Chapter 4

The psychological processes underlying

selection

This chapter sets out to examine some of the psychological

processes involved in selection. Such processes will affect, to a
greater or lesser degree, everyone involved in the selection process:
people designing selection systems, people actively involved in
selecting candidates, and the candidates themselves. In discussing
these processes no definitive account of the ways in which they
affect the people involved in selection, and the selection process,
will be offered. The main reason for this is that there is surprisingly
little research investigating the existence and effects of the well-
established psychological processes discussed in this section within
the selection process. So when reading this material, consider for
yourself how the various stages of the selection process might be
affected and influenced by the processes covered.

Objectives
When you have completed this chapter you should be able to:
„ describe research findings of the evidence of the influence of non-verbal
communication in selection processes
„ discuss ways in which stereotyping may affect the selection process
„ outline what is meant by attribution theory and describe the ways in which
attributions, and attributional biases, may impact the selection process
„ describe the ways in which people may combine different sources of
information when making decisions at various stages of the selection process.

The effects of non-verbal information on judgements

Non-verbal information provides a wealth of potentially useful
material about what people are like. This information includes
both physical characteristics such as the colour of a person’s skin,
the shape of their body, their height, and the features of their face,
and also dynamic phenomena such as their way of walking, the way
they hold themselves, their gestures, and their facial expressions.
There has for many years been evidence that people believe that
the physical features of people are linked to their personality.
Secord (1958) describes a study in which people were given
descriptions of two men and then asked to rate their facial features.
The descriptions were:

Birkbeck University of London © 2009

1
Postgraduate studies with the Department of Organizational Psychology

This man is ruthless and brutal. He is extremely hostile, quick

tempered and overbearing. He is well known for his boorish and
vulgar manner and is a very domineering and unsympathetic
character.

This man is warm-hearted and honest. He has a good sense of

humour and is intelligent and unbiased in his opinion. He is
responsible and self-confident with an air of refinement.

Do you find that images of the physical appearance of these two

people form in your mind as you read the descriptions? Well, the
subjects in Secord’s study certainly believed that there was a likely
link between the personalities and the physical feature of the men,
rating them differently on 25 of 32 physiognomic characteristics.
Overall, the person described positively was viewed as having
average features, whereas the man described negatively was viewed
as being likely to have abnormal features.
What sorts of processes may be at work when the character of a
person is judged from their appearance? Secord (1958) suggested
that there are five such processes.
1 Temporal extension is said to occur when a fleeting
expression is judged to represent an enduring feature of
someone’s behaviour. So if you see someone smile once or twice
when you first meet them you may leap to the conclusion that
they smile a lot of the time.
2 Parataxis occurs when the characteristics of an individual who
is known are generalized to someone who is not known. This
might occur if an individual meets a new acquaintance, and
something about this new acquaintance is reminiscent of an
uncle. In these circumstances the uncle’s personal
characteristics might be generalized to the new acquaintance.
3 Categorization occurs if someone looks at an individual’s
facial features, and, on the basis of this, places the individual in
a category of people about which they have a set of beliefs. For
example, you might look at someone’s face and decide that they
are of Middle-Eastern origin. As a consequence you may then
attribute to them the personality characteristics you associate
with Middle-Eastern people.
4 Functional quality is the idea that we link the functions of
various parts of the face to elements of personality. So the eyes
are for seeing, and if you notice that someone has big eyes you
might assume that they have the ability to read your
personality, and the personality of others, relatively easily.
5 Metaphorical generalization occurs when the several facial
features are considered together, and inferences made from
them. A man with a big smile, a round face, and well-groomed
hair might be viewed as good-natured, and this might in turn
lead to the assumption that he is also generous.
Whatever the validity of this early attempt to classify the
psychological processes which lead us to link the physical features
of faces to the personalities of their owners, it does draw our
attention to the fact that these process may be varied and complex.
But it is not only faces that are associated with dispositions. There
is evidence that people believe that the physique of people – the
shape of their bodies – is also linked to their dispositions. Sheldon
(1940) examined photographs of thousands of people and

