Upload
charity-skinner
View
212
Download
0
Tags:
Embed Size (px)
Citation preview
Motivation and Cognition
Art Markman
Special thanks to:
C. Miguel BrendlKyungil KimClaude MessnerTodd MaddoxGrant Baldwin
Issues
• Motivation and Cognition
• Why do we care about motivation?
• Components of the motivational system– Goals– Energy (motivation) to act
• Why hasn’t motivation been a big part of the study of cognition?
• Why is it coming back?
What is a goal?
• This has proven to be a hard question to answer– Much of the motivational system is not accessible to
consciousness
• What I’d like to do in this talk– Review 9 phenomena that constrain a theory of goals– Talk about some studies we’re doing
9 Phenomena• People can talk about the reasons for their actions
• Talking about actions can cause interference
• Difficulty predicting future preferences
• Expressed attitudes need not coincide with action
• Affective states have a loose correlation with motivational states
• The world can prime goals
• Goals prime means for achieving goals
• Means remind people of goals
• Intentions to perform actions influence behavior
People can talk about reasons
• Sometimes they are right
• People who express protected values often act in accordance with them– If you say you will not buy a desk with
rainforest wood, often you will not
• Some choices are made, because they can be justified– Shafir and Tversky: Trip to Hawaii after exam– Attraction effect gets stronger when a reason
must be given
Talking about reasons can hurt
• Not everything can be talked about
• Wilson et al.– People are less satisfied with their choice of a poster
when they justify their choice than when they don’t
• Berridge– Addicts will press a button more for a solution
containing a low dosage of their drug, but they cannot explicitly identify which solutions contained the drug
Mis-prediction of future preferences• When a future choice will take place in a
different context, people’s predictions are bad– Early devaluation studies with smokers
• Brendl, Markman, & Messner
– Kahneman & Snell: People cannot predict the amount of ice cream they would want the future
– Read et al.: People cannot predict their future movie choices
• Pick a highbrow film for later
• Pick a comedy for now
– Simonson• Greater variety seeking when picking for the future
Attitude Behavior (in)consistency
• People’s behavior need not coincide with their expressed attitude
• Nisbett & Wilson; Kruglanski– People often pick the right-most item in a display– Justify with a discussion of quality of the item– Strength of the effect mediated by Need for Closure– Attitudes are affected by active goals
• If the attitude is elicited under different goal conditions than is the behavior, the two do not correspond.
Affective States & Motivation
• We use affective states as a guide to active goals
• They are not necessarily a good marker– Hunger and eating– Craving and need to use a drug (Tiffany)– We were able to dissociate need to eat from hunger
• Brendl, Markman & Messner, popcorn study
Contextual Priming of Goals
• Zeigarnik Effect– People remember unfinished actions better than
finished ones (also Patalano & Seifert)– Ovsiankina: Seeing a mailbox primes the
action to mail a letter
Priming among goals and means• Lexical decision tasks
– Words for desired states prime words for actions related to them (Kruglanski et al.)
• Implementation intentions (Gollwitzer)– People must learn to associate particular means
with goals– Forming specific intentions for satisfying a goal
increases success at carrying out goal– Wertenbroch: People arrange their worlds to
take this priming into account• Smokers trying to quit will stop buying cartons
Studies in my lab on Motivation
• A number of ongoing projects– Influence of goals on preferences
• Structure of goals
– Influence of motivation on task performance• Regulatory fit and flexibility.
What is a Goal?
• This pattern can be used to determine what a goal is.
• We can use the pattern of valuation and devaluation to determine what is inside and outside the goal– What is valued is inside the goal– What is devalued is outside the goal
Design• Need to eat
– High: Small amount of bread with salted butter– Low: Large amount of bread with unsalted butter– Had to use male participants
• Females at UT would not eat prescribed amount
• Time of day– 9am or 4pm
• Item type– Breakfast foods– Dinner Foods– Nonfoods
Hypotheses• Preference ratings
• Are Goals General?– Need to eat means need to eat food– Breakfast and dinner foods should show valuation– Nonfoods should show devaluation
• Are Goals Specific?– Need to eat is specific to particular meal– Interaction between time of day and need
• Breakfast foods valued in the morning
• Dinner foods valued in the evening
Results of Goal Specificity Study
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
BreakfastFoods
Dinner Foods Nonfoods
Item Type
MorningEvening
* *
* *
Framework
• Approach goals– Potential positive states of the environment
• Avoidance goals– Potential negative states of the environment
• Regulatory focus theory (Higgins)– Promotion Focus
• Sensitivity to gains/nongains
– Prevention Focus• Sensitivity to losses/nonlosses
Regulatory Fit
• Tasks have a reward structure
• Typical structure– Reward for correct answers– No points for incorrect answers– Mild promotion focus
• Performance bonus
• Social approval
– This is a regulatory fit
Consider the bigger picture
• Almost all cognitive research is in one cell– Does this matter?
• As Cognitive Psychologists, we believe we are learning general facts about cognition
Reward Structure of Task
Gains LossesPromotion Focus Fit MismatchPrevention Focus Mismatch Fit
Studying Regulatory Fit
• How can we study this systematically?
• Need a good manipulation of regulatory focus
• Need a task for which we can assess performance
• Need to be able to manipulate reward structure
Perceptual Classification
• Classification of simple perceptual stimuli– Small number of underlying dimensions– Easy to vary category structure– Extensive set of tools for modeling
performance of individual participants– Can assess the strategy they use in the task
• Hypothesis– Regulatory fit promotes cognitive flexibility
General PredictionsSituational Incentive Focus
Reward Task Type Structure Promotion Prevention
Rule-based: Flexibility Gains Better Worse
Rule-based: Flexibility Losses Worse Better
Rule-based: No flexibility Gains Worse Better
Rule-based: No flexibility Losses Better Worse
Two Studies
• Gains condition: Positive payoffs
• Loss condition: Losses on each trial
• Promotion focus: Entry into drawing for exceeding criterion
• Prevention focus: Get entry at start, keep entry if criterion is exceeded
• Performance should be best in this task when there is a fit– Complex conjunctive rule should be found
earlier when there is a fit
For gains
0.70
0.80
0.90
1.00
1 2 3 4 5 6 7 8 9 10 11 12
Block (48 trials/block)
Promotion Prevention
Modeling for Gains
0.00
0.20
0.40
0.60
0.80
1 2 3 4 5 6 7 8 9 10 11 12Block (48 trials/block)
Promotion Prevention
Losses
0.70
0.80
0.90
1.00
1 2 3 4 5 6 7 8 9 10 11 12
Block (48 trials/block)
Promotion Prevention
Modeling-Losses
0.00
0.20
0.40
0.60
0.80
Block (48 trials/block)
Promotion Prevention
Conclusions
• Regulatory fit supports flexibility
• Not a main effect of regulatory focus
• Most cognitive studies operate under conditions of regulatory fit
• Many phenomena may reflect fit/mismatch– Cognitive deficits in disorders– Effects of pressure on performance
Summary• For many years, theories of goals were
based on intuitions guided by little data– Lewin; Carver & Scheier
• Research is acknowledging that key aspects of goals are not consciously accessible– Experimental methods reflect this
• Data are permitting a new generation of theories of the cognitive structure of goals
• New work focuses on influence of motivation on cognitive performance.