat anticipation, and chains of catastrophic,negative thinking (Borkovec et al., 1983; Davey, 1993, Mathews,1990). The HSM provides a theoretical basis for understanding how in-dividuals process information. This model has been successfully appliedto persuasion, risk perception and decision-making, and additionallyhas relevance to clinical psychopathology. Key dispositional factorsexhibited by worriers – negative mood, IU, and low PSC – are all likelyto widen the gap between individuals' actual level of confidence andtheir desired level of confidence (the sufficiency threshold), and theHSM suggests that in order to meet their sufficiency thresholds, wor-riers will deploy systematic processing. It is an exciting prospect forboth psychopathology research and clinical practice that raised suffi-ciency thresholds and increased systematic processing may provide afuller understanding of the proximal mechanisms underlying persever-ative worry— an understanding that is derived from core psychologicalknowledge which will help to explain why worry is such a universalphenomenon.

Acknowledgements

We thank the two anonymous reviewers for their helpful com-ments on an earlier version of this manuscript. This article was writtenwhile GCLD and FM were funded by ESRC grant RES-062-23-2336.SD was funded by a University of Sussex Graduate Teaching AssistantStudentship.

References

Allport, G. W. (1954). The Nature of Prejudice. Cambridge: Addison-Wesley.Alter, A. L., Oppenheimer, D.M., Epley, N., & Eyre, R. N. (2007). Overcoming intuition:

Metacognitive difficulty activates analytic reasoning. Journal of ExperimentalPsychology. General, 136(4), 569–576. http://dx.doi.org/10.1037/0096-3445.136.4.569.

Ambady, N., & Gray, H. M. (2002). On being sad and mistaken: Mood effects on the accu-racy of thin-slice judgments. Journal of Personality and Social Psychology, 83(4),947–961. http://dx.doi.org/10.1037/0022-3514.83.4.947.

American Psychiatric Association (2000). Diagnostic and Statistical Manual of Mental Dis-orders (4th ed., text rev ed.)Washington, DC: Author.

Axsom, D., Yates, S., & Chaiken, S. (1987). Audience response as a heuristic cue in persua-sion. Journal of Personality and Social Psychology, 53, 30–40.

Batra, R., & Stayman, D.M. (1990). The role of mood in advertising effectiveness. TheJournal of Consumer Research, 17(2), 203–214.

Beevers, C. G. (2005). Cognitive vulnerability to depression: A dual process model. ClinicalPsychology Review, 25, 975–1002.

Behar, E., DiMarco, I. D., Hekler, E. B., & Staples, A.M. (2009). Current theoretical models ofgeneralized anxiety disorder (GAD): Conceptual review and treatment implications.Journal of Anxiety Disorders, 23, 1011–1023.

Belzer, K. D., D'Zurilla, T. J., & Maydeu-O., A. (2002). Social problem solving and trait anx-iety as predictors of worry in a college student population. Personality and IndividualDifferences, 33(4), 573–585. http://dx.doi.org/10.1016/S0191-8869(01)00173-8.

Berenbaum, H. (2010). An initiation–termination two-phase model of worrying. ClinicalPsychology Review. http://dx.doi.org/10.1016/j.cpr.2010.06.011.

Birrell, J., Meares, K., Wilkinson, A., & Freeston, M. H. (2011). Toward a definition of intol-erance of uncertainty: A review of factor analytical studies of the Intolerance ofUncertainty Scale. Clinical Psychology Review, 31(7), 1198–1208. http://dx.doi.org/10.1016/j.cpr.2011.07.009.

Bodenhausen, G. V., Sheppard, L. A., & Kramer, G. P. (1994). Negative affect and socialjudgement: The differential impact of anger and sadness. European Journal of SocialPsychology, 24, 45–62. http://dx.doi.org/10.1002/ejsp.2420240104.

Bohner, G., Bless, H., Schwarz, N., & Strack, F. (1988). What triggers causal attributions?The impact of valence and subjective probability. European Journal of Social Psycholo-gy, 18(4), 335–345. http://dx.doi.org/10.1002/ejsp.2420180404.

Bohner, G., Erb, H. -P., & Siebler, F. (2008). Information processing approaches to persua-sion: Integrating assumptions from the dual- and single-processing perspectives. InW. D. Crano, & R. Prislin (Eds.), Attitudes and Attitude Change. New York: PsychologyPress, Taylor & Francis Group.

Bohner, G., Moskowitz, G. B., & Chaiken, S. (1995). The interplay of heuristic and system-atic processing of social information. European Review of Social Psychology, 6, 33–68.http://dx.doi.org/10.1080/14792779443000003.

1054 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

Borkovec, T. D. (1994). The nature, functions, and origins of worry. In G. C. L. Davey, & F.Tallis (Eds.), Worrying: Perspectives on Theory, Assessment and Treatment. Chichester,UK: John Wiley & Sons, Ltd.

Borkovec, T. D., Hazlett-Stevens, H., & Diaz, M. L. (1999). The role of positive beliefs aboutworry in generalized anxiety disorder and its treatment. Clinical Psychology &Psychotherapy, 6, 126–138. http://dx.doi.org/10.1002/(SICI)1099-0879(199905)6:2b126::AID-CPP193N3.0.CO;2-M.

