Transcript The Psychology of food choice

The Psychology of food
choice
E.P. Köster
Center for Innovative Consumer
Studies (CICS)
Wageningen University
The Netherlands
Food choice and psychology
• Food choice and eating and drinking are among the
most important human behaviours.
• Psychology is the study of behaviour and its
conscious and non-conscious determinants.
• Nevertheless, for a long time eating, drinking and
food choice were studied very little by psychology.
• What were the reasons for this neglect and why is it
changing now ?
• How will and should this change affect our thinking
about food choice and modify our methods to
predict it ?
Eating, Drinking and Food Choice
• A seemingly simple, but in fact very complicated
behaviour that is influenced by many interacting factors.
• A predominantly learned behaviour with the exception of
a dislike for bitter and a preference for sweet and it has
very little inborn rules, except for the physiological
mechanisms involved in hunger, thirst and satiety.
• A very frequent behaviour that to a large extent is
determined by implicit and unconscious intuitive
motives, that are hard to investigate.
• A behaviour that has many non-psychological
determinants, demanding interdisciplinary research.
• For a long time psychology has rejected explanations
implying unconscious determinants of behaviour.
Basic questions and the disciplines that
try to answer them
Why : Biology, Physiology,
does
Motivation and Decision Psychology
Who : Biology, Sociology,
eat
Social, Differential and Developmental Psychology
What : Sensory, Consumer and Food Science, Marketing,
and
Perception, Memory and Learning Psychology
Where: Economy, Sociology, Marketing, Consumer Science,
and
Memory, Emotion, Social and Decision Psychology
When : Economy, Sociology, Marketing, Consumer Science,
?
Memory, Emotion, Social and Decision Psychology
Essential factors that influence eating and drinking behaviour and food choice (Mojet)
Genetic factors
Immuno system
Brain imaging
2.1
Cognition
Emotion Motivation
Decision making
3.1
Age/Gender
Physical condition
Sensory acuity
2.2
Oro-gastrointestinal physiology
2.3
Appearance
Interaction Taste Smell
Texture Trigeminal
1.1
Personality traits
Neophobia
3.3
Biological &
Physiological
factors
2
Situational
factors
4
Modelling
Data-integration
&
Co-ordination
0
Intrinsic product
characteristics
Perception
1
Socio-cultural
factors
5
Extrinsic product
characteristics
Expectations
6
Claims/brandlabel
packaging
6.1
Integrity
Sustainability
6.2
Time
Social surroundings
Physical surroundings
4.1
Coping
Assimilation
Habituation
4.2
Psychological
factors
3
Complexity
Adaptation
Dynamic contrast
1.2
Irritation
Boredom
Aversion
1.3
Memory
Previous experiences
Learning
3.2
Intentionality
Signification
Attribution
4.3
Cultural, economical
influence
5.1
Trust in industry
& gouvernement
5.2
Risk perception
Copyright J. Mojet ATO 18-11-2001
6.3
Changeing beliefs,
norms, habits,attitudes
5.3
Relevant developments in Psychology
• Learning and memory
– Incidental vs. intentional; Implicit vs. explicit
• Development
– Sensitive periods
• Active vs. Passive perception
– How does it function? vs. What can be perceived?
• Thinking, reasoning and conscious control
– Intuitive thinking and biases vs. rationality
• Motivation, Emotion and Deciding
– Conscious vs. Unconscious
Learning and memory
• Food choice is a learned behaviour. It is mostly based
on incidental learning and implicit memory.
• Many different forms of learning are involved in the
formation of food habits and depending on the form of
acquisition, some food habits are more resistant (+) to
change than others (-):
– Imprinting and conditioning (pre- and peri-natal)*
(+++)
– Praise, reward and punishment (parents or others)* (++)
– Imitation (parents, peers, idols)
(±)
– “Sensory” learning (complexity, boredom, exposure)* (++)
– Cognitive learning (advice, labeling, risk perception) (-,+)
* Largely implicit and unconscious
Development
• Sensitive periods for stable food habit
formation and/or change
– Peri-natal (Schaal, Menella, Nicklaus, Haller)
– Early childhood (Birch, Rigal, Hanse)
– Late adolescence (new “traditions”)*
– Child birth (women) *
– Retirement *
* Only anecdotal evidence
Active versus passive perception
• Psychophysics and sensory analysis have learned
us a lot about the possibilities and limitations of our
perceptual systems, but they have told us little about
how and to what extent these systems influence and
control our actions.
