Chapter 2

Thought is the internal language and symbols we use – it is often conscious, or at least something we
are or could be aware of.

Cognition is broader; it also refers to mental processing that can be largely automatic. Cognition acts a
bit like a com- puter program or operating system: it operates automatically in the background,
running all the functions of the computer.

Because psychologists felt that theories should be based on publicly observable and repli- cable data,
there was a shift away from studying internal (cognitive) events towards external, publicly observable
events. The ultimate expression of this change in emphasis was American behaviourism of the early
twentieth century (e.g. Skinner, 1963; Thorndike, 1940; Watson, 1930)

Behaviourists focused on overt behaviour (e.g. a hand wave) as a response to observable stimuli in the
envi- ronment (e.g. an approaching bus), based on past punishments and rewards for the behav- iour
(e.g. being picked up by the bus).

Drawing on Gestalt psychology, Lewin (1951) believed that social behaviour is most usefully
understood as a function of people’s perceptions of their world and their manipulation of such percep-
tions. Cognition and thought were placed centre stage in social psychology.

The cognitive emphasis in social psychology has had at least four guises (Jones, 1998; Taylor, 1998):
cognitive consistency, naive scientist, cognitive miser and motivated tactician.

Gestalt psychology - Perspective in which the whole influences constituent parts rather than vice
versa.

Cognitive consistency - A model of social cognition in which people try to reduce inconsistency
among their cognitions, because they find inconsistency unpleasant.

Naive psychologist
(or scientist) - Model of social cognition that characterises people as using rational, scientific-like,
cause–effect analyses to understand their world. This model underpins the attribution theories of
behaviour that dominated social psychology in the 1970s. Attribution - The process of assigning a
cause to our own behaviour, and that of others.

Cognitive miser - A model of social cognition that characterises people as using the least complex
and demanding cognitions that are able to produce generally adaptive behaviours.

motivated tactician:

a fully engaged thinker who has multiple cognitive strategies available and chooses among them
based on goals, motives, and needs.

The most recent development in social cognition is social neuroscience, sometimes called cognitive
neuroscience or social cognitive neuroscience. Social neuroscience is largely a methodology in which
cognitive activity can be monitored by the use of functional Magnetic Resonance Imaging (fMRI),
which detects and localises electrical activity in the brain associ- ated with cognitive activities or
functions.
Configural model - Asch’s Gestalt-based model of impression formation, in which central traits play
a disproportionate role in configuring the final impression.

Central traits - Traits that have a disproportionate influence on the configuration of final
impressions, in Asch’s configural model of impression formation.

peripheral traits - Traits that have an insignificant influence on the configuration of final
impressions, in Asch’s configural model of impression formation.

primacy - An order of presentation effect in which earlier presented information has a

disproportionate influence on social cognition.

recency - An order of presentation effect in which later presented information has a disproportionate
influence on social cognition.

we are biased towards negativity; a negative impression is much more difficult to change

We may be particularly sensitive to negative information for two reasons:

1 The information is unusual and distinctive – unusual, distinctive or extreme information attracts
attention (Skowronski & Carlston, 1989).

2 The information indirectly signifies potential danger, so its detection has survival value for the
individual and ultimately the species.

personal constructs - Idiosyncratic and personal ways of characterising other people.

Implicit personality theories - Idiosyncratic and personal ways of characterising other people and
explaining their behaviour.

Social judgeability - Perception of whether it is socially acceptable to judge a specific target.

Cognitive algebra - Approach to the study of impression formation that focuses on how people
combine attributes that have valence into an overall positive or negative impression.

Summation - A method of forming positive or negative impressions by summing the valence of all
the constituent person attributes.

Averaging - A method of forming positive or negative impressions by averaging the valence of all the
constituent attributes.

Weighted averaging - Method of forming positive or negative impressions by first weighting and
then averaging the valence of all the constituent person attributes.

Schema - Cognitive structure that represents knowledge about a concept or type of stimulus,
including its attributes and the relations among those attributes.

Script - A schema about an event

person schemas - Person schemas are knowledge structures about specific individuals.

role schemas - Role schemas are knowledge structures about role occupants
roles - Patterns of behaviour that distinguish between different activities within the group, and that
interrelate to one another for the greater good of the group.

Content-free schemas - Content-free schemas do not contain rich information about a specific
category but rather a limited number of rules for processing information.