Borkovec, T. D., & Inz, J. (1990). The nature of worry in generalized anxiety disorder:A predominance of thought activity. Behaviour Research and Therapy, 28, 153–158.http://dx.doi.org/10.1016/0005-7967(90)90027-G.

Borkovec, T. D., Ray, W. J., & Stöber, J. (1998). Worry: A cognitive phenomenon intimatelylinked to affective, physiological, and interpersonal behavioral processes. CognitiveTherapy and Research, 22(6), 561–576. http://dx.doi.org/10.1023/A:1018790003416.

Borkovec, T. D., Robinson, E., Pruzinsky, T., & DePree, J. A. (1983). Preliminary explorationof worry: Some characteristics and processes. Behaviour Research and Therapy, 21(1),9–16. http://dx.doi.org/10.1016/0005-7967(83)90121-3.

Borkovec, T. D., & Roemer, L. (1995). Perceived functions of worry among generalized anx-iety disorder subjects: Distraction from more emotionally distressing topics? Journalof Behavior Therapy and Experimental Psychiatry, 26(1), 25–30. http://dx.doi.org/10.1016/0005-7916(94)0064-5.

Brain, K., Henderson, B. J., Tyndel, S., Bankhead, C., Watson, E., & Clements, A. (2008). Pre-dictors of breast cancer-related distress following mammography screening in youn-ger women on a family history breast cancer screening programme. Psycho-Oncology,17(12), 1180–1188. http://dx.doi.org/10.1002/pon.1355.

Buhr, K., & Dugas, M. J. (2006). Investigating the construct validity of intolerance of uncer-tainty and its unique relationship with worry. Journal of Anxiety Disorders, 20(2),222–236. http://dx.doi.org/10.1016/janxdis.2004.12.004.

Burger, J. M., & Cooper, H. M. (1979). The desirability of control. Motivation and Emotion,3(4), 381–393. http://dx.doi.org/10.1007/BF00994052.

Butler, G., & Mathews, A. (1983). Cognitive processes in anxiety. Advances in BehaviourResearch and Therapy, 5(1), 51–62. http://dx.doi.org/10.1016/0146-6402(83)90015-2.

Butler, G., & Mathews, A. (1987). Anticipatory anxiety and risk perception. CognitiveTherapy and Research, 11(5), 551–565. http://dx.doi.org/10.1007/BF01183858.

Cacioppo, J. T., Petty, R. E., & Morris, K. J. (1983). Effects of need for cognition on messageevaluation, recall, and persuasion. Journal of Personality and Social Psychology, 45,805–818. http://dx.doi.org/10.1037/0022-3514.45.4.805.

Carleton, N. R., Norton, P. J. M.A., & Asmundson, G. J. G. (2007). Fearing the unknown:A short version of the Intolerance of Uncertainty scale. Journal of Anxiety Disorders,21(1), 105–117.

Carleton, N. R., Sharpe, D., & Asmundson, G. J. G. (2007). Anxiety sensitivity and intoler-ance of uncertainty: Requisites of the fundamental fears? Behaviour Research andTherapy, 45(10), 2307–2316. http://dx.doi.org/10.1016/j.brat.2007.04.006.

Carter, W. R., Johnson, M. C., & Borkovec, T. D. (1986). Worry: An electrocortical analysis.Advances in Behaviour Research and Therapy, 8(4), 193–204. http://dx.doi.org/10.1016/0146-6402(86)90004-4.

Cartwright-Hatton, S., & Wells, A. (1997). Beliefs about worry and intrusions: TheMeta-Cognitions Questionnaire and its correlates. Journal of Anxiety Disorders, 11(3),279–296. http://dx.doi.org/10.1016/S0887-6185(97)00011-X.

Cervone, D., Kopp, D. A., Schaumann, L., & Scott, W. D. (1994). Mood, self-efficacy, andhigher performance standards: Lower moods induce higher standards of performance.Journal of Personality and Social Psychology, 67(3), 499–512. http://dx.doi.org/10.1037/0022-3514.67.3.499.

Chaiken, S., Liberman, A., & Eagly, A. H. (1989). Heuristic and systematic information pro-cessingwithin and beyond the persuasion context. In J. S. Uleman, & J. A. Bargh (Eds.),Unintended Thought. New York: Guilford Press.

Chaiken, S., & Maheswaran, D. (1994). Heuristic processing can bias systematicprocessing: Effects of source credibility, argument ambiguity, and task importanceon attitude judgement. Journal of Personality and Social Psychology, 66(3), 460–473.http://dx.doi.org/10.1037/0022-3514.66.3.460.

Chaiken, S., & Trope, Y. (Eds.). (1999). Dual-process Theories in Social Psychology. NewYork: The Guilford Press.

Chen, S., & Chaiken, S. (1999). The Heuristic–SystematicModel in its broader context. In S.Chaiken, & Y. Trope (Eds.), Dual-process Theories in Social Psychology. New York: TheGuilford Press.

