Lec 3 Flashcards
Tracking kiwis from birth to adulthood
- What does SC predict
Methods
- Followed 1K kids from NZ for 30+yr
- Measured SC at ages, 3, 5, 7, 9, & 11
- Examined their life outcomes in health, wealth, and crime
Results: SC predicts: Health, Wealth, & Public Safety
- High SC kids → as adult, they hv better wealth and health and low crime, less likely tb single parent
Dunedin SC study vs marshmallow study
- Critique of Dunedin SC study
Dunedin SC study
- Mischel’s marshmallow study: SC in childhood predicts good life
- Dunedin study: better b/c it shows effects still exist after controlling for SES and IQ
- Critique:
- Is this SC or impulsivity?
- IOW: is it due to mild desire OR strong SC
- Is this SC or impulsivity?
The resource model of self-control
- 3 claims of resource model
- 1 Central resource
- 2 Limited resource
- sequential-task paradigm
- 3 SC can be improved (~ to muscle)
Evidence – SC exhaustible resource
- Muraven et al 1998
- Methods (IV, DV)
- Results
- Vohs & Heatherton, 2000
- Methods & deception
- Results
- Inzilcht & Gutsell 2007
- Methods
- Results
- Vohs et al 2008 - Decide grad courses
- H
- IV, DV
- Control vs manipulation gp
- Results
- Issue
- Gino et al, 2011
- H
- IV, DV
- Results
- Issue
*
Resource Model should be celebrated: SC rs took off in 90s b/c Baumeister’s Resource model
Claim 1: SC is a central resource
- Willpower/SC powers many things (ex. exercise, empathy, diet, anger, racism, stroop task, academics, delay gratification, marriage fidelity)
- Empathy is effortful: thinking in your perspective is easy, thinking what others are thinking is effortful
Claim 2: SC is an exhaustible/limited resource
- If I exert SC at Time 1 (not get angry at kids)
- If I exert SC at Time 2 (write assignment)
- You might not be able to exert SC in Time 2 due to low fuel
- Rs use sequential-task paradigm to study this claim
- sequential-task paradigm: Ppl do SC task at time 1; when they do SC task at time 2, they fail to show SC
- Rs ASSUME this is due to no energy in body
Claim 3: SC is like a muscle that can be improved w/ exercise
- Practicing non-habitual tasks can increase general self-control
- Even on tasks unrelated to practice domain
- Ex. regulate emo while watching at movie at Time 1 reduce performance on subsequent tasks (ex. solve hard puzzle, squeeze exercise handgrip, WM, stroop task, etc)
Evidence – SC exhaustible resource
Muraven et al 1998
- Ppl watch documentaries (gross)
- Conditions
- A: ppl need to controlling and not express
- B: control group
- Then, ppl have to push a hand grip – measure SC and willpower
- Results: those who restrain themselves at Time 1 (watching TV), they held onto the handgrip for less time than the control gp
- This supports the idea the SC is a central resource and can be depleted
Vohs & Heatherton, 2000
- Same design
- Except ppl were told to eat icecream and rate the flavors
- Rs: they are actually measuring how much ice cream ppl consumed
- Results: those who were told to restrain their emotions while watching the gross doc ate more icecream
Inzilcht & Gutsell 2007
- Restraining emotions while watching gross doc (Mondo Cane) → deplete resources → took more time on stroop task
Vohs et al 2008 – Making decisions depletes control
- Hypothesis: Making a series of decisions will deplete SC
- IV: Ppl make a series of decisions on grad courses
- DV: how long will ppl practice on an upcoming test (procrastination)
- the less they practice indicates they have less SC
- Control gp: didn’t make decisions/ assigned to courses
- Manipulation gp: made decisions
- Results: those who made more decisions did not study as much for the test (aka low SC)
- ISSUE: N = 24! → This suggests p hacking is used
Gino et al, 2011 – Depletion leads to unethical b
