Self-Control and Health Behaviour - PowerPoint Presentation

1. Self-Control and Health BehaviourMartin S. HaggerJohn Curtin Distinguished Professor, Curtin University, Perth, AustraliaFinland Distinguished Professor (FiDiPro), University of Jyväskylä, Finland

2. What is ‘Self-Control’?Self-control is central to goal-directed behaviourCapacity to actively engage in directed, persistent behaviour to obtain a distal goalForego small, short term gainsReflects people’s ability or capacity to:Override impulsesResist temptationBreak habitsAvoid the ‘dominant response’Why is self-control important?

3. Problems Associated with Lack of Self-ControlObesityDrinking Too Much AlcoholEating DisordersSmokingDrug addictionSexually Transmitted DiseaseUnplanned PregnancyPersonal debtGambling problemsCrimeAntisocial-behaviour

4. Adaptive Outcomes Linked to Self-ControlSuccess at schoolSuccess in the workplaceCohesive personal relationshipsGood healthAbility to cope with difficulties and barriersReduced susceptibility to social ills like crime and drug use

5. Typical approachesDispositional or individual difference perspectiveDelaying gratification (Mischel et al., 1988)‘Hot’ vs. ‘cool’ distinctionCool = reflective, deliberative, slow(er), effortful: “do it if it makes sense”Hot = impulsive, emotive, spontaneous, fast, low effort: “if it feels good, do it”Theoretical Perspectives

6. Self-control is associated with health-related outcomes across multiple studiesSelf-control and Health Behaviourde Ridder et al. (2012). Taking stock of self-control: A meta-analysis of how trait self-control relates to a wide range of behaviours. Personality and Social Psychology Review.Self-ControlHealthBehaviour.26 [.23;28]

7. What processes might determine action of trait self-control on health behaviour?Organizing tendency, planningOrganise/structure long-term goalsRecognise/predict costs and consequencesRecognise deviations from goalsRecognise cues to/signals of poor regulationDetect attention/motivation shifts & rectify them (Gottfredson & Hirshci, 1990; Inzlicht & Schmeichel, 2012)Work ethic, motivation, effortful tendencyConscientiousness, HEXACO (Ashton et al., 2004)‘Non’-impulsivityImpulse suppression (Friese & Hofmann, 2009)Explaining Self-Control Effects: Mechanisms

8. Mechanisms by which trait self-control may predict health behaviourTrait self-control may compel individuals to strategically form beliefs and intentionsBeliefs and intentions related to effortful, deliberative decisions to actHypothesis: Effect of trait self-control on health behaviour would be mediated by beliefs and intentionsSelf-Control and Intentional Health behaviour

9. PerceivedControlAttitudesSubjective NormsIntentionsBehaviourSource: Ajzen (1991). Organizational Behavior and Human Decision ProcessesTheory of Planned Behaviour

10. Proposed Model-0.38Social Cognitive VariablesIntentionHealth BehaviourTrait Self-ControlEffects of self-control mediated by social cognitive variables and intentionsBeliefs/expectanciesAttitudesSubjective normsPerceived controlSelf-efficacyIntentionsPlans to engagein behaviour in future

11. Individuals often do not follow through on their intentions (Orbell & Sheeran, 1999)Inclined abstainers or unsuccessful intenders (Rhodes & de Bruin, 2014)As trait self-control is related to effortful engagement, individuals with high trait self-control better enact their intentionsHypothesis: Effect of intentions on health behaviour would be moderated by trait self-controlSelf-Control and Intentional Health Behaviour

12. Proposed Model-0.38Trait Self-ControlSocial Cognitive VariablesIntentionHealth BehaviourTrait Self-ControlI-B relationship conditionalon level of self-controlBehavioursFruit & vegJunk foodDiet restrictingWalkingDental flossingPhysical activityAlcohol intakeBinge drinking‘Pre-drinking’Hagger (2014). Annals of behavioural Medicine. http//:dx.doi.org/10.1007/s12160-014-9631-x