Self-schemas represent and store information about themselves in a similar but more complex and
varied way than information about others. Self-schemas form part of people’s concept of who they
are, the self-concept

Fuzzy sets - Categories are considered to be fuzzy sets of features organised around a prototype.

Prototype - Cognitive representation of the typical/ideal defining features of a category.

Family resemblance - Defining property of category membership.

Associative network - Model of memory in which nodes or ideas are connected by associative links
along which cognitive activation can spread.

Accentuation principle - Categorization accentuates perceived similarities within and differences

between groups on dimensions that people believe are correlated with the categorization. The effect is
amplified where the categorization and/or dimension has subjective importance, relevance or value.

Social identity theory - Theory of group membership and intergroup relations based on self-
categorization, social comparison and the construction of a shared self-definition in terms of ingroup-
defining properties.

Self-categorization theory - Turner and associates’ theory of how the process of categorizing oneself
as a group member produces social identity and group and intergroup behaviours.

Stereotypes are not only consensual beliefs held by members of one group about members of another
group; they are also more general theories or social representations of the attributes of other groups.
Stereotypes persist if they are readily accessible to us in memory and they seem to make good sense
of people’s attitudes and behaviour (i.e. they neatly fit ‘reality’). Changes in accessibility or fit will
change the stereotype. Stereotypes can clarify social role, power differentials and intergroup conflicts,
and they can justify the status quo or contribute to a positive sense of ingroup identity.

How we use, acquire and cHange scHemas

Accessible schemas that are habitually used or salient in memory, and schemas that are relevant to
oneself in that context, are more likely to be invoked. So, for example, a racist would tend to use
racial schemas more than someone who was not racist. Also, people tend to cue mood-congruent
schemas and schemas that are based on earlier rather than later information. These fairly automatic
schema-cueing processes are functional and accurate enough for immediate interactive purposes.
They have circumscribed accuracy.

If the costs of being wrong are increased, people are more attentive to data and may use more accurate
schemas. The costs of being wrong can be important where people’s outcomes (i.e. rewards and
punishments) depend on the actions or attitudes of others. Under these circumstances, people probe
for more information, attend more closely to data, particularly to schema-inconsistent information,
and generally attend more carefully to other people. The costs of being wrong can also be impor- tant
where people need to explain or justify their decisions or actions.
There are some individual differences that may influence the degree and type of schema use:

 Attributional complexity – people vary in the complexity of their explanations of other

people
 Uncertainty orientation – people vary in their interest in gaining information versus
remaining uninformed but certain
 Need for cognition – people differ in how much they like to think deeply about things
 Need for cognitive closure – people differ in how quickly they need to tidy up cognitive
loose ends and move to a decision or make a judgement
 Cognitive complexity – people differ in the complexity of their cognitive processes and
representations

People also differ in accessibility. Accessibility - ease of recall of categories or schemas that we
already have in mind.

Acquiring schemas

Schema acquisition and devel- opment involve a number of processes:

 Schemas become more abstract, less tied to concrete instances, as more instances are
encountered
 Schemas become richer and more complex as more instances are encountered: greater
experience with a particular person or event produces a more complex schema of that
person or event
 With increasing complexity, schemas also become more tightly organised: there are more
and more complex links between schematic elements
 Increased organisation produces a more compact schema
 Schemas become more resilient
 All things being equal, this entire process should make schemas generally more accurate,
in the sense of accurately mapping social reality.

Because schemas appear to be accurate, they impart a sense of order, structure and coherence to a
social world that would otherwise be highly complex and unpredictable. Because of this, schemas do
not easily change.

People think a lot about their schemas, marshalling all sorts of supportive evidence. People also
protect their schemas by relying uncritically on their own earlier judgements – they justify and
rationalise by using prior judgements, which are in turn based on even earlier judgements.

Rothbart (1981) has studied how social categorization works and suggested three ways in which
schemas can change:

1 Bookkeeping – slow change in the face of accumulating evidence.

2 Conversion – sudden and massive change once a critical mass of disconfirming evidence has
accumulated.
3 Subtyping – schemas morph into a subcategory to accommodate disconfirming evidence. For
example, a woman who believes that men are violent might, through encountering many who are not,
form a subtype of non-violent men to contrast with violent men.

Schema change may also depend on the extent to which schemas are either logically or practically
disconfirmable compared with ones that cannot be disconfirmed at all.
A schema that is logically disconfirmable is relatively easily changed by counter-evidence: if my
schema of a stranger is that he is honest, then evidence that he has cheated is very likely to change my
schema (honest people do not cheat).

