A classic study: Richard LaPiere, 1934

1
health behaviour lecture 2 social
cognition models the theory of planned behaviour
falko sniehotta, t8 william guild
buildingf.sniehotta_at_abdn.ac.uk
2
What are Attitudes?

• We here indicate that attitudes are
  predispositions to respond to some class of
  stimuli with certain classes of responses and
  designate the three major types of response as
  cognitive, affective, behavioural. (Rosenberg
  Hovland, 1960, p. 3 )
• ... attitudes are enduring systems of positive
  or negative evaluations, emotional feelings, and
  pro or con action tendencies with respect to
  social objects. (Krech, Crutchfield,
  Ballachey, 1962, p.139)

3
Three Component Model of Attitudes
cognitive reactions
Object
Attitudes
affective reactions
behavioural reactions
4
For a long time, attitudes were seen as key
predictor of behaviour
5
A classic study Richard LaPiere, 1934

• LaPiere travelled with a young Chinese couple
  through the USA in 1930.
• They made 251 visits to the hotels and
  restaurants, and in only one case were they
  refused service.
• Six months later, he sent a letter to each
  establishment enquiring whether they would accept
  members of the Chinese race as guests in your
  establishment.
• 92 percent of the restaurants and hotels replied
  no. Only one person gave a definite yes.
• Conclusion people's verbal report of their
  attitudes might not be very good predictors of
  their actual behaviour

6
Attitudes alone do not explain behaviour

• The LaPiere study was followed by many other
  studies showing that people do not always act in
  accordance with their attitudes.
• Further determinants behave been suggested that
  explain peoples decision to act in certain ways.

7
Theory of Reasoned ActionAjzen Fishbein, 1980
8
Expectancy-Value Model of Attitudes(Fishbein
Ajzen, 1976)

• A person holds many beliefs about an attitude
  object an object is seen as having many
  attributes
• Associated with each attribute is an evaluative
  response (i.e., an attitude)
• Through a learning process evaluative responses
  are associated with the attitude object
• Learned evaluative responses summate
• On future occasions, an attitude object will
  elicit this summated evaluative response
• An attitude is a function of a person's beliefs
  about an object and the evaluative responses

9
Expectancy-Value Model of Attitudes

• An attitude can be conceptualized by the
  following equation
• Ao ?(bi x ei)
• Ao The attitude toward an object (o)
• bi A belief about the object's attributes
• ei The evaluation of an attribute
• A belief is the subjective probability that an
  object has a given attribute. A value could be
  assigned to designate the degree to which the
  attitude object has a given attribute
• The evaluation is a rating of the attribute along
  some evaluative dimension (e.g., good-bad).

10
Expectancy-Value Model of Attitudes

• One's attitude is a sum of the product of each
  belief times its evaluation
• Beliefs are held in a hierarchy
• An attitude is determined at any given time by
  the 5 to 9 most salient beliefs in one's belief
  hierarchy
• Types of beliefs
• Descriptive belief Based on direct experience
  with the attitude object. Held with maximum
  confidence
• Inferential belief Belief based on an inference
  process. Infer a belief from other beliefs
• Informational belief Belief based on information
  from an outside source

11

• Behavioural Beliefs about regular physical
  exercise
• A What do you see as the advantages of engaging
• in regular exercise over the next six months?
• B What are the likely positive outcomes of
  engaging
• in regular exercise over the next six months?
• B What do you see as the disadvantages of
  engaging
• in regular exercise over the next six months?
• A What are the likely negative outcomes of
  engaging
• in regular exercise over the next six months?

12

• Normative Beliefs
• A Who (i.e. which individuals or groups) would
  approve of you engaging in regular exercise over
  the next six months?
• B Who (i.e. which individuals or groups) would
  want you to engage in regular exercise over the
  next six months?
• B Who (i.e. which individuals or groups) would
  disapprove of you engaging in regular exercise
  over the next six months
• A Who (i.e. which individuals or groups) wouldnt
  want you to engage in regular exercise over the
  next six months?