13. 6 data setsCorrelational design with a prospective measure of behaviourTrait self-control: Tangney et al. (2004) scale or self-discipline scale from HEXACO-C scale (Ashton et al., 2004)Social cognitive variables: intention, subjective norms, perceived behavioural control/self-efficacy, outcome expectancies, and risk perceptionsSelf-report measures of behaviour:Sample 1: fruit and veg, fast foodSample 2: restrict diet, binge drinking, physical activitySample 3: alcohol intake, physical activity (walking)Sample 4: dental flossingSample 5: out-of-school exerciseSample 6: pre-drinkingData analysed using bootstrapped path analyses (Hayes, 2017) with the lavaan package in RTesting the Mediation and Moderation Effects

14. Findings-0.38Social Cognitive VariablesIntentionHealth BehaviourTrait Self-ControlIndirect effects (7/10 behaviours)Fruit & vegFast food (-ive)Diet restrictionBinge drinkingPhysical activityLeisure-time exercisePre-drinkingDirect effects (6/10)Fast food (-ive)Diet restrictionBinge drinkingPhysical activityAlcohol intakeLeisure-time exerciseIndirect effects only (2/10)Fruit & vegPre-drinkingDirect effects only (1/10)Alcohol intakeHagger, Hankonen, Kangro, Reis, Lintunen, Polet & Hamilton (2018). Under Revision.

15. FindingsTrait Self-ControlIntentionHealth BehaviourDiet restrictionBinge drinkingAlcohol consumptionHagger, Hankonen, Kangro, Reis, Lintunen, Polet & Hamilton (2018). Under Revision.

16. Reasonably good evidence to support mediation effect – trait self-control associated with beliefs/intentions, which translate to behaviourBut direct effects of trait self-control remained – not mediated by beliefs/intentionsMaybe that reflects non-conscious/non-deliberative effectsEvidence of a moderating effect on the intention-behaviour relationship less strongMay be most relevant in behaviours where there is a strong impulsive element (eating, alcohol consumption)No effect in behaviours where there is less impulse control (physical activity)Summary and Conclusion

17. Other Theoretical Perspectives: Ego-DepletionThe ‘Strength’ or ‘Resource Depletion’ Model of Self-Control (Baumeister et al., 1998)Acts of self-control consume a limited ‘energy’ resourceEngaging in an act of self-control results in reduced performance in subsequent acts of self-controlFocuses on a ‘state’ perspective – resource availabilityThe state of self-control failure termed ego-depletionExtremely popular and influential perspective

18. The Strength Model: Initial HypothesesPredictions:Self-control is a finite resource that becomes depleted after a period of exertion (Baumeister et al., 1998)Self-control ‘strength’ is reduced on subsequent SC tasks (Baumeister & Vohs, 2007)Restored after a period of recuperation (Tyler & Burns, 2008) Could be improved through training (Muraven et al., 1999)

19. Experimental Tests of Ego-DepletionSequential-task paradigmExperimental Task NotRequiringSelf-controlSelf-controlTaskControlSelf-controlTaskDependent measureTASK 1TASK 2

20. What types of behaviours demand self control ‘resources’?Example behavioursExample lab tasksEating ‘tasty’ unhealthy foodControlling impulsesSmoking & alcohol drinkingControlling thoughtsCompulsive spendingControlling emotionsEffortful decision makingMaking choicesExercise & physical activity

21. Hagger, M. S., Wood, C., Stiff, C., & Chatzisarantis, N. L. D. (2010). Ego depletion and the strength model of self-control: A meta-analysis. Psychological Bulletin, 136, 495-525. doi: 10.1037/a0019486Meta-Analysis of Ego-Depletion Effect

22. Meta-Analysis of Ego-Depletion Effectd = 0.44Self-controlResourceDepletionTaskPerformanceTaskDifficultyd = 0.94d = 0.62FatigueSource: Hagger, Wood, Stiff & Chatzisarantis (2010). Psychological Bulletin.Effortd = 0.64

