Attribution and explaining

Psychology 2A

Dr Cristina Marinho
A photo-still of Will Smith slapping Chris Rock at the 94th Academy Awards in
March 2022. The incident had many people wondering why Smith behaved in this
way.
Overview

• Attribution theories
• Heider
• Kelley’s Covariation theory

• Evaluation of attribution theories

• Biases in attribution

• Real world explanations (Discursive Psychology)

The development of attribution theory
Heider (1958)

 What attributions?

 Why does attribution matter?

 What information may the attribution process provide?

 How can we best portray the process of attribution?

The development of attribution theory
Heider (1958)
 What attributions? Causal attributions – we look for causes of
behaviour, occurrences, experiences

 Why does attribution matter? Locating a cause can help make the
world less worrying and unpredictable

 What information may the attribution process provide? Information

about whether the cause is within the person (internal) or within
the situation (external)

 How can we best portray the process of attribution? People are

like ‘naïve psychologists’
 Kelley’s (1967, 1973)
Covariation model
• Harold Kelley is the best-known attribution theorist
• Like his predecessors, he thought that people tried to
discover causes of behaviour
• The causes could be internal (e.g. personality) or external
(e.g. social pressure)
• People act like scientists trying to identify which factor (e.g.
personality) most closely covaries with the behaviour they
want to explain, and then they assign a causal role to this
factor
• NB Kelley’s model assumes people conduct a rudimentary
analysis of variance (ANOVA)
 Covariation model
Kelley proposed that we used three kinds of information to
determine the cause of the target behaviour

1. Consensus – ‘Do other people act like this/have this

experience?’

2. Distinctiveness – ‘Across multiple contexts, is it unusual for

the actor (the person whose behaviour I am trying to explain)
to act like this/have this experience?’

3. Consistency – ‘Does this keep happening – does the actor

always act like this/have this experience in this sort of
situation?’
 Covariation model
We can apply this to the example we started with: why did Will
Smith slap Chris Rock?

1. Consensus – ‘Do other people act like this/have this

experience?’ Did other people take turns to slap Chris Rock?

2. Distinctiveness – ‘Across multiple contexts, is it unusual for

the actor (the person whose behaviour I am trying to explain)
to act like this/have this experience?’ How usual – or unusual
– is it for Will Smith to slap someone?

3. Consistency – ‘Does this keep happening – does the actor

always act like this/have this experience in this sort of
situation?’ Does Will Smith consistently slap, rather than clap
at award ceremonies? Does he consistently slap Chris Rock?
 Kelley’s attribution theory

consistency distinctiveness consensus attribution

Low Discounting
(search for
different cause)

consistency
High + High + High External attribution

High + Low + Low Internal attribution

(Adapted from Hogg & Vaughan)

Quick check
• Tala invited several friends to meet up. One of them, Himari,
declined
• Tala is trying to understand Himari’s behaviour (what caused
him to decline the invitation?)

• She knows that:

- Other friends accepted the invitation
- Himari often declines invitations
- Himari has declined previous invitations from Tala

 Would Tala make an internal or external attribution for Himari’s

behaviour?
Evidence

• McArthur (1972) tested the Covariation theory:

• Participants were presented with 8 possible

configurations of consensus, consistency and
distinctiveness through scenarios
• And asked to make internal or external attributions for a
range of behaviours

 The theory was generally supported (although there was a

tendency to underuse consensus information)
 Problems
with Kelley’s Covariation model (1)
• Experiments show people can use covariation information, but this
does not mean they normally do so (and Alloy & Tabachnik, 1984
show they are poor at using covariation)

• Covariation requires multiple observations which we may not have –

for example, of how usual or unusual the actor’s behaviour (or
experience) is both in general (distinctiveness) and in this specific
situation (consistency)