13
Theory of Planned BehaviorAjzen, 1985
14

• Control Beliefs
• What are the things that might make it easier for
  you
• to engage in regular exercise over the next six
  months?
• What are the things that might facilitate you
• engaging in regular exercise over the next six
  months?
• What are the things that might make it more
  difficult for you to engage in regular exercise
  over the next six months?
• What are the things that might hinder or prevent
  you
• from engaging in regular exercise over the next
  six
• months?

15

• Operationalization of the Model
• Belief-based (indirect) measures
• Pilot studies - semi-structured interviews or Qs
• Select the modal salient beliefs
• Ajzen Fishbein (1980) suggest that those
  beliefs that exceed a certain frequency should
  be chosen

16
Measurement of TPB cognitions
17

• Behavioural Beliefs and Outcome Evaluations
• 1. Engaging in regular exercise over the next six
• months would improve my fitness
• Unlikely 1 2 3 4 5 6 7 Likely
• 2. Improving my fitness would be
• Bad 1 2 3 4 5 6 7 Good
• Bad -3 -2 -1 0 1 2 3 Good
• Bad ? ? ? ? ? ? ? Good

18

• Normative Beliefs and Motivation to Comply
• 1. My friends think that I should/should not
  engage in
• regular exercise over the six months
• Think 1 2 3 4 5 6 7 Think I
• I should should not
• With regard to exercising, how much do you want
• to do what your friends think you should?
• Not 1 2 3 4 5 6 7 Very much
• at all

19

• Control Beliefs and Power
• The availability of cheap sports facilities would
  make engaging in regular exercise over the next
  six months
• More 1 2 3 4 5 6 7 More
• easy difficult
• Cheap sports facilities are available
• Never 1 2 3 4 5 6 7 Frequently

20

• Operationalization of the Model
• Direct measures
• Guidelines
• Icek Ajzen (2006)
• http//www.people.umass.edu/aizen/pdf/tpb.measurem
  ent.pdf
• Jill Francis et al (2004)
• http//www.rebeqi.org/ViewFile.aspx?itemID212

21

• Attitude
• My engaging in regular exercise over the
• next six months would be
• Bad 1 2 3 4 5 6 7 Good
• Pleasant 1 2 3 4 5 6
  7 Unpleasant
• Enjoyable 1 2 3 4 5 6
  7 Unenjoyable
• Wise 1 2 3 4 5 6
  7 Foolish
• Harmful 1 2 3 4 5 6
  7 Beneficial

22

• Subjective Norm
• People who are important to me would approve/
• disapprove of me engaging in regular exercise
• over the next six months
• Would 1 2 3 4 5 6 7 Would
• approve disapprove
• 2. People who are important to me think I should/
• should not engage in regular exercise over the
  next six months
• Think 1 2 3 4 5 6 7 Think I
• I should should not

23

• Perceived Behavioural Control
• How much control do you feel you have over
  engaging in regular exercise over the next six
  months?
• No 1 2 3 4 5 6 7 Complete
• control control
• 2. My engaging in regular exercise over the next
  six months would be
• Easy 1 2 3 4 5 6 7 Difficult

24

• Intention
• I intend to engage in regular exercise over the
• next six months
• Definitely 1 2 3 4 5
  6 7 Definitely
• do not do
• 2. How likely is it that you will engage in
  regular exercise over the next six months?
• Likely 1 2 3 4 5
  6 7 Unlikely

25

• Norman, P., Conner, M., Bell, R. (2000)
• The theory of planned behaviour and exercise.
• British Journal of Health Psychology, 5, 249-261.

26

• Respondents
• 110 patients attending a health promotion clinic
  at their general practice
• 87 followed-up by postal questionnaire at six
  months
• Time 1 - TPB Questionnaire
• Time 2 - Exercise Behaviour

27
Attitude Taking regular physical activity over
the next six months would be bad-good,
unpleasant-pleasant, etc Subjective Norm People
who are important to me think that I
should/should not take regular physical activity
over the next six months Perceived Behavioural
Control How much control do you feel you have
over taking regular physical activity over the
next six months? Intention I intend to take
regular physical activity over the next six
months
28