- Hypothesis: Controlling thoughts will increase cheating
- IV: Ppl watch a video and told to ignore words appearing on screen
- Then ppl were given math test they self score on a computer
- They are given money for each item they get correct
- DV: % of students who over-reported how much money they are owed on math task
- Since you can cross check the actual score and reported score on the computer, you can see how much they cheated on/overreported on correct answers
- Results:
- Everyone overreported
- The depleted ppl overreported more
- Issue: they do not use error bars to report results
Israeli judge study
- results
- TB interpretation of results
- 2 Issues of the study (mentioned in lec)
Muraven et al., 1999
- H
- IV, DV
- Reported results
- 3 Issues
Israeli judge study
- Results:
- if you have your hearing early on and after lunch → more likely to have parole (65%)
- If you have their hearing before lunch → less likely to have parole (0%)
- TB: low glucose → judges depleted and rely on heuristic/ safe option = no parole
- Issue:
- 1 order of hearings is non-random, the “Easiest” cases last
- 2 reported effect size is very high & sketchy (d=1.96)
- NOTE: Men taller than women (d=2.0)
Muraven et al., 1999 Practicing self-control
- Hypothesis: Practicing self-control for 2 weeks will improve self-control (in general)
- IV: Participants monitor their posture or moods, or keep food diary
- All of these take effort
- DV: Susceptibility to depletion before and after intervention
- (how long to hold handgrip after thought suppression)
- Results: typically depletion effect
- B4 depletion intervention, they hold on it for more time
- After depletion → hold on handgrip for less time
- Issues
- 1 Results have no sig effect
- 2 Woefully underpowered, n=69= 13 Ss per cell!
- 3 doesn’t make sense that watching your posture can have a dramatic effect (aka help resist temptations)
- This could by p hacking and a fluke
Theoretical problems w/ Resource model
- What is the resource in resource depletion?
- Issue w/ resource model
- Theoretical soup stone
- Galliott et al study on SC & glucose
- 3 claims
- 5 issues
What is the resource in resource depletion?
- Baumeister – thinks it’s glucose
- Controlling oneself at Time 1 affects performance at Time 2
- Maybe due to limited resource, ex. glucose
- IOW: SC exertion at time 1 depletes levels of brain glucose
- Issue: not replicable & theoretically/stat impossible
- Resource model = theoretical soup stone
- Theoretical soup stones: concepts that seem essential to some process, but upon closer inspection are unnecessary and do no theoretical work.
Galliott et al: 9 studies on blood glucose
- 3 claims
- SC consumes glucose
- Drops in glucose mediate ego depletion
- Glucose supplementation (e.g., sugary drink) can increase self-control & reverse depletion
Problems
- Theoretically improbable
- Control consumes little brain glucose (i.e. half a tic tac for thinking)
- When looking at PET scans that can measure glucose consumption, brain uses very little glucose when thinking
- Statistically unlikely
- Given low power, 9 significant effects unlikely, “incredible” even = too good tb true
- Statistically irreproducible
- Failures to replicate
- Merely swishing glucose (but not consuming it) increases control
Theoretical problems w/ Resource model (cont)
- 1 motivational incentives
- Muruvan study w/ money
- 2 subjective perceptions
- Clarkson et al
- 3 Lay beliefs
- Job’s study
- 4 Personal Prayer
1 Challenges to the resource model – motivational incentives
- Muraven follow up study
- You do a task at time 1 → ego depleted → at time 2, you are given cash to do task 2 → ego depletion effect disappears
- If a resource is depleted, how can money make people not depleted?