• It is most suitable for unintentional behaviour or unintended events

(we normally assume that intentions are internal, so the choice of
cause is redundant) (Malle, 2011)
 Problems
with Kelley’s Covariation model (2)
• It is possible that identifying a cause as internal or external is
sufficient, but in many circumstances, we will want to be more
specific (Dickerson, 2024)
• Even in McArthur’s (1972) studies, people underused consensus
information
• Methodologically:
• Person and situation measures are treated as DVs (people were
not asked to provide causal explanations – which may have been
more complex) (Malle, 2011)
• People don’t spontaneously search for covariation information
when trying to explain behaviour (Lalljee et al., 1984)
 Summary
• Attribution Theory became popular in Social Psychology, especially
– the idea that we seek causal attributions for behaviour and
occurrences
– that identifiable types of information influence our attributions in
ways that could be specified

• It was extended and applied, for example,

– Understanding causes and consequences of attributions people
make for achievement of tasks (Weiner, 1979, 1985)
– Understanding use and consequences of causal attributions in
relationship conflict (e.g. Fincham & Bradbury, 1993)

• However, the utility and accuracy of the concept of the ‘naïve

psychologist’ was questioned
 ‘Cognitive miser’
Attributional Biases This model of the
person views people
as taking cognitive
• The attribution process is subject to
short-cuts, which are
forms of bias
usually adaptive and
sufficient if not
• Three particularly influential biases accurate.
are the:
– Fundamental Attribution Error (FAE) ‘Motivated tactician’
– Actor observer difference This model of the
– Self-serving Bias person views people
as having multiple
cognitive strategies,
• They assume a model of the person as from which they
a cognitive miser or motivated tactician choose according to
rather than a naïve psychologist. personal goals,
motives and needs.
a. Fundamental Attribution Error (FAE)
• The FAE is the tendency for observers to attribute behaviour to
the person (internal) rather than factors in the situation (external).

• It has been shown in many experiments (e.g. Jones and Harris,

1967)

• And outside the lab, it explains

o people’s tendency to attribute road accidents to the driver
rather than the vehicle or road conditions (Barjonet, 1980).
o people’s tendency to attribute poverty and unemployment to
the person rather than social conditions
o How medical professionals can over-simplify the cause of a
medical problem by locating it within the person
 Evidence for the FAE (Jones & Harris, 1967)

(From Myers et al, 2021:129)

Cultural and developmental factors

• Several research have questioned whether the fundamental

attribution error is fundamental.

• In Western cultures, young children explain action in

situational terms and learn to make dispositional attributions
(Kassin & Prior, 1985). Hindu children, however, shift further
towards situational attributions as they get older (Miller,
1984)

• These differences reflect cultural norms and values (Chiu &

Hong, 2007)
Dispositional
attributions as
a function of
age and
cultural
background

Source: Based on data from Miller (1984); figure from Hogg & Vaughan, 2022).
 b. Actor-observer
effect The actor-observer effect is
the tendency to attribute our
own behaviours to the
situation and others’
behaviours to the person
(Jones & Nisbett, 1972).

It is shown when participants

use dispositional items on a
checklist to describe friends’
behaviour and situational
items when describing their
own behaviour (Nisbett et
al., 1977).
(From Dickerson, 2024:186)
b. Actor-observer effect: explanations
Two main explanations for the effect are:

1. Informational difference
Actors have more information about themselves and how
the situation is impacting them (Jones & Nisbett, 1972)

2. Difference in perceptual focus

Storms (1973) argued that we blame what we see (actors
see the situation, observers see the actor).
McArthur and Post (1977) found that observers make more
dispositional attributions for an actor’s behaviour when the
actor was strongly (cf. dimly) illuminated.
c. false consensus effect

• In Kelley’s (1972) model, consensus information was used

by people to make causal attributions.
• McArthur’s (1972) studies suggested that consensus was
underused or ignored.
• Ross, Greene and House (1977) suggested that people use
false consensus information
 i.e. we see our own behaviour as being more typical
than it is
• The effect is especially strong for important beliefs
(Granberg, 1987) or those we about which we feel certain
(Marks & Miller, 1985)
d. Self-Serving bias

The self-serving attribution

bias suggests that we
attribute our successes to
ourselves and our failures to
factors outside of ourselves.
(From Dickerson, 2024: 194)

e.g. US politicians attribute Mezulis, Abramson, Hyde and

success to internal factors Hankin’s (2004) meta-analysis
(hard work) and failures to found that a large self-serving bias
situational factors (wider was present in ‘nearly all samples’
voting trends) (Kingdon, – but that it varied with age, culture
1967) and psychopathology of
participants.
 Critical Review of biases
• The four attributional biases explain variations in how we explain
behaviour and occurrences.