Chen, S., Duckworth, K., & Chaiken, S. (1999). Motivated heuristic and systematic processing.Psychological Inquiry, 10(1), 44–49. http://dx.doi.org/10.1207/s15327965pli1001_6.

Chen, J., Liu, X., Rapee, R. M., & Pillay, P. (2013). Behavioural activation: A pilot trial oftransdiagnostic treatment for excessive worry. Behaviour Research and Therapy, 51,533–539.

Chen, S., Shechter, D., & Chaiken, S. (1996). Getting at the truth or getting along: Accuracy-versus impression-motivated heuristic and systematic processing. Journal of Personalityand Social Psychology, 71(2), 262–275. http://dx.doi.org/10.1037/0022-3514.71.2.262.

Cohen, A.R., Stotland, E., & Wolfe, D.M. (1955). An experimental investigation of need forcognition. Journal of Abnormal and Social Psychology, 51, 291–294. http://dx.doi.org/10.1037/h0042761.

Covin, Roger, Ouimet, Allison J., Seeds, Pamela M., & Dozois, David J. A. (2008). Ameta-analysis of CBT for pathological worry among clients with GAD. Journal ofAnxiety Disorders, 22(1), 108–116.

Darke, Peter R., Chaiken, Shelly, Bohner, Gerd, Einwiller, Sabine, Erb, Hans-Peter, &Hazlewood, J. Douglas (1998). Accuracy motivation, consensus information, and thelaw of large numbers: Effects on attitude judgment in the absence of argumentation.Personality and Social Psychology Bulletin, 24(11), 1205–1215. http://dx.doi.org/10.1177/01461672982411007.

Dash, Suzanne R., & Davey, Graham C. L. (2012). An experimental investigation of the role ofnegative mood in worry: The role of appraisals that facilitate systematic informationprocessing. Journal of Behavior Therapy and Experimental Psychiatry, 43(2), 823–831.

Davey, G. C. L. (1993). A comparison of three cognitive appraisal strategies: The role ofthreat devaluation in problem-focussed coping. Personality and IndividualDifferences, 14(4), 535–546. http://dx.doi.org/10.1016/0191-8869(93)90146-T.

Davey, G. C. L. (1994). Worrying, social problem-solving abilities, and socialproblem-solving confidence. Behaviour Research and Therapy, 32(3), 327–330.http://dx.doi.org/10.1016/0005-7967(94)90130-9.

Davey, G. C. L. (2006a). The catastrophising interview procedure. In G. C. L. Davey, & A.Wells (Eds.), Worry and Its Psychological Disorders. Chichester, Uk: Wiley.

Davey, G. C. L. (2006b). A mood-as-input account of perseverative worrying. In G. C. L.Davey, & A. Wells (Eds.),Worry and Its Psychological Disorders. Chichester, UK: Wiley.

Davey, G. C. L., Hampton, J., Farrell, J., & Davidson, S. (1992). Some characteristics of worrying:Evidence for worrying and anxiety as separate constructs. Personality and IndividualDifferences, 13(2), 133–147. http://dx.doi.org/10.1016/0191-8869(92)90036-0.

Davey, G. C. L., Jubb, M., & Cameron, C. (1996). Catastrophic worrying as a function ofchanges in problem-solving confidence. Cognitive Therapy and Research, 20,333–344. http://dx.doi.org/10.1007/BF02228037.

Davey, G. C. L., & Levy, S. (1998a). Internal statements associated with catastrophic wor-rying. Personality and Individual Differences, 26(1), 21–32. http://dx.doi.org/10.1016/S0191-8869(98)00162-7.

Davey, G. C. L., & Levy, S. (1998b). Catastrophic worrying: Personal inadequacy and a persev-erative iterative style as features of the catastrophizing process. Journal of AbnormalPsychology, 107(4), 576–586. http://dx.doi.org/10.1037/0021-843X.107.4.576.

Davey, G. C. L., Startup, H. M., MacDonald, B., Jenkins, D., & Patterson, K. (2005). The use of‘as many as can’ versus ‘feel like continuing’ stop rules during worrying. CognitiveTherapy and Research, 29, 155–169. http://dx.doi.org/10.1007/s10608-005-3162-5.

Davey, G. C. L., Tallis, F., & Capuzzo, N. (1996). Beliefs about the consequences of worrying.Cognitive Therapy and Research, 20(5), 499–520. http://dx.doi.org/10.1007/BF02227910.

De Martino, B., Harrison, N. A., Knafo, S., Bird, G., & Dolan, R. J. (2008). Explainingenhanced logical consistency during decision making in autism. The Journal ofNeuroscience, 28(42), 10746–10750.

Dobson, K. S., Hollon, S. D., Dimidjian, S., Schmaling, K. B., Kohlenberg, R. J., & Gallop, R. J.(2008). Randomized trial of behavioral activation, cognitive therapy, and antidepres-sant medication in the prevention of relapse and recurrence in major depression.Journal of Consulting and Clinical Psychology, 76(3), 468–477.