- Analogy: If the gas is really empty and I’m really motivated to get home, putting money on the car won’t make the car move
2 Challenges to the resource model – subjective perceptions
- Clarkson et al
- Do task at time 1 → ppl are depleted
- Ppl did a questionnaire, rs make them believe they have a lot of energy → when doing SC task at time 2, ego depletion is gone
- Those who were not depleted but were told they were fatigued → showed SC failure
3 Challenges to the resource model – Lay beliefs
- Job’s study
- Looked at ppl w/ diff SC beliefs
- Some ppl think SC is limited – need to save your energy for later things
- Some ppl think SC is unlimited, the more you practice it, the more you have it
- Results: If you believe SC is renewable, you won’t show SC deficit in time 2
- This is inconsistent with a resource account b/c if people really failed at Time 2 because they had no resources, you will be depleted regardless of your beliefs
4 Challenges to the Resource Model - Personal Prayer
- If ppl believed in God and prayed, the SC depletion at Time 2 disappeared
The Why of Ego Depletion - Balance between labour vs. leisure
- 4 reasons
- 1 lack motivation for working
- 2 Law of less work
- 3 value of work vs leisure
- 4 motivated slacking
- Overall
Inzlicht et al 2014 - The How of Ego Depletion
- Preliminary process model/ shift in motivation
The Why of Ego Depletion - Balance between labour vs. leisure
- 1 ppl lack motivation not energy
- 2 “Law of less work;” people avoid cognitive labour to seek leisure
- People sometimes forego money to avoid labor
- 3 After working, the value of more work decreases, value of leisure increases
- If I worked towards smth for a while, I get some benefit/value
- After a while, I don’t value those goods as much as I used to
- Ex. If I value money or food, once I have enough, I don’t need anymore
- I can do smth I relax and enjoy
- 4 Depletion as self-justified indulgence, motivated slacking
- “I’ve worked hard, now I deserve a break”
- In general: work hard at time 1 → ppl think “after working hard, I deserve a break” → show less control at Time 2, prefer leisure
Inzlicht et al 2014 - The How of Ego Depletion
- Shifts in Motivation/ preliminary process model
- Reduced motivation for “have to” or “ought to” or “should” goals (labor)
- Increased motivation for “want to” goals (leisure)
Motivational shift I: I don’t feel like regulating myself
- Murray study
- other study - self-justified indulgence
- Methods
- Results
Caution: Shifts in Motivation ≠ Less Motivation; explain ex.
Motivational shift I: I don’t feel like regulating myself
- Murray study
- When you give ppl money, you increase ppl’s motivation to exert effort
- Other study - self-justified indulgence
- 1 Hv ppl exert effort
- 2 rate how important their goals are
- Results: after ppl exert effort/ are depleted, they see their goals as less urgent/important/ necessary
Caution: Shifts in Motivation ≠ Less Motivation
- After doing work, you’re tired
- You want candy (motivated) → You would walk for 10 minutes to get candy from the shop
- IOW: more motivated to walk all the way to get candy/icecream, less motivated to do work
- This shift in motivation DOES NOT MEAN less motivation overall.
- Shift of motivation explains how ppl how more energy to do leisure things after work
Motivation Shift II: I want gratification
- Concept
- Wagner & Heatherton
Beyond the resource model
- 3 issues w/ model
Motivation Shift II: I want gratification
- After you exert SC in Time 1 → Desire for junk food increases + Center for rewards are more sensitive
- Wagner & Heatherton
- previous acts of control increase activity in the L OBFC (area associated w/ rewards & desiring smth)
Beyond the resource model
- 1 data does not support resource model: it doesn’t make sense that motivation can reverse depletion when resource is empty
- 2 Depletion is similar to fatigue; it can be reversed by motivation
- 3 Depletion reflects shifts in priorities
- Ex when exhausted OR when 2 goals were in conflict— impulsive goals vs LT goals requiring SC, less motivation for have to/work goals; more motivation for want to/leisure goals
The replication crisis & ego depletion - Problem 1
- Publication bias affects meta-analyses
- Funnel plots
- Use
- How it is plotted
- How to spot publication bias
Funnel Plot – Ego depletion
- 2 issues
- 3 techniques to adjust for biases & results
- Main point
Problem 3 - The replication crisis & ego depletion
- McCullough study
- Lurquin et al, 2016 - depletion and WM
- H
- IV, DV
- Results
- 3 reasons why this is a strong study
Problem 3 - Massive, coordinated replication failure
- Hagger et al - registered replication report (RRR)
- Methods
- Results
- Conclusion
Problem #1 problems in meta-analyses
- At 2010: ego depletion = med to large effect but it has issues
- Publication Bias: Published record ≠ Actual record
- affects meta-analysis b/c it creates inflated effect sizes
- funnel plots: used to spot biases in systematic reviews and meta-analyses
- Studies w/ largest samples are plotted near the avg/center
- studies w/ smaller samples will be spread on both sides of the avg
- This creates a funnel shape
- Deviation from this shape can indicate publication bias.