Social encoding refers to the way in which external social stimuli are represented in the mind of the
individual. There are at least four key stages (Bargh, 1984):

1 Pre-attentive analysis – an automatic and non-conscious scanning of the environment.

2 Focal attention – once noticed, stimuli are consciously identified and categorized.

3 Comprehension – stimuli are given meaning.

4 Elaborative reasoning – the stimulus is linked to other knowledge to allow for complex inferences.

Social encoding depends markedly on what captures our attention. In turn, attention is influenced by
salience, vividness and accessibility.

SALIENCE

Salience refers to the property of a stimulus that makes it stand out relative to other stimuli.

People can be salient because: they are novel or figural; they are behaving in ways that do not fit prior
expectations of them; they are important to your goals, they dominate your visual field or you have
been told to pay attention to them

Salient people attract attention, relative to non-salient people, tend to be considered more influential
in a group, more personally responsible for their behaviour and less influenced by the situation, and
they are generally evaluated more extremely.

Novelty, Figural, Unusual behaviour for that person, Unusual behaviour for people in general,
Unusual behaviour for people in that category, Person is important to your goals, Visual field
dominance, and Being instructed to watch the person lead to SALIENCE.

SALIENCE leads to - Seen as more influential, Behaviour reflects dispositions, Behaviour

uninfluenced by situation, Extremitised evaluation, and Coherent impression

VIVIDNESS.

Vividness is an intrinsic property of the stimulus itself. Vivid stimuli are ones that are: emotionally
attention-grabbing, concrete and image-provoking, close to you in time and place

ACCESSIBILITY.

Priming occurs when we become conscious of features of a stimulus domain that are highly
accessible in memory; they come easily to mind and are useful in making sense of the intrinsically
ambiguous nature of social information. They are categories that we often use, have recently used and
are consistent with current goals, needs and expectations (Bruner, 1957, 1958).

For example, people who are very concerned about sex discrimination (i.e. it is an accessible
category) may find that they see sexism almost everywhere: it is readily primed and used to interpret
the social world.
Once primed, a category tends to encode stimuli by interpreting them in a category-consistent
manner. This is particularly true of ambiguous stimuli. However, when people become aware that a
category has been primed, they often contrast stimuli with the category: they interpret them in a
category-incongruent manner

Memory for people

Social psychological approaches to person memory draw on cognitive psychological theories of

memory and mainly adopt what is called an associative network or propositional model of memory.

The general idea is that we store propositions (e.g. ‘The student reads the book’, ‘The book is a social
psychology text’, ‘The student has a ponytail’) consisting of nodes or ideas (e.g. book, ponytail,
student, reads) that are linked by relationships between ideas. The links are associative in so far as
nodes are associated with other nodes (e.g. student and ponytail), but some associative links are
stronger than others. Links become stronger the more they are activated by cognitive rehearsal.

In terms of our general impression of someone, we are more likely to recall information that is
inconsistent rather than consistent with our impression. This is because inconsistent information
attracts attention and generates more cogni- tion and thought, and this strengthens linkages and
retrieval routes.

However, better recall of inconsistent information does not occur when:

 ●  we already have a well-established impression (Fiske & Neuberg, 1990);

 ●  the inconsistency is purely descriptive and not evaluative (Wyer & Gordon, 1982);
 ●  we are making a complex judgement (Bodenhausen & Lichtenstein, 1987);
 ●  we have time afterwards to think about our impression (Wyer & Martin, 1986).

Most person-memory research concerns traits. The storage of trait information is organised with
respect to two continua: social desirability (e.g. warm, pleasant, friendly) and competence (e.g.
intelligent, industrious, efficient). Behaviour is usually perceived as purposeful action, so memory for
behaviour may be organised with respect to people’s goals. Memory for appearance is usually based
on directly observable concrete information (‘Boris has unruly yellow hair and a large face’) and is
stored as an analogue rather than a proposition. In other words, appearance is stored directly, like a
picture in the mind, which retains all the original spatial information rather than as a deconstructed set
of propositions that have symbolic meaning. Laboratory studies reveal that we are phenomenally
accurate at remembering faces: we can often recall faces with 100 per cent accuracy over very long
periods of time.

Social memory can be organised by person or by group.

Social inference

It addresses the inferential processes (which can be quite formal and abstract, or intuitive and
concrete) that we use to identify, sample and combine information to form impressions and make
judgements. There are two distinct ways in which we process social information: (a) we can rely
automatically on general schemas or stereotypes in a top-down deductive fashion; or (b) we can
deliberatively rely on specific instances in a bottom-up inductive fashion.