Dugas, M. J., Buhr, K., & Ladouceur, R. (2004). The role of intolerance of uncertainty in theetiology and maintenance of generalized anxiety disorder. In R. G. Heimberg, C. L.Turk, & D. S. Mennin (Eds.), Generalized Anxiety Disorder: Advances in Research andPractice (pp. 143–163). : Guilford Press.

Dugas, M. J., Freeston, M. H., & Ladouceur, R. (1997). Intolerance of uncertainty andproblem orientation in worry. Cognitive Therapy and Research, 20, 593–606.http://dx.doi.org/10.1023/A:1021890322153.

Dugas, M. J., Gagnon, F., Ladouceur, R., & Freeston, M. H. (1998). Generalized anxietydisorder: A preliminary test of a conceptual model. Behaviour Research and Therapy,36, 215–226. http://dx.doi.org/10.1016/S0005-7967(97)00070-3.

Erb, H. -P., Pierro, A., Mannetti, L., Spiegel, S., & Kruglanski, A.W. (2007). Biased processing ofpersuasive information: On the functional equivalence of cues and message arguments.European Journal of Social Psychology, 37, 1057–1075. http://dx.doi.org/10.1002/ejsp.420.

Evans, J. S. B. T. (2008). Dual-processing accounts of reasoning, judgment, and socialcognition. Annual Review of Psychology, 59, 255–278.

Festinger, L. (1957). A Theory of Cognitive Dissonance. Stanford: Stanford University Press.Fiske, S. T., & Taylor, S. E. (1991). Social Cognition (2 ed.)New York: McGraw-Hill, Inc.Forehand,Mark, Gastil, John, & Smith, Mark A. (2004). Endorsements as voting cues: Heu-

ristic and systematic processing in initiative elections1. Journal of Applied SocialPsychology, 34(11), 2215–2231.

Freeston, M. H., Dugas, M. J., & Ladouceur, R. (1996). Thoughts, images, worry, and anxiety.Cognitive Therapy and Research, 20(3), 265–273. http://dx.doi.org/10.1007/BF02229237.

Freeston, M. H., Rhéaume, J., Letarte, H., Dugas, M. J., & Ladouceur, R. (1994). Why do peopleworry? Personality and Individual Differences, 17(6), 791–802. http://dx.doi.org/10.1016/0191-8869(94)90048-5.

Frost, R. O., Marten, P., Lahart, C., & Rosenblate, R. (1990). The dimensions of perfectionism.Cognitive Therapy and Research, 14(449–468). http://dx.doi.org/10.1007/BF01172967.

Gigerenzer, G., & Gaissmaier, W. (2011). Heuristic decision making. Annual Review ofPsychology, 62, 451–482.

Giner-Sorolila, R., & Chaiken, S. (1997). Selective use of heuristic and systematic process-ing under defense motivation. Personality and Social Psychology Bulletin, 23(1), 84–97.http://dx.doi.org/10.1177/0146167297231009.

Griffin, R. J., Zheng, Y., ter Huurne, E., Boerner, F., Ortiz, S., & Dunwoody, S. (2008).After the flood. Science Communication, 29(3), 285–315. http://dx.doi.org/10.1177/1075547007312309.

Hakamata, Y., Lissek, S., Bar-Haim, Y., Britton, J. C., Fox, N. A., & Leibenluft, E. (2010). Atten-tion bias modification treatment: A meta-analysis toward the establishment of noveltreatment for anxiety. Biological Psychiatry, 11(1), 982–990.

Hale, Jerold L., Lemieux, Robert, & Mongeau, Paul A. (1995). Cognitive processingof fear-arousing message content. Communication Research, 22(4), 459–474.http://dx.doi.org/10.1177/009365095022004004.

Hayes, S., Hirsch, C. R., & Mathews, A. (2008). Restriction of working memory capacityduring worry. Journal of Abnormal Psychology, 117(3), 712–717. http://dx.doi.org/10.1037/a0012908.

Heppner, P. P., & Lee, D. -G. (2002). Problem-solving, appraisal and psychological adjust-ment. In C. R. Snyder, & S. J. Lopez (Eds.), Handbook of Positive Psychology. New York:Oxford University Press.

Hirsch, Colette R., MacLeod, Colin, Mathews, Andrew, Sandher, Oneet, Siyani, Amruti, &Hayes, Sarra (2011). The contribution of attentional bias to worry: Distinguishingthe roles of selective engagement and disengagement. Journal of Anxiety Disorders,25(2), 272–277.

1055S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

Hirsch, C. R., & Mathews, A. (2012). A cognitive model of pathological worry. BehaviourResearch and Therapy, 50, 636–646.

Hoehn-Saric, R., Schlund, M. W., &Wong, S. H. (2004). Effects of Citalopram onworry andbrain activation in patients with generalized anxiety disorder. Psychiatry Research:Neuroimaging, 131, 11–21. http://dx.doi.org/10.1016/j.psychresns.2004.02.003.

Johnston, W. M., & Davey, G. C. L. (1997). The psychological impact of negative TV newsbulletins: The catastrophizing of personal worries. British Journal of Psychology,88(1), 85. http://dx.doi.org/10.1111/j.2044-8295.1997.tb02622.x.