Funnel Plot – Ego depletion
- 1 The effect sizes are so big → doubtful
- 2 the funnel plot is asymmetrical, it is a linear plot
- 3 techniques to adjust for biases
- Trim & Fill; suggest the effect is a small +ve effect
- PEESE; suggest effect includes 0
- PET; suggest the effect is -ve
- Point: bias corrections cannot determine if depletion is real; inconclusive
- Garbage in garbage out effect w/ meta analysis
- 3 techniques to adjust for biases
Problem 2: indiv replication failures
- Replication failures are hard to publish
- McCullough study: after doing a bias-corrected meta-analysis, ego depletion effects are non-existent or super small.
- Lurquin et al, 2016 - depletion and WM
- H: Controlling thoughts will impair subsequent WM
- IV: Participants watch a video and told to ignore words appearing on screen at time 1
- DV: Working memory performance at time 2
- Results: no significance in WM performance; fail to replicate
- Strong study
- 1 Large samples, n=200
- 2 Hypotheses pre-registered in advance
- 3 Confirmatory and immune from p-hacking
Problem 3 - Massive, coordinated replication failure
- Hagger et al - registered replication report (RRR)
- 1: 23 labs w/ 2400 participants tried to replicate Sripada et al 2014
- 2 Pre-registered hypotheses
- 3: Method approved by Baumeister and he advised
- SC exertion Time 1:
- 1 Ppl watched words flashed on screen
- 2 Ppl hit a key if the word has letter e that was not w/in 2 spaces of another vowel - create ego depletion effect
- SC failure time 2: do a task at Time 2
- SC exertion Time 1:
- Results: the effect overlaps 0
- only 2 labs found +ve effect; 1 lab found -ve effect
- Baumeister was salty
- Conclusion: we don’t know if ego depletion is real, and when or why it happens
Pregistered replication report (RRR)
- Hagger RRR study results
- 3 critiques
- Dai et al 2014
- Methods
- Results
- Issue
Pregistered replication report (RRR)
- Hagger RRR was very influential (most cited in 2016) - showed the ego depletion is even smaller
- Critiques
- 1 Only replicated one study
- 2 The depleting task was not strong enough
- counter argument: most depleting tasks are weak too
- 3 The DV did not involve breaking a habit, some ppl think this is essential
- This DV has been used successfully before
Dai et al 2014
- Methods: Tracked down caregivers and whether they sanitized their hands
- Results: compliance rate decreases over the shift but recovers after the breaks
- Issue: not pre-registered study
Endgar article
Studies
- Baumeister and Tice study
- Methods
- Results
- Conclusion
- Hagger 2010
- Baumeister and Tierney - exercise
- Journal Perspectives on Psych science study
- reproducibility crisis study
- Carter 2007 study - replicate lemonade effect
- lemonade/glucose effect
- Results from 2 other labs
- Carter 2010 study - meta-analysis
- 2 issues
- reanalysis results
Baumeister and Tice
- 1 Put cookies next to radishes
- 2 Have students in the lab
- Some can only eat radishes
- Some can only eat cookies
- Control gp
- 3 All did a puzzle that was impossible to complete; rs timed how long it took them to give up
- Results
- cookie ppl OR control condition ppl worked on the puzzle for 20 min
- radishes ppl lasted for 10 min only
- Conclusion: ego depletion -
- 1 We have a limited supply of willpower
- 2 Willpower uses mental energy
- 3 Willpower is like a muscle, and it can get exhausted
- This is replicated
- Hagger 2010: Meta-analysis confirmed ego depletion is real
- Baumeister and Tierney: exercise boost willpower/ SC
- Journal Perspectives on Psych science study:
- Methods: recruited 2000 subjects across 12 labs
- Results: Study found zero-effect for ego depletion
- reproducibility crisis study
- 40/100 psych experiments were replicated
- Why? original authors made stat errors (aka bad rs)
- Carter 2007: can’t reproduce the lemonade effect (drinking sugary drinks can replenish SC)
- Other labs produced the lemonade effect w/o consuming lemonade (ex. swishing then spitting out lemonade, not drinking; alter beliefs)
- Carter 2010 - Meta-analysis
- Issue 1: Meta-analysis only used published studies - biased
- Issue 2: included studies w/ contradictory measures of SC
- Ex. one study suggested depleted subjects will give more money charity; other study said ppl will spend less time to help stranger
- reanalysis = no effect
Q1: Does SC really wane over time?
- SC effect - after bias correction
- 3 Issues w/ statistics on bias correction
- Tuk et al 2015
- Methods
- Results
Q2: Is ego depletion a form of mental fatigue?