23. Research suggests that depleted individuals (Martin-Ginis & Bray, 2011) Spent less time on a cycling task Planned to exercise less Engaged in fewer bouts of exercise over 8-weeksTraining self-control improves exercise adherence (Oaten & Cheng, 2007)Exercise program improves self-control – reduced smoking and less dietary fat consumed (Oaten & Cheng, 2008)Depletion and Exercise

24. Depletion and ExerciseDesign: 2 (depletion: depletion task vs. non-depleting task) x 2 (time: pre- vs. post)Depleting task: Counting and coordination taskDependent variables: Energy expenditure on cycle ergometer/RPEEnergy expenditure difference (kJ)Source: Martin-Ginis & Bray (2011). Psychology & HealthRating of perceived exertion diff.

25. Training self-control and exerciseSource: Oaten & Cheng (2006). Br. J. Health Psychol.Design: 8-week staggered entry RCTTraining task: Computerised visual tracking taskDependent variables: Free visits to gym

26. Ego-Depletion and Eating BehaviourChronic dieters exposed to temptation and availability exhibited increased eatingAmount of ice-creameaten (grams)Source: Vohs & Heatherton (2000). Psychol. Sci.

27. Ego-Depletion and Dietary BehaviourAim: Examine effect of ego-depletion on eating behaviour in people with elevated and normal BMIHypothesis: People with elevated BMI (low dietary restraint) will be more susceptible to state self-control depletion and eat more food when depleted

28. Design and MethodParticipants. UG students (N = 83; M age = 19.00, SD = 1.21) identified as elevated (≥25) or normal (<25) BMIDesign. 2 (ego-depletion: depletion vs. control) x 2 (eating restraint: High BMI vs. normal BMI) between-participants experimental designEgo-depletion manipulation. Counting and co-ordination task (easy vs. hard versions)

29. Design and MethodProcedure. Participants identified using pre-screening and fasted for 2hr prior to study. Randomly allocated to ego-depletion or control condition. After depletion or control task participants engaged in a ‘taste-and-rate’ perception task for cookies and smarties.Dependent measures. Amount of food eaten in ostensible taste perception task (cookies and smarties).

30. ResultsSource: Hagger, M. S., Panetta, G., Leung, C.-M., Wong, G. G., Wang, J. C. K., Chan, D. K.-C., . . . Chatzisarantis, N. L. D. (2013). Chronic inhibition, self-control and eating behaviour: Test of a 'resource depletion' model. PLoS ONE, 8, e76888.

31. Ego-Depletion and SmokingSelf-control has been implicated in attempts to quit smokingEgo-depleted smokers were more likely to smoke during a break than non-depleted smokers (Shmueli & Prochaska, 2009)Ego-depletion could be overturned among smokers given the opportunity to smoke a cigarette (Heckman, Ditre, & Brandon, 2012)No research has implicated cigarette cravings on self-control resource depletion

32. Aim: Examine effect of smoking cravings on ego-depletion in smokersHypothesis: Smokers exposed to smoking cues will exhibit poorer performance on self-control tasks compared to when they are exposed to neutral cuesTwo experimentsExpt. 1: Pictorial cuesExpt. 2: Cue-exposure taskEffects will be independent of trait self-controlEgo-Depletion and Smoking

33. Design and MethodParticipants. Regular smokers smoking >10 cigarettes per day. Expt. 1: N = 19, 12 males, 7 females, M age = 23.47, SE = 0.69; Expt. 2: (N = 32, 13 males, 19 females; M age = 20.13, SE = .25Design. 2 (cue exposure: cigarette vs. neutral) x 2 (order of presentation: cigarette first vs. neutral first) mixed-model designCue exposure. Expt. 1: Smoking related images vs. neutral images; Expt. 2: Cigarette cue-exposure task vs. drinking straw task