Asch’s configural model (impressions are based on holistic images) and Anderson’s cognitive algebra
model (impressions are based on integration of pieces of information).
Normative models - Ideal processes for making accurate social inferences.

Behavioural decision theory - Set of normative models (ideal processes) for making accurate social
inferences.

The first stage in making an inference involves gathering data and sampling information from those
data. In doing this, people rely too heavily on schemas. Person schemas are quickly, and often
unconsciously, activated and used as the basis for candidate assessment. This reliance on person
schemas is referred to as ‘clinical judgement’ and, although by no means all bad, it can produce
subop- timal inferences and judgements.

People can also be overly influenced by extreme examples and small samples (small sam- ples are
rarely representative of larger populations; this is called the law of small numbers); and they can be
inattentive to biases in samples and to how typical a sample is of its population. For example, in
Europe there is substantial media coverage of hate speech by radical ‘Islamists’ who promote anti-
Western violence and terrorism. From this, some people seem happy to infer that all 1.6 billion
Muslims in the world behave like this. However, this inference is flawed.

Regression

Individual instances are often more extreme than the average of the population from which they are
drawn: over a number of instances, there is a regression to the population mean. For example, a
restaurant you have just visited for the first time may have been truly excellent, causing you to extol
its virtues to all your friends. However, the next time you go, it turns out to be mediocre. On the next
visit, it is moderately good, and on the next fairly average. This is an example of regression.

Base-rate information is general information, usually factual and statistical, about an entire class of
events. For instance, if we knew that only 5 per cent of university lecturers gave truly awful lectures,
or that only 7 per cent of social security recipients preferred being on the dole to working, this would
be base-rate information. Research shows that people chronically underuse this information in making
inferences, particularly when more concrete anecdotal cases exist. people often fail to see the
relevance of base-rate information, relative to other information, to the inference task (Bar-Hillel,
1980). People increase their use of base-rate information when it is made clear that it is more relevant
than other information (e.g. case studies) to the inferential task.

Judgements of covariation are judgements of how strongly two things are related.

When people assume that a relationship exists between two variables, they tend to overestimate the
degree of correlation or see a correlation where none actually exists. This phenomenon, called
illusory correlation. Chapman reasoned that there are two bases for illusory correlation: associative
meaning (items are seen as belonging together because they ‘ought’ to, on the basis of prior expecta-
tions) and paired distinctiveness (items are thought to go together because they share some unusual
feature). Distinctiveness-based illusory correlation may help to explain stereotyping, particularly
negative stereotypes of minority groups.

In real life, negative events are distinctive because they are perceived to be rarer than posi- tive events
(Parducci, 1968), and minority groups are distinctive because people often have relatively few
contacts with them. Thus, the conditions for distinctiveness-based illusory correlation are met. There
is also evidence for an associative-meaning basis to negative ste- reotyping of minority groups: people
have preconceptions that negative attributes go with minority groups.

People use heuristics to reduce complex problem-solving to simpler judgemental operations. The
three key heuristics are: (1) representativeness, (2) availability and (3) anchoring and adjustment.
The representativeness heuristic is a relevance judgement that disregards base-rate information,
sample size, quality of information and other normative principles. For example, consider the
following information: ‘Steve is very shy and withdrawn, invariably helpful, but with little interest in
people, or in the world of reality. A meek and tidy soul, he has a need for order and structure, and a
pas- sion for detail’. The representativeness heuristic would quickly lead to the inference that Steve is
a librarian rather than, say, a farmer, surgeon or trapeze artist, and in general that would probably be
correct.

The availability heuristic is used to infer the frequency or likelihood of an event on the basis of how
quickly instances or associations come to mind. Where instances are readily available, we tend to
inflate frequencies. For example, exposure to many media reports of violent Muslim extremists will
make that information available and will tend to inflate our estimate of the overall frequency of
violent Muslims. Similarly, in forming an impression of Paul, who has short hair, wears big boots and
carries a cane, you might overestimate the likeli- hood that he will be violent because you have just
seen the film A Clockwork Orange.

Anchoring and adjustment is a heuristic that ties inferences to initial standards. So, for example,
inferences about other people are often anchored in beliefs about ourselves: we decide how intelligent,
artistic or kind someone else is with reference to our own self-schema.