2 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

concluded that there are three basic body shapes: the ectomorph
who is thin with little muscle, the endomorph who is round and fat,
and the mesomorph who is muscled and athletic. Ryckman,
Robbins et al. (1989) asked students to rate these three body types
on various personality characteristics and found that there were
clear and systematic differences between the way that they viewed
them. For example, ectomorphs were viewed as intelligent and
neat, and mesomorphs were seen as healthier and more hard-
working than endomorphs.
Other research shows that a person’s height (Wilson, 1968),
unusual physical features (Bull, 1979), the accent in their voice,
their hairstyle, and their choice of clothes (Argyle, 1988), how
physically attractive they are (Kalick, 1988), and whether or not
they wear spectacles (Argyle & McHenry, 1971), are all used by
people as a guide to the nature of their personality.
It should be noted that physical characteristics affect judgements
about more than personality. In studying attractiveness Kalick
(1988) found that being attractive is also associated with greater
social status, and specifically with having a better job, having been
to a better college, and earning more money.
Watkins and Johnston (2000) showed that such factors may, at
least in certain circumstances, have a real impact on selection.
They investigated the impact of physical attractiveness and
curriculum vitae quality on the evaluation of job applicants in the
screening phase of the selection process. One hundred and eighty
participants were asked to imagine they were a recruiting officer
and to screen an application for the position of graduate trainee
manager. Participants read a job advertisement and one of two
versions of a curriculum vitae, which differed in quality. Attached
to the front page of each curriculum vitae was a passport-sized
head-and-shoulders photograph of either an average or an
attractive female. A control condition with no attached photograph
was also included. Participants judged the likelihood with which
they would offer an interview to the applicant, the quality of the
application, and the likely starting salary they would offer the
applicant. Results indicated that attractiveness had no impact
when the quality of the application was high, but that
attractiveness was an advantage when the application was average.
When the curriculum vitae quality was average the attractive
applicant was evaluated more positively than the control, no
photograph, applicant.
Evidence for the impact of applicant attractiveness on selection
also comes from a study by Jackson, Hunter et al. (Jackson, 1995).
The authors carried out a study of the relationship between
attractiveness and perceived and actual competence, and found
that although there was no actual relationship between
attractiveness and competence in adults, attractive people were
perceived as more competent than less attractive people. This
effect was stronger for males than for females (that is, males were
more likely to associate competence with physical attractiveness),
and stronger when explicit information about competence was
absent than when it was present.

Birkbeck University of London © 2009 3

Postgraduate studies with the Department of Organizational Psychology

The implications of the research findings

What implications do all of these findings have for selection and
assessment? Selection methods can be distinguished on the basis
of the extent to which the participants actively form impressions of
one another.
At one end of the continuum is the more traditional type of biodata
questionnaire in which information is collected about the
demographic characteristics and the personal history of job
applicants. This information is then processed mechanically
according to a set of rules established through statistical analysis of
the variables which predict good job performance. The person
involved in this traditional biodata-based selection process does
not have to interpret or make sense of information about the job
candidate – they merely have to feed it into the pre-existing bio-
data information processing system, and this will generate a
number representing the job applicant’s predicted job
performance.
At the other end of the continuum is the unstructured interview.
Here the interviewer is given a considerable amount of freedom to
decide what sort of questions they ask the applicant. As a result the
interviewer is likely to collect different sorts of information from
different job applicants. Furthermore, the information the
interviewer receives from each applicant needs to be actively
interpreted. If the applicant says that they left their last job
because they felt that they needed a new challenge, the interviewer
needs to decide what this means. Does it mean that the last job was
very boring, and that anybody would have wanted to leave it after a
while? Or does it mean that the applicant is someone who gets
bored easily, and so is likely to become quickly bored in any new
job and so perform badly and leave? Such active interpretation is
not only being made by the person doing the interviewing – it is
also being carried out by the interviewee. The interviewee is
forming an impression of the interviewer, perhaps trying to decide
what the interviewer is looking for, and then trying to say the
things they assume the interviewer wants to hear. In short, the
interview involves an intense and dynamic social interaction in
which those involved are playing roles, in which they have at least
partially different goals, and in which they are actively seeking to
form an impression of each other. In forming an impression they
need to decide what sort of information is relevant and important,
and then to combine different pieces of information in order to
build up an overall view.
Other types of selection process can be placed somewhere on the
continuum between the traditional biodata form (processed
mechanically and with little or no active involvement or
interpretation by the person doing the selecting), and the
unstructured interview (which involves a considerable amount of
active involvement and active interpretation by both the person
doing the selecting and the job applicant).
More information is required about the psychological processes
which take place when selection methods that actively involve the
person doing the selecting are used. That is, more information is
required about the processes involved when the person doing the

4 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

selecting forms an impression of the job applicant (and vice versa).

It is helpful to know whether these processes are likely to lead to
accurate or inaccurate results; and it is also helpful to know what,
if anything, can be done to improve the accuracy of these processes
and their capacity to provide the best possible assessment of the
suitability of a job applicant. It is to these issues that