• They are supported through extensive research and support a model

of people’s thinking as adaptive and strategic.

• However, the distinction between ‘external’ and ‘internal’ attributions

has limited utility (e.g. it matters whether we attribute failure to lack of
effort or ability; Hewstone, 1989).

• Attributing ‘responsibility’ may be more important for predicting future

behaviour (Cheng et al, 2017)

• The biases are not, as initially thought, universal

The different foci of attribution
theories and discursive psychology
Attribution research
is traditionally
focused on internal
cognitive processes.

Discursive psychology research is

concerned with how people
talk about causes, reasons,
responsibility and so on for
behaviours, and what is thereby
achieved.
From Dickerson, 2024
 Discursive psychology (DP)
and accounting for behaviour
• DP is concerned with how behaviour is made an accountable
matter
• To understand this, DP focuses on sequences of actual talk
in which people hold themselves or are held to account.
• DP is interested in what our accounts achieve (e.g. denying
responsibility or agency)
• DP works with actual talk and accounting rather than
vignettes provided by the researcher
• DP is interested in the details of what is being done here and
now rather than with testing models of attribution, using
theory as explanation, or seeking general laws of how we
explain things
 Blame attributions and mitigated
confessions (Archer & Parry, 2019)
• Archer and Parry studied two celebrity TV confessions

• The researchers were interested in how the celebrity guests

attributed, accepted or denied responsibility for their
transgressions

• Analysis: They show how the celebrity interviewees produced

descriptions which minimized the extent and severity of their
transgressions, reduced their agency, transformed the nature
of the transgression

• In this way, the celebrity interviewees reduced their

responsibility for the transgression
 Ways of reducing responsibility for a transgression
(Archer and Parry, 2019)

• Minimise the extent and severity (e.g. of violence) through

descriptive terms used ‘lashed out’ carries an implication of
unplanned/spontaneous behaviour
(almost as a reflex), which offers a vague
description of the physical aggression
20 Wa: yeh kno:w (.) there (.) is a ti:me
21 when you can’t get a word in a:nd (.) .hhh you kinda
22 lash- I’d- I lashed out.
23 (0.8)
24 Wa: There wuz no- I couldn’t end the argument
25 so I (0.4) I lashed out at her. … after that IR choose ‘punch’ to
characterise the aggression, which
targets the interviewee as the agent of a
forceful action-type (more deliberate and
less reactive than ‘lashing out’); in
response IE denies ‘punching’ and uses
instead ‘clap’, which retains agency but
downgrades the forcefulness of his action
 Ways of reducing responsibility for a transgression
(Archer and Parry, 2019)
• Reducing or attributing agency to others
joint agents

Extract [5] Morgan/Waterman

09 Mo: An what were the ci:rcumstances.
10 Wa: ↑Quite honestly: I ca:n’t remember.<We wuh goin
11 through=uh- through a horrible horrible ti:me (.) and uh
12 .hhhh we wuh- <as they sa:id we were=arguing a great deal
13 a:nd >you cou(ldn’t) w-< you know >if y-< The problem
14 wi=stro:ng intelligent women °if° (0.2) ((cough/laugh))
15 some people=might disa:gree: bout that >but=(uh/I’ll)<
16 .hhh i:s. thet(h) (0.5) they-ca:n-a:rgue implicates (indirectly ‘strong,
intelligent women’) his
wife’s role an
(following admission that she had a black eye) responsibility/blame