• Table 1 Descriptive Statistics
• _____________________________________________
• Items Alpha Mean SD
• _____________________________________________
• Intention 3 0.95 2.01 1.52
• Attitude 5 0.85 2.21 0.97
• Subjective Norm 1 - 1.40 1.77
• Per Behav Control 4 0.87 1.71 1.33
• ______________________________________________

29

• Table 2 Correlations
• _____________________________________________
• Intention Behaviour
• _____________________________________________
• Intention - .32
• Attitude .33 .23
• Subjective Norm .16 -.22
• Per Behav Control .74 .37
• _____________________________________________
• p lt .05, p lt .01, p lt .001

30

• Table 3 Regression Analysis Predicting
  Intention
• ____________________________________
• Intention
• ____________________________________
• Attitude .04
• Subjective Norm .14
• Per Behav Control .71
• ____________________________________
• R2 .53, p lt .001

31

• Table 4 Regression Analysis Predicting
  Behaviour
• _________________________________
• Behaviour
• _________________________________
• Intention -.01
• Per Behav Control .39
• _________________________________
• R2 .15, p lt .05, p lt .01

32
TPB meta-analysis (Conner Armitage, 2001)
r.50
BBs
ATT
r.49
r.47
r.50
r.34
NBs
SN
BEH
INT
r.43
r.52
r.37
CBs
PBC
33
Evidence

• More than 1000 studies have tested the TPB in
  various behaviour domains.
• There is compelling evidence that the TPB
  accounts for about 50 of the variance in
  intentions and about 28 of the variance in
  behaviour.
• But do interventions based on the TPB actually
  change behaviour?

34
APPLICATION OF THE THEORY OFPLANNED BEHAVIOUR IN
BEHAVIOURCHANGE INTERVENTIONSA SYSTEMATIC
REVIEW

• Hardeman et al., 2002 Psych Health

35
Hardeman review of TPB interventions

• 24 studies were found
• TPB was mainly used to measure process and
  outcome variables and
• less frequently to develop the interventions

36
Hardeman review Results

• half of the interventions were effective in
  changing intention,
• two-thirds in changing behaviour, with small
  effect sizes.
• Effectiveness was unrelated to use of the theory
  to develop interventions
• Sparse evidence about mediation of effects by TPB

37
Does Changing Behavioral Intentions Engender
Behavior Change? A Meta-analysis of the
Experimental Evidence

• Thomas L. Webb Paschal Sheeran
• In press, Psychological Bulletin

38
Does Changing Behavioral Intentions Engender
Behavior Change?

• Method Review of studies that assigned
  participants randomly to a treatment that
  significantly increases the strength of
  respective intentions relative to a control
  condition, and comparing differences in
  subsequent behavior
• Results 47 experimental tests of
  intention-behavior relations that satisfied these
  criteria.
• medium-to-large change in intention (d .66)
  leads to a small-to-medium change in behavior
  (d .36).

39
Experimental tests of the TPB

• McCarty (1981) 22 orthogonal experimental
  design based on the TRA (attitudes subjective
  norms) ? no effect on cognitions or behaviour
• Chatzisarantis Hagger (2005) persuasion based
  communications targeting salient vs. non salient
  beliefs ? small effects on intention, no effect
  of behaviour

40
A full factorial experimental test of the Theory
of Planned Behaviour

• Sniehotta (2008)

41
Full factorial experimental test of the Theory of
Planned Behaviour

• N579 university students completing an online
  survey
• Design orthogonal 2 (behavioral-belief-interventi
  on or not) 2 (normative-belief-intervention or
  not) 2 (control-belief-intervention or not)
  factorial
• Randomisation computerised, double blind
• Primary outcome objectively recorded attendance
  to university sports facilities over 2 months
• Secondary outcomes post-intervention TPB
  measures
• Interventions Online delivered persuasive
  messages addressing salient modal behavioural,
  normative and control beliefs elicited in a prior
  pilot study (N52).