34. Cue-exposure task(cigarette version)Rating:1.2.3.4.5.Rating:1.2.3.4.5.Rating:1.2.3.4.5.1.2.3.

35. Cue-exposure task(drinking straw version)Rating:1.2.3.4.5.Rating:1.2.3.4.5.Rating:1.2.3.4.5.1.2.3.

36. Design and MethodDependent measure. Baseline-adjusted handgrip duration (Expts. 1 & 2) and incongruent Stroop task response latency (Expt. 2 only)Procedure. Baseline handgrip measurementSmoking or neutral cuesDependent handgrip taskRepeat with cue set not received in 2Additional measures. Trait self-control (Tangney et al. (2004) and urge to smoke (QSU-Brief; Cox et al., 2001).

37. Ego-Depletion and Smoking: Results for Handgrip TaskHandgrip task durationNeutral CueSmoking CueExperiment 1Experiment 2Source: Hagger, M. S., Leung, C. M., Leaver, E., Esser, K., te Pas, N., Keatley, D. A., . . . Chatzisarantis, N. L. D. (2013). Cue-induced smoking urges deplete cigarette smokers’ self-control resources. Annals of behavioural Medicine, 46, 394-400.

38. Ego-Depletion and Smoking: Results for Stroop TaskStroop Response LatencyNeutral CueSmoking CueExperiment 2Source: Hagger, M. S., Leung, C. M., Leaver, E., Esser, K., te Pas, N., Keatley, D. A., . . . Chatzisarantis, N. L. D. (2013). Cue-induced smoking urges deplete cigarette smokers’ self-control resources. Annals of behavioural Medicine, 46, 394-400.

39. Summary and ImplicationsSmoking cravings deplete self-control resources Independent of trait self-control and order of presentationSmokers attempting to quit might have problems with self-control in other domains e.g. eating restraintImportant to manage exposure to temptations among quitting smokersInterventions to boost self-control might help

40.

41. Is the Depletion Effect ‘Real’?Carter & McCullough, 2015, Journal of Experimental Psychology: GeneralCarter & McCullough, 2014, Frontiers in Psychology

42. Is the Depletion Effect ‘Real’?Recent re-analyses of our meta-analysis and new meta-analyses have suggested that:“Meta-analytic evidence does not support self-control …[as] a limited resource” (Carter et al., 2015)Small study/publication bias based on ‘funnel’ plotsEffect sizes computed using:Carter et al. (2015) incl. unpublished data: g = .43Duvall & Tweedie (2000) ‘trim and fill’ method: g = 0.24Egger et al. (1997) PET: g = 0.27 or PEESE = g = 0.003

43. Ego-depletion effect sizes.24.003.26.43.62Carter et al. (2015) ‘trim andfill’Carter et al. (2015) incl.unpublishedHagger et al. (2010) meta-analysisCarter et al. (2015) PEESECarter et al. (2015) ‘top 10’.27Carter et al. (2015) PET.17Tuk et al. (2015) large-scale study

44. “…estimates of the true [ego-depletion] effect fluctuate, and do not converge on a reasonable range of estimates; it might be meaningfully different from zero, or it might not. Inzlicht et al. (2015)What is really needed is a high-powered and good-faith attempt to replicate the basic effect with large samples to estimate the true size of the ego depletion effect.”Ego-depletion effect size estimates

45. Multi-Lab Registered Replication ReportsRRR inspired by ‘replication crisis’ in psychology (Open Science Collaboration, 2015, Science)Initiative of Perspectives on Psychological ScienceAims to provide substantiation of effects in multiple labs with standardized methodsMethods approved by original authorsOutcome is a ‘meta-analysis’ of the effects examining effect size & confidence intervals

46. Ego-Depletion RRRRRR uses methods adopted by Sripada et al. (2014)Used fully computerized tasksStandardization of tasks for replicationDeveloped with and approved by Sripada, Kessler and Baumeister