37 Wa: ↑O:h I’m- I’m unbelievably asha:med thet it ever=ha:ppened.

38 (0.5) #I mean r#e::ally asha:med.# An it’s never=happened
39 befo:re an it’s never happened si:nce so:: .hhh u::hmm
One-off incident – without
agent ‘it… happened’
Ways of reducing responsibility for a transgression
(Archer and Parry, 2019) admission to transgression but transgression
is about failing to ‘trying to stop the culture’ –
so admission of guilt or responsibility for being
• Transforming the nature of the transgression inactive in the face of prevailing doping
culture, rather than responsible (agent) in
taking banned, performance enhancing
Oprah Winfrey’s interview with Lance Armstrong substances
2013 (Archer & Parry,
2019, pp. 605, 606).

124 Wi: You said to me earlier that you don’t

125 think it was possible to win without doping
126 Ar: Not in that generation (.) a- an I’m not here to talk about
127 others in that genera[tion it’s it’s been well documented
128 Wi: [mm hmm]
implicate others; constructs his lack of
129 Wi: Mm hmm responsibility for that culture
130 Ar: U:hm (.) I didn’t invent the culture
131 Wi: Mm hmm
132 Ar: But I didn’t try to stop the culture (.) an that’s w- tha
133 that’s my mistake (.) and that’s what I have to be sorry for
134 and that’s what something- .hhh and the sport is now paying
135 (.) the price because of that and so (.) I am sorry for that
 Nigel Lawson’s tent
Edwards and Potter (1992)
• Accounts or explanations given in a dispute between the UK Chancellor
(Nigel Lawson) and journalists over a story allegedly leaked by Lawson
and reported in the newspaper, which Lawson later denied had
happened

• Analysis: Researchers shows how attribution for behaviour (publishing

the story) is done through description (i.e. our descriptions give clues as
to the cause or intention of the described behaviour)

• Their work also shows how people construct, establish and use
consensus as a resource in accounting for what happened (or
undermine it as corroboration)
 Nigel Lawson’s tent
Potter and Edwards (1990)
• Constructing consensus to warrant the journalists’ accounts as
factual

The reporters, it seemed, had Mr David Winnick (Walsall, North): As

unanimously got it wrong. Could all the Sunday newspapers carried
so many messengers really be so virtually the same story, is the
much in error? It seems doubtful. Chancellor saying that every journalist
(Guardian, 8 November) who came to the briefing – he has not
denied that there was one –
misunderstood what he said (Hansard,
7 November: 26)
 Nigel Lawson’s tent
Potter and Edwards (1990)
• Constructing consensus as corroboration to dismiss the
journalists’ accounts as factual

Mr Lawson: I am grateful to my hon. Friend. In fact, the statements, as I

said, - the statements that appeared in the press on Sunday bore no relation
whatever to what I in fact said. What I have said to them [reporters] is that,
while we were absolutely, totally committed to maintaining
Ms Short: They will have their shorthand notes
Mr Lawson: Oh yes, they will have their shorthand notes and they will know
it, and they will know they went behind afterwards and they thought there was
not a good enough story and so they produced that. They will know that I said
that, while … (Hansard, 7 Nov:26)
 Explaining in everyday conversation
invitation

1 B Uh if you’d care to come over and visit a little

while
2 this morning I’ll give you a cup of coffee
3 A hehh Declining the invitation but without immediate
no – nor even ‘thanks but no’; A qualifies the
4 Well decline ‘I don’t think I can make it’ after delay
and talk that displays appreciation
5 that’s awfully sweet of you,
6 I don’t think I can make it this morning .hh uhm
7 I’m running an ad in the paper an-and uh I have
to
8 stay near the phone

Source: Potter (1996, cited in Atkinson & Drew, 1979, p. 58; Transcript simplified from the original)
Summary: discursive psychology and
explanations
• Discursive psychologists study explanations or accounts in
real life talk or texts

• They show the importance of attending to the details of how

actions or events are described

• Explanations or accounting has an action-orientation (e.g. to

deny responsibility, or decline an invitation)