Kahlor, L., Dunwoody, S., Griffin, R. J., Neuwirth, K., & Giese, J. (2003). Studyingheuristic–systematic processing of risk communication. Risk Analysis, 23(2), 355–368.http://dx.doi.org/10.1111/1539-6924.00314.

Koerner, N., & Dugas, M. J. (2008). An investigation of appraisals in individuals vulnerableto excessive worry: The role of intolerance of uncertainty. Cognitive Therapy andResearch, 32, 619–638. http://dx.doi.org/10.1007/s10608-007-9125-2.

Kofta, M., & Sedek, G. (1998). Uncontrollability as a source of cognitive exhaustion: Impli-cations for helplessness and depression. In M. Kofta, G. Weary, & G. Sedek (Eds.), Per-sonal Control in Action: Cognitive and Motivational Mechanisms. New York: PlenumPress.

Kraus, S. J. (1995). Attitudes and the prediction of behaviour: A meta-analysis of the em-pirical literature. Personality and Social Psychology Bulletin, 21, 58–75.

Kruglanski, A. W., Thompson, E. P., & Spiegel, S. (1999). Separate or equal? Bimodal no-tions of persuasion and a single-process "unimodel" (p.p. 293–313). In S. Chaiken,& Y. Trope (Eds.), Dual-process theories in social psychology. New York: Guilford Press.

Ladouceur, R., Blais, M. H., Freeston, M. H., & Dugas, M. J. (1998). Problem solving andproblem orientation in generalized anxiety disorder. Journal of Anxiety Disorders,12(2), 139–152. http://dx.doi.org/10.1016/S0887-6185(98)00002-4.

Ladouceur, R., Talbot, F., & Dugas, M. J. (1997). Behavioral expressions of intolerance of un-certainty in worry. Behavior Modification, 21(3), 355–371. http://dx.doi.org/10.1177/01454455970213006.

Langlois, F., Freeston, M. H., & Ladouceur, R. (2000). Differences and similarities betweenobsessive intrusive thoughts and worry in a non-clinical population: Study 1.Behaviour Research and Therapy, 38, 157–173. http://dx.doi.org/10.1016/S0005-7967(99)00027-3.

Leigh, E., & Hirsch, C. R. (2011). Worry in imagery and verbal form: Effect on resid-ual working memory capacity. Behaviour Research and Therapy, 49(2), 99–105.http://dx.doi.org/10.1016/j.brat.2010.11.005.

Leventhal, H. (1970). Findings and theory in the study of fear communications. Advancesin experimental social psychology, 5, 119–186.

Leynes, A. P. (2002). The effect of specific test queries on source-monitoring event-relatedpotentials. Brain and Cognition, 50(2), 218–233. http://dx.doi.org/10.1016/S0278-2626(02)00505-5.

Leynes, A., & Phillips, M. C. (2008). Event-Related Potential (ERP) evidence for varied recol-lection during source monitoring. Journal of Experimental Psychology: Learning, Memory,and Cognition, 34(4), 741–751. http://dx.doi.org/10.1037/0278-7393.34.4.741.

Livingston, S. D., & Sinclair, L. (2008). Taking the watchdog off its leash: Personal preju-dices and situational motivations jointly predict derogation of a stigmatized source.Personality and Social Psychology Bulletin, 34, 210–223. http://dx.doi.org/10.1177/0146167207310028.

MacLeod, A. K., Williams, M. J., & Bekerian, D. A. (1991). Worry is reasonable: The role ofexplanations in pessimism about future personal events. Journal of AbnormalPsychology, 100(4), 478–486. http://dx.doi.org/10.1037/0021-843X.100.4.478.

Maheswaran, D., & Chaiken, S. (1991). Promoting systematic processing in low-motivationsettings: Effect of incongruent information on processing and judgement. Journal ofPersonality and Social Psychology, 61(1), 13–25. http://dx.doi.org/10.1037/0022-3514.61.1.13.

Martin, R., & Hewstone,M. (2003). Majority versusminority influence:When, not wheth-er, source status instigates heuristic or systematic processing. European Journal of So-cial Psychology, 33(3), 313–330. http://dx.doi.org/10.1002/ejsp.146.

Mathews, A. (1990). Why worry? The cognitive function of anxiety. Behaviour Researchand Therapy, 28(6), 455–468. http://dx.doi.org/10.1016/0005-7967(90)90132-3.

McEvoy, Peter M., & Mahoney, Alison E. J. (2011). Achieving certainty about the structureof intolerance of uncertainty in a treatment-seeking sample with anxiety and depres-sion. Journal of Anxiety Disorders, 25(1), 112–122.

Meeten, F., & Davey, G. C. L. (2011). Mood-as-input hypothesis and perseverative psychopa-thologies. Clinical Psychology Review, 31(8), 1259–1275. http://dx.doi.org/10.1016/j.cpr.2011.08.002.