47. Ego-Depletion RRROne-way between-participants design (depletion vs. no-depletion) sequential-task paradigmDepleting task: Computerized version of letter ‘e’ task (lonely ‘e’ vs. no lonely ‘e’ versions)Dependent task: Multi-source interference task (MSIT) (Bush & Shin, 2006)Administered by E-PrimeStandardized protocol & materials made available

48. Ego-Depletion RRRHypothesis: Difference across experimental groups on primary DVDependent variables:PRIMARY: Reaction-time variability for MSITSECONDARY: Overall reaction time for MSITAdditional variables:Letter ‘e’ task performance (error rates)Effort, difficulty, fatigue, frustrationUG/PG student participant poolsLabs pre-register with PPS and OSF

49. Depletion RRR24 labs registered and completedLabs independently conducted replication, analysed and submitted results for meta-analysisResults and log of any deviations posted on OSF

50. Source: Hagger, M. S., Chatzisarantis, N. L. D., Alberts, H., Angonno, C. O., Batailler, C., Birt, A., . . . Zwienenberg, M. (2016). A multi-lab pre-registered replication of the ego-depletion effect. Perspectives on Psychological Science, 11, 546-573. doi: 10.1177/1745691616652873

51. ResultsMean RTV: d = 0.04, CI95 = -0.07; 0.14, I2 = 36.08%, Q = 34.42 (p = .045)

52. ResultsLetter E task performance: d = -1.82, CI95 = -1.98; -1.67, I2 = 56.57%, Q = 50.65 (p = .001)

53. Ego-depletion effect sizes.24.003.26.43.62Carter et al. (2015) ‘trim andfill’Carter et al. (2015) incl.unpublishedHagger et al. (2010) meta-analysisCarter et al. (2015) PEESECarter et al. (2015) ‘top 10’.27Carter et al. (2015) PET.38Dang (2017) Updated meta-analysis.04Hagger et al. (2016) RRR

54.

55. Lesson 1: Not everyone’s going to be happy or accept the findings(Expect a lot of attention!)Ego-depletion replication – Some lessons learned

56. Reactions

57. Reaction“Foolish”“Debacle”“An unfortunate mistake”

58. Reaction“Caution is required in drawing implications from the results of this RRR for the phenomenon of ego depletion writ large”

59.

60. Lesson 2: Expect a lot of questions/re-analyses/re-interpretationEgo-depletion replication – Some lessons learned

61. Catalysis of New ResearchReanalysesSchimmack (April, 2016): R-Index and PowergraphBlázquez et al. (2017): p-Uniform methodWitte & Zenker (2017): Research program strategyAdditional analysesDang (2016): Task effort as moderatorDrummond & Philipp (2017): Effort as mediator of ego-depletionConceptual reviews/commentsBaumeister & Vohs (2016); Sripada et al. (2016)Hagger et al. (2016)Lurquin & Miyake (2017)Lee et al. (2016)

62. Lesson 3: This isn’t the end of it!(Pave the way for additional replications and data to contribute to further debate)Ego-depletion replication – Some lessons learned

63. Vadillo, M. A., Gold, N., & Osman, M. (2018). Searching for the bottom of theego well: Failure to uncover ego depletion in Many Labs 3.Royal Society Open Science, 5. doi: 10.1098/rsos.180390New replications: “Many Labs 3”

64. Ego-depletion effect size estimates.24.003.26.43.62Carter et al. (2015) ‘trim andfill’Carter et al. (2015) incl.unpublishedHagger et al. (2010) meta-analysisCarter et al. (2015) PEESECarter et al. (2015) ‘top 10’.27Carter et al. (2015) PET.04.007Vadillo et al. (2018) replicationHagger et al. (2016) RRR

65. Dang, J. (2017). An updated meta-analysis of the ego depletion effect. Psychological Research. doi: 10.1007/s00426-017-0862-xNew meta-analysis