- Define mental fatigue
- Similarities b/s mental fatigue and ego depletion
- mental fatigue vs ego depletion
- Ego depletion vs sleep deprivation
Problems w/ bias-correction estimates
- publication bias effect seen can also be caused by running a power analysis
- 5 studies were compared; that estimators that correct for publication bias sometimes fare worse than those that do not correct
- no consensus among statisticians on such estimators
Tuk et al 2015
- Meta-analysis of 9 ego depletion studies w/ 580 ppl
- No data selection and publication bias
- Results: small, homogeneous, real effect
- The effect slightly overlap 0 range
Q2: Is ego depletion a form of mental fatigue?
- mental fatigue = -ve feeling after long periods of cog activity
- low arousal, -ve mood (ex. tiredness, weariness exhaustion) + distracted/bored
- Decline in cog and b performance
- Mental fatigue and ego depletion – similar
- once you apply effort → -ve feelings (ex. fatigue)
- lead to worse performance
- can be overcome by task motivation
- mental fatigues vs ego depletion:
- Mental fatigue happens after long periods
- Depletion happens after a few min
- Ego depletion vs sleep deprivation
- diff body behavior: ego deprivation can be reversed by motivation; sleep deprivation cannot (IOW: you still wanna sleep)
Q3: What is the resource that is depleted by Ego Depletion?
- maybe glucose
- alt theory
Q4: How can changes in motivation and expectations replenish a depleted resource?
- 2 ways
- 1 rewards
- 2 beliefs
- 3 types of beliefs
- Conclusion
- Resource theorists rebuttal
- Motivation theorists rebuttal
*
Q3: What is the resource that is depleted by Ego Depletion?
- Researchers infer resource depletion b/c performance declines in sequential task paradigm
- Some suggest low glucose may cause depletion but this can’t be replicated
- Can’t be replicated b/c original studies hv low stat power
- Brain glucose is stable and plentiful regardless of mental activity
- alternative theory: ingest glucose → increase motivation → improve SC
Q4: How can changes in motivation and expectations replenish a depleted resource?
- Rewarding SC cancel’s SC refractory period
* offering incentives, reframing temptations as tests of willpower removes depletion
* Ex. Smoking, watch fav TV, receive surprise gift
- Rewarding SC cancel’s SC refractory period
- Change perceptions prevents depletion
* 1 Belief of self: If ppl believe they have stamina → show SC; vc
* 2 Belief on SC: if ppl believe SC is renewable → no drop in SC; vv
* 3 Perception of effort: If ppl see they same task is fun → no SC drop; vv
* These findings don’t align w/ the resource account (ex. if car has no gas, putting money on the car won’t make it move)
* IOW: ego depletion is explained by motivation not resource
* Resource theorists rebuttal: this only happens when there is a little depletion
* Motivation theorists rebuttal: Severe fatigue (ex. 2 hr cycling) can be overcome by motivations
- Change perceptions prevents depletion
Q5: Has the resource model unwittingly become a model about motivation?
- Revised Resource theory
- Motivation theorists rebuttal
Q6: Do SC exercises increase SC?
- Resource model POV
- Results from other studies
- EF study
- Conclusion
- small acts of SC study
- Result
- meta-analysis that corrected for bias
- Result
- 3 ways SC training may improve SC
*
- EF study
Q5: Has the resource model unwittingly become a model about motivation?
- Resource theorists think resources are only partially depleted; this way they can include motivation theories in their model
- Motivation theorists: we just need the motivation theory
- SC decreases b/c the task are not interesting → apply less effort
Q6: Do SC exercises increase SC?
- Resource model suggest that SC exercises increase SC
- Studies
- computer games that uses WM and inhibitory control does not generalize
- SC relies on a facet of EF (ie inhibitory control)
- If EF training fail to generalize, SC training may not generalize
- Other studies 2: If ppl do small acts of SC for 2 weeks, results are inconsistent
- SC Meta-analysis that corrected for publication bias: SC training may not generalize
- SC training can improve SC by
- 1 increasing SC capacity
- 2 by changing one’s beliefs on willpower → lower aversiveness of SC
- 3 make SC a habit and effortless
- SC training can improve SC by