• Goodale and Milner: In vision two independent systems:
– Knowing what we see: a ball, a glass
– Using seeing in catching a ball or picking up a glass
• Why, in sensory analysis, do we devote much more
attention to the sensory properties than to the question
what function they have in life?
– Are sensory properties essential or are they just learned signals
for other forms of satisfaction, evokers of expectations ?
Intuition and Reasoning
• Intuitive thinking: Heuristics and biases in
decision making under uncertainty.
Kahneman and Tversky1972, 1973, 1979; Tversky and
Kahneman, 1974, 1984. Kahneman 2003.
Perception
Process
Content
Intuition
Reasoning
System 1
System 2
Fast
Parallel
Automatic
Effortless
Associative
Slow-learning
Emotional
Percepts
Current
stimulation
Stimus-bound
Slow
Serial
Controlled
Effortful
Rule-governed
Flexible
Neutral
Conceptual representations
Past, Present and Future
Can be evoked by language
Intuitive judgment: biases and limitations
• People (even trained statisticians) make
many mistakes in simple problems,
because they do not reason, but use
intuitive evidence for fast solutions without
checking.
• Example:
A baseball bat and a ball cost $ 1.10.
The bat costs $1.00 more than the ball.
How much does the ball cost ?
Bat and ball
• Intuitive answer most commonly given:
Ball = $ 0.10
• Time consuming reasoning solution:
Bat + Ball = $ 1.10
Bat = Ball + $ 1.00
(Ball + $ 1.00) + Ball = $ 1.10
2 Balls = $0.10
1 Ball = $ 0.05
1 Bat = $ 1.05
Intuition and corrective control by reasoning
• When will corrective thoughts (system 2) intervene?
– When people make errors in the intuitive system
and
– When these errors are recognised as errors by reasoning
• Corrective operations are impaired by:
–
–
–
–
Time pressure
Concurrent involvement in other cognitive tasks
Incongruency with bio-rhytms (morning-evening people)
Being in a good mood
• Corrective operations are positively correlated with:
– Intelligence
(average in the population)
– Need for cognition
(low in relation to food)
– Exposure to statistical thinking (very low even among
educated people)
• Conclusion: Corrective control occurs seldom
Motivation, Emotion and Deciding
• Conscious vs. Unconscious
–
–
–
–
–
–
–
–
Zajonc (1968): Priority of emotion over cognition
Kahneman and Tversky (1972,1973,1979, 2003)
Nisbett and Wilson (1977): Telling more than we know.
Wilson, Dunn(1986):Analyzing reasons vs. Focussing on
feelings. Latter more effective than first
Bargh (1990) Preconscious determinants of social
interaction; Role of automaticity.
Wilson, Schooler 1991): Thinking too much: Introspection
harmful for decisions.
Damasio (1994): Essential role of emotion in thinking and
behaviour. Unawareness and physiological underpinning
LeDoux (1998):Defensive behaviour; “emotional
unconscoius”; physiological underpinning
Food choice behaviour is changing under
the influence of exposure
• At first encounter
– Specific exploration and novelty reduction
– Optimal Arousal Level
• During consumption
– Adaptation
– Sensory Specific Satiety
• Over longer periods
– Change in Optimal Arousal Level
– Product boredom
– Slowly rising aversion
Theory of Dember and Earl
A
B
Optimum
Perceived Complexity
Theory of
Dember
and Earl
Theory of Walker
B
Optimum
A
Appreciation
Experience
METHOD
• Subjects: N=168
51% females, 49% males
49% 18-25 years, 51% 26-35 years
• Stimuli
7 orange drinks equal in caloric value and of very similar
taste, but varied in complexity by small additions of natural
aromas.
• Scaling
9-point bipolar rating scales measuring the attributes
Appreciation, Surprising, Ample, Confusing, Fantastic,
Elaborated, Shattered, Marked.