Mennin, D. S., Heimberg, R. G., Turk, C. L., & Fresco, D.M. (2002). Applying an emotion reg-ulation framework to integrative approaches to generalized anxiety disorder. ClinicalPsychology: Science and Practice, 9(1), 85–90. http://dx.doi.org/10.1093/clipsy.9.1.85.

Mennin, Douglas S., Heimberg, Richard G., Turk, Cynthia L., & Fresco, David M. (2005). Pre-liminary evidence for an emotion dysregulation model of generalized anxiety disorder.Behaviour Research and Therapy, 43(10), 1281–1310. http://dx.doi.org/10.1016/j.brat.2004.08.008.

Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990). Development and vali-dation of the Penn State Worry Questionnaire. Behaviour Research and Therapy, 28,487–495. http://dx.doi.org/10.1016/0005-7967(90)90135-6.

Monin, B. (2003). The warm glow heuristic: When liking leads to familiarity. Journal ofPersonality and Social Psychology, 85, 1035–1048.

Moulding, R., & Kyrios, M. (2007). Desire for control, sense of control and obsessive–compulsive symptoms. Cognitive Therapy and Research, 31, 759–772. http://dx.doi.org/10.1007/s10608-006-9086-x.

Mushtaq, F., Bland, A.R., & Schaefer, A. (2011). Uncertainty and cognitive control. Frontiersin Psychology, 2, 249. http://dx.doi.org/10.3389/fpsyg.2011.00249.

Niceli, M., & Castelfranchi, C. (2001). Anxiety as an ‘epistemic’ emotion: An uncertaintytheory of anxiety. Anxiety, Stress & Coping: An International Journal, 18(4), 291–319.http://dx.doi.org/10.1080/10615800500209324.

Nolde, S. F., Johnson, M. K., & Raye, C. L. (1998). The role of prefrontal cortex during testsof episodic memory. Trends in Cognitive Sciences, 2(10), 399–406. http://dx.doi.org/10.1016/S1364-6613(98)01233-9.

O'Keefe, D. J. (2002). Persuasion: Theory & Research, Vol. 2, Thousand Oaks, CA: Sage Pub-lications, Inc.

Petty, R. E., Briñol, P., & Tormala, Z. L. (2002). Thought confidence as a determinant of per-suasion: The self-validation hypothesis. Journal of Personality and Social Psychology,82, 722–741. http://dx.doi.org/10.1037/0022-3514.82.5.722.

Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on responses to argumentquantity and quality: Central and peripheral routes to persuasion. Journal ofPersonality and Social Psychology, 46(1), 69–81. http://dx.doi.org/10.1037/0022-3514.71.1.39.

Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persuasion. In L.Berkowitz (Ed.), Advances in Experimental Social Psychology, Vol. 19. (pp. 123–205)New York: Academic Press.

Petty, R. E., &Wegener, D. T. (1999). The elaboration likelihoodmodel: Current status andcontroversies. In S. Chaiken, & Y. Trope (Eds.), Dual-process Theories in Social Psychol-ogy. New York: The Guilford Press.

Pieters, R. G. M., & Verplanken, B. (1995). Intention–behaviour consistency: Effects of con-sideration set size, involvement and need for cognition. European Journal of Social Psy-chology, 25, 531–543.

Pratt, P., Tallis, F., & Eysenck, M. (1997). Information-processing, storage characteristicsand worry. Behaviour Research and Therapy, 35, 1015–1023. http://dx.doi.org/10.1016/S0005-7967(97)00057-0.

Purdon, C. L., & Harrington, J. (2006). Worry in psychopathology. In G. C. L. Davey, & A.Wells (Eds.), Worry and Its Psychological Disorders: Theory, Assessment and Treatment.Chichester, UK: Wiley.

Ratneshwar, S., & Chaiken, S. (1991). Comprehension's role in persuasion: The case of itsmoderating effect on the persuasive impact of source cues. The Journal of ConsumerResearch, 18, 52–62.

Rector, Neil A., Kamkar, Katy, Cassin, Stephanie E., Ayearst, Lindsay E., & Laposa, Judith M.(2011). Assessing excessive reassurance seeking in the anxiety disorders. Journal ofAnxiety Disorders, 25(7), 911–917.

Robichaud, M., & Dugas, M. J. (2005). Negative problem orientation (Part I): Psycho-metric properties of a new measure. Behaviour Research and Therapy, 43(3), 391–401.http://dx.doi.org/10.1016/j.brat.2004.02.007.

Roemer, L., & Borkovec, T. D. (1993). Worry: Unwanted cognitive activity that controlsunwanted somatic experience. In D.M. Wegner, & J. W. Pennebaker (Eds.), Handbookof Mental Control. Englewood Cliffs, NJ, US: Prentice-Hall, Inc.

Roemer, L., Molina, S. M. S., & Borkovec, T. D. (1997). An investigation of worry contentamong generally anxious individuals. The Journal of Nervous and Mental Disease,185(5), 314–319.

Roemer, L., & Orsillo, S. M. (2005). An acceptance-based behaviour therapy for generalizedanxiety disorder. In S. M. Orsillo, & L. Roemer (Eds.), Acceptance and Mindfulness-basedApproaches to Anxiety: Conceptualization and Treatment. New York: Springer.