Sniehotta, under review
42
Rational of the experiment
Behavioural Belief intervention
? Attitudes
Normative Belief Intervention
? Subjective Norms
? Behaviour
? Intention
Control Belief Intervention
? PBC
43
Regression analyses
Attitudes
R2 .57
R2 .13
SR attendance
Behavioural Intentions
Subjective Norm
Perceived Behavioural Control
44
Regression analyses

• Post intervention TPB measures were highly
  predictive of attendance over the following 2
  months, with both intention (ß.167 p.012) and
  PBC (ß.241 plt.001) significantly contributing
  to the prediction (R2.134 R2adjusted.129
  plt.001).
• Adding attitudes and subjective norm to the
  equation did not improve the prediction.
• Thus, correlation analyses fully support the TPB.

45
Effects of the BB intervention
Behavioural Belief intervention
(F(1,5306.046 p.014 eta2.011).
46
Effects of the NB intervention
Normative Belief Intervention
(F(1,5314.428 p.036 eta2.008),
(F(1,53110.667 plt.001 eta2.020).
(F(1,5335.075 p.025 eta2.010)
47
Effects of the CB intervention
(F(1,5644.980 p.026 eta2.009).
Control Belief Intervention
48
What is wrong with the TPB?

• Attenuation problem
• Technology problem
• Bottle-neck problem
• Non-linearity problem

49
The TPB is a perfect theory to explain why
reasoned action interventions fail to change
behaviour
r.50
BBs
ATT
r.49
r.47
r.50
r.34
NBs
SN
BEH
INT
r.43
r.52
r.37
CBs
PBC
Based on Armitage Conner, 2001
50
The intention threshold Intentions are
neccessary but NOT sufficient conditions of
behaviour change
Physical Exercise Following Cardiac
Rehabilitation
After Orbell Sheeran, 1998 Sheeran, 2002 Data
from Sniehotta, Scholz Schwarzer, 2005)
51
Conclusions

• The experiment if the first strict test of the
  TPB.
• While the assumption of the TPB concerning
  intentions are confirmed
• The effect of behaviour does not seem to be
  mediated by cognition as the theory would suggest.

52
Overall conclusions

• The TPB is the most frequently model of behaviour
  in the social sciences
• While there is compelling evidence for the TPB to
  be predictive of behaviour,
• The evidence for the TPB as behaviour change
  theory is quite limited.
• In spite of good intentions, many people fail to
  translate intentions into behaviour.

53
Icek Aizen (Ajzen) http//www-unix.oit.umass.edu/
aizen/
54
Why do people behave unhealthily?

• Behaviour is acquired through an lifelong ongoing
  interaction with the environment.
• Causal explanation for acts of behaviour are
  difficult to find and they would not necessarily
  tell us how to change behaviour.
• Social cognitive models specify modifiable
  immediate (proximal) determinants of behaviour.
• They assume that other causes and determinants
  are mediated by these proximal determinants of
  behaviour.

55
Social cognition models of behaviour

• Social cognitive constructs such as
  self-efficacy, attitudes, subjective norms, etc.
  are used to predict behaviour based on empirical
  or theoretically derived regression equations.
  These models estimate the amount of behaviour, or
  its probability of occurring. The process of
  health behaviour change is construed in a linear
  and continuous way.

56
Core prinicples of social cognition models

• Parsimony and economy
• Behaviour specific predictors (principle of
  correspondence)
• Specific cognitions that are addressable in
  interventionsm, rather and inconscious processes
• Principle of mediation All other variables are
  thought to be mediated by the models predictors.

57
Social cognition models Interventions

• Interventions aim to progress individuals on the
  contunuum towards behaviour or a higher
  likelihood of behaviour.
• The same interventions should be suitasble for
  all individuals (no moderators with respect to
  preparedness, previous experiance, etc.)

58
Imagine this as a result of a regression. On the
y axis you see the predicted values from the
regression (a linear combination of predictors),
on the x axis the observed values. The slope of
the regression line indicates a relationship. If
we know the values of predictors and the
regression equation we can make decent
estimations of participants observed values.
Without knowing the observed values, we can make
predictions with some accuracy based on a simple
model. This model captures SOME of the original
variance but the deviances of the observed values
to the regression line indicates that there is
variance in the measures that we do not account
for. If we rely on the regression model, this
variance is lost, i.e. it is unaccounted for.
Method Background Regression analysis