• Procedure
– Initial rating of all 7 drinks
– Repeated exposure to one of the 7 drinks
– Final rating of all 7 drinks
Effects of exposure to simple products
5,9
5,8
Appreciation
5,7
5,6
5,5
5,4
5,3
5,2
5,1
5
Simple
Initial rating
Type of product
Complex
Final rating
Effects of exposure to complex products
5,8
5,7
5,6
Appreciation
5,5
5,4
5,3
5,2
5,1
5
4,9
4,8
Simple
Initial rating
Type of product
Complex
Final rating
Change of the optimal level of complexity
7
6
5
4
initial measurement
3
(before exposure)
2
f inal measurement (a
1
f ter exposure)
N=
43
43
1
47
47
2
40
40
3
16
16
4
22
22
5
Experimental groups
Cases weighted by the number of indiv idual optima
Sensory and consumer research
• Almost entirely based on conscious methods
(scaling and questionnaires)
• Bringing people in analytical attitudes by asking
them questions that they normally never ask
themselves, but just answer in their actions (Do
you like this? Is this just right on salt ?)
• Strong adherence to a rational consumer model
such as the theory of planned behaviour and
asking unanswereable questions about attitudes
and intentions.
(Wilson, Strangers to ourselves 2002)
Theory of reasoned action or planned
behaviour (Ajzen and Fishbein) .
• Based on the idea that behaviour is directly provoked by
rational and conscious intentions which are in turn influenced
by attitudes and by beliefs about ones’ own values and
control possibilities and about the judgments of ‘important’
others.
• Objections to the theory:
– Based on (low) correlations (but no scatter plots)
– Past behaviour (learning) is a better predictor of actual
behaviour than attitudes or intentions.
(Bem 1972: We base our opinions about our attitudes and
intentions on past behaviour. Wilson 2002: narratives
about ourselves)
– Based on conscious and rational decision making
(see Kahneman and Tversky, Zajonc, Damasio, LeDoux
and Wilson)
Towards a new way of studying food choice
Reductionist vs. Deductionist approach
• Should we continue to dissect human behaviour into many
sub-behaviours that can easily be studied in the laboratory
and after studying these, try to recombine them ? (Sensory
properties, Hedonics, Context etc.)
Or
• Should we try to find ways to study natural complex behaviour
in its totality and to deduct the determinants of such integrated
behaviour from the effects of variations in the circumstances
under which it occurs?
(Observation and Situational Analysis)
Situations
• A situation is an event in the life of an individual
in which his personal history is organised in his
perception of, and expectations about the things
around him. Intentionality is the organiser.
– Hunger shows “eatables”
– Hunger on a cold winter morning shows …?
– Hunger on a hot summer beach shows ….?
• Situations are created by the meaning the
subjects unconsciously convey to the
surroundings (signification by intentionality).
They are not the surroundings themselves.
• Although highly individual, situations are shared
Situational analysis
• Collecting typical eating situations
• Segmenting people on the basis of the
frequency with which these situations occur in
their life.
• Simulating situations with people that are in
different segments. Measuring reactions to foods
• Simulating different situations with people from
the same segment or different segments
• Situations can be evoked by stories, images,
etc.
Central Location Test vs. Home Party
8
Central Location
Hedonic scale values
7
Home Party
6
5
4
3
2
1
0
P
Q
R
Pure spirits
S
A
B
C
Mixed drinks
D
Simulated vs. real Disco experience
35
Simulation
Disco
30
Numberof choices
25
20
15
10
5
0
Non-alcoholic Pure spirits
Mixed A
Mixed B
Mixed C
Mixed D
Reductionist vs. Deductionist approach
Reductionist
Deductionist
• Isolated behaviours
• Experimental manipulation
of single determinants
(sensory, hedonic, social)
• Measuring mean effects of
and interactions between
determinants
• Reconstitution of the total
behaviour on the basis of
mean group results (?!?)
• Re-integration of data by
the experimenter
• Total behaviour
• Situational manipulation or
situational comparison
with subject segmentation
• Observation and hypothetical
interpretation of behavioural
effects and differences
• Verification of interpretations
in similar situations with new
and/or old subjects
• Integration of behaviour by
the individual subjects
Conclusions
• More interdisciplinary research and a deductionist
approach is needed to make real progress.
• Many of our consciousness-oriented and rather
explicit methods should be replaced or critically
evaluated depending on the research goal.
(OK in technical research, not in hedonic research)
• More emphasis on research that shows real
behavioural or physiological effects is needed.
(less correlational research on statements only;
observation of actual choice in typical situations
rather than just asking about liking or wanting)