Ruscio, A.M., Borkovec, T. D., &Ruscio, J. (2001). A taxometric investigationof the latent struc-ture of worry. Journal of Abnormal Psychology, 110(3), 413–422. http://dx.doi.org/10.1037/0021-843X.110.3.413.

Salkovskis, Paul M., Richards, H. Candida, & Forrester, Elizabeth (1995). The relationshipbetween obsessional problems and intrusive thoughts. Behavioural and CognitivePsychotherapy, 23(03), 281–299. http://dx.doi.org/10.1017/S1352465800015885.

Salters-Pedneault, K., Roemer, L., Tull, M. T., Rucker, L., & Mennin, D. S. (2006). Evidence ofbroad deficits in emotion regulation associated with chronic worry and generalizedanxiety disorder. Cognitive Therapy and Research, 30(4), 469–480. http://dx.doi.org/10.1007/s10608-006-9055-4.

Schwarz, N., & Clore, G. L. (1983). Mood, misattribution, and judgments of well-being: In-formative and directive functions of affective states. Journal of Personality and SocialPsychology, 45(3), 513–523. http://dx.doi.org/10.1037/0022-3514.45.3.513.

Scott, W. D., & Cervone, D. (2002). The impact of negative affect on performance stan-dards: Evidence for an affect-as-information mechanism. Cognitive Therapy andResearch, 26, 19–37. http://dx.doi.org/10.1023/A:1013837520031.

Shah, A. K., & Oppenheimer, D.M. (2008). Heuristics made easy: An effort-reductionframework. Psychological Bulletin, 134(2), 207–222.

Simon, H. A. (1957). Models of Man. New York: John Wiley & Sons, Inc.Sookman, D., & Pinard, G. (2002). Overestimation of threat and intolerance of uncertainty

in obsessive compulsive disorder. In R. O. Frost, & G. Steketee (Eds.), Cognitive Ap-proaches to Obsessions and Compulsions: Theory, Assessment and Treatment. Oxford:Elsevier Science, Ltd.

Startup, H. M., & Davey, G. C. L. (2001). Mood-as-input and catastrophic worrying. Journal ofAbnormal Psychology, 110(1), 83–96. http://dx.doi.org/10.1037//0021-843X.110.1.83.

Startup, H. M., & Davey, G. C. L. (2003). Inflated responsibility and the use of stoprules for catastrophic worrying. Behaviour Research and Therapy, 41, 495–503.http://dx.doi.org/10.1016/S0005-7967(02)00245-0.

Startup, H. M., Freeman, D., & Garety, P. A. (2007). Persecutory delusions and catastrophicworry in psychosis: Developing the understanding of delusion distress andpersistence.Behaviour Research and Therapy, 45, 523–537. http://dx.doi.org/10.1016/j.brat.2006.04.006.

Steginga, S. K., & Occhipinti, S. (2004). The application of the heuristic–systematic pro-cessing model to treatment decision making about prostate cancer. Medical decisionmaking, 24(6), 573–583. http://dx.doi.org/10.1177/0272989X04271044.

Sternheim, L., Konstantellou, A., Startup, H. M., & Schmidt, U. (2011). What does uncertaintymean to women with anorexia nervosa? An interpretative phenomenological analysis.European Eating Disorders Review, 19(1), 12–24. http://dx.doi.org/10.1002/erv.1029.

Stöber, J., & Joormann, J. (2001). A short form of the worry domains questionnaire:Construction and factorial validation. Personality and Individual Differences, 31(4),591–598. http://dx.doi.org/10.1016/S0191-8869(00)00163-X.

1056 S.R. Dash et al. / Clinical Psychology Review 33 (2013) 1041–1056

Sugiura, Y. (2007). Responsibility to continue thinking and worrying: Evidence of incre-mental validity. Behaviour Research and Therapy, 45, 1619–1628. http://dx.doi.org/10.1016/j.brat.2006.08.001.

Swann, W. B., Stephenson, B., & Pittman, T. S. (1981). Curiosity and control: On the deter-minants of the search for social knowledge. Journal of Personality and Social Psychology,40(4), 635–642.

Sweet, M., & Pelton-Sweet, L. M. (2008). The social foundation of team-based learning: Stu-dents accountable to students. New Directions for Teaching and Learning, 116, 29–40.

Szabó, M., & Lovibond, P. F. (2002). The cognitive content of naturally occurring worry ep-isodes. Cognitive Therapy and Research, 26(2), 167–177. http://dx.doi.org/10.1023/A:1014565602111.

Tallis, F., & DeSilva, P. (1992). Worry and obsessional symptoms: A correlational analysis.Behaviour Research and Therapy, 30, 103–105.

Tallis, F., Eysenck, M. W., & Mathews, A. (1991). Elevated evidence requirements and worry.Personality and Individual Differences, 12, 505–520. http://dx.doi.org/10.1016/0191-8869(91)90128-X.

Tetlock, Philip E. (1983). Accountability and the perseverance of first impressions. SocialPsychology Quarterly, 46(4), 285–292.