66. Ego-depletion effect size estimates.24.003.26.43.62Carter et al. (2015) ‘trim andfill’Carter et al. (2015) incl.unpublishedHagger et al. (2010) meta-analysisCarter et al. (2015) PEESECarter et al. (2015) ‘top 10’.27Carter et al. (2015) PET.38Dang (2017) Updated meta-analysis.04Hagger et al. (2016) RRR.007Vadillo et al. (2018) replication

67. Garrison, K., Finley, A., & Schmeichel, B. J. (2018). Ego depletion reduces attention control: Evidence from two high-powered preregistered experiments. Personality and Social Psychology Bulletin. doi: 10.17605/OSF.IO/PGNY3

68. Ego-depletion effect size estimates.24.003.26.43.62Carter et al. (2015) ‘trim andfill’Carter et al. (2015) incl.unpublishedHagger et al. (2010) meta-analysisCarter et al. (2015) PEESECarter et al. (2015) ‘top 10’.27Carter et al. (2015) PET.04Hagger et al. (2016) RRR.007Vadillo et al. (2018) replication.38Dang (2017) Updated meta-analysis.20Garrison et al.(2018)

69. Dang, J. H., Liu, Y., Liu, X. P., & Mao, L. H. (2017). The ego could be depleted, providing initial exertion is depleting: A preregistered experiment of the ego depletion effect. Social Psychology, 48, 242-245. doi: 10.1027/1864-9335/a000308Lurquin, J. H., Michaelson, L. E., Barker, J. E., Gustavson, D. E., von Bastian, C. C., Carruth, N. P., & Miyake, A. (2016). No evidence of the ego-depletion effect across task characteristics and individual differences: A pre-registered study. PLoS ONE, 11, e0147770. doi: 10.1371/journal.pone.0147770

70. Lurquin et al.(2016)Ego-depletion effect size estimates.24.003.26.43.62Carter et al. (2015) ‘trim andfill’Carter et al. (2015) incl.unpublishedHagger et al. (2010) meta-analysisCarter et al. (2015) PEESECarter et al. (2015) ‘top 10’.27Carter et al. (2015) PET.04Hagger et al. (2016) RRR.007Vadillo et al. (2018) replication.38Dang (2017) Updated meta-analysis.20.48Dang et al. (2017)pre-registeredGarrisonet al.(2018).22

71. Baumeister, R. F., & Vohs, K. D. (2016). Misguided effort with elusive implications. Perspectives on Psychological Science, 11, 574-575. doi: 10.1177/1745691616652878

72. New replication:Vohs, Baumeister, & SchmeichelResults. 26 labs = directionally consistent effect, 11 labs = directionally opposite effect.Bayesian analysis. Priors, d = .30, SD = .15, truncated at zero. Bayesian factors suggest data are 4 times more likely under null than pre-registered hypothesis.Bayesian meta-analysis. d = .08Conclusion. “If the depletion hypothesis is nonetheless true, the effect is relatively small” – Wagenmakers; “There is a real (non-zero) but small effect” - Funder

73. Where next?Explore task parameters. Task type, duration, effort on initial taskNew designs. Within-participants designs?Unification. How does vigilance and mental fatigue fit with ego-depletion – hour-long tasks?Capitalize on context and ‘real world’ applications. Context and individual differences allow opportunity to study depletion in real contexts – e.g. diet, smoking

74. Summary and ImplicationsTrait self-control is linked to health behaviour, with beliefs as a mediatorTrait self-control may moderate intention-behaviour relationship in some behavioursLimited self-control resources implicated in health behaviour (e.g., dieting, smoking)Meta-analyses of ego-depletion effect indicate substantial range (d =.003 to .62)Two registered replication report indicates that zero is a probable value for ego-depletionFuture research should consider ‘real world’ applications e.g. prejudice, health behaviours

75. Thank yoUmartin.hagger@curtin.edu.auwww.healthbehaviourchange.comwww.martinhagger.com