Tiedens, L. Z., & Linton, S. (2001). Judgment under emotional certainty and uncertainty:The effects of specific moods on information processing. Journal of Personality andSocial Psychology, 81(6), 973–988. http://dx.doi.org/10.1037/0022-3514.81.6.973.

Tormala, Z. L., Rucher, D.D., & Seger, C. R. (2008). When increased confidence yieldsincreased thought: A confidence-matching hypothesis. Journal of Experimental SocialPsychology, 44, 141–147. http://dx.doi.org/10.1016/j.jesp.2006.11.002.

Trumbo, C. W. (1999). Heuristic–systematic information processing and risk judgment.Risk Analysis, 19(3), 391–400. http://dx.doi.org/10.1111/j.1539-6924.1999.tb00415.x.

Trumbo, C. W. (2002). Information processing and risk perception: An adaptation ofthe heuristic–systematic model. The Journal of Communication, 52(2), 367–382.http://dx.doi.org/10.1111/j.1460-2466.2002.tb02550.x.

Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging frequency andprobability. Cognitive Psychology, 5, 207–232.

Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases.Science, 185, 1124–1131.

Uleman, J. S. (1989). The self-control of thoughts: A framework for thinking aboutunintended thought. In J. S. Uleman, & J. A. Bargh (Eds.), Unintended Thought(pp. 425–449). New York: Guilford Press.

Vasey, M., & Borkovec, T. D. (1992). A catastrophising assessment of worrisome thoughts.Cognitive Therapy and Research, 15, 505–520. http://dx.doi.org/10.1007/BF01175138.

Vaughn, L. A., Malik, J., Schwartz, S., Petkova, Z., & Trudeau, L. (2006). Regulatory fit as inputfor stop rules. Journal of Personality and Social Psychology, 91, 601–611. http://dx.doi.org/10.1037/0022-3514.91.4.601.

Verkuil, Bart, Brosschot, Jos F., Borkovec, T. D., & Thayer, Julian F. (2009). Acute autonomiceffects of experimental worry and cognitive problem-solving: Why worry aboutworry. International Journal of Clinical and Health Psychology, 9(3), 439–453.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measuresof positive and negative affect: The PANAS scales. Journal of Personality and SocialPsychology, 54(8), 1063–1070.

Weary, G., & Jacobson, J. A. (1997). Causal uncertainty beliefs and diagnostic informationseeking. Journal of Personality and Social Psychology, 73(839–848). http://dx.doi.org/10.1037/0022-3514.73.4.839.

Wells, A. (1994). Attention and the control of worry. In G. C. L. Davey, & F. Tallis(Eds.), Worrying: Perspectives on Theory, Assessment and Treatment. Chichester,UK: Wiley.

Wells, A. (1995). Meta-cognition and worry: A cognitive model of generalized anxietydisorder. Behavioural and Cognitive Psychotherapy, 23, 301–320. http://dx.doi.org/10.1017/S1352465800015897.

Wells, A. (1999). A metacognitive model and therapy for generalized anxiety disorder.Clinical Psychology & Psychotherapy, 6(2), 86–95. http://dx.doi.org/10.1002/(SICI)1099-0879(199905)6:2b86::AID-CPP189N3.0.CO;2-S.

Wells, A. (2006). The metacognitive model of worry and generalised anxiety disorder. InG. C. L. Davey, & A. Wells (Eds.),Worry and Its Psychological Disorders. Chichester, Uk:John Wiley & Sons Ltd.

Wells, A., & Carter, K. (1999). Preliminary tests of a cognitive model of generalized anxietydisorder. Behaviour Research and Therapy, 37, 585–594. http://dx.doi.org/10.1016/S0005-7967(98)00156-9.

Wells, A., & Carter, K. (2001). Further tests of a cognitive model of generalized anxietydisorder: Metacognitions and worry in GAD, panic disorder, social phobia, depres-sion, and nonpatients. Behavior Therapy, 32(1), 85–102. http://dx.doi.org/10.1016/S0005-7894(01)80045-9.

Wells, Adrian, & Cartwright-Hatton, Sam (2004). A short form of the metacognitionsquestionnaire: Properties of the MCQ-30. Behaviour Research and Therapy, 42(4),385–396. http://dx.doi.org/10.1016/S0005-7967(03)00147-5 (http://dx.doi.org/10.1016/S0005-7967(03)00147-5)

Wells, A., & King, P. (2006). Metacognitive therapy for generalized anxiety disorder:An open trial. Journal of Behavior Therapy and Experimental Psychiatry, 37(3),206–212.

Wells, A., & Papageorgiou, C. (1998). Relationships betweenworry, obsessive–compulsivesymptoms and meta-cognitive beliefs. Behaviour Research and Therapy, 36(9),899–913. http://dx.doi.org/10.1016/S0005-7967(98)00070-9.

Zanna, M. P., Kiesler, C. A., & Pilkonis, P. A. (1970). Positive and negative attitudinal affectestablished by classical conditioning. Journal of Personality and Social Psychology,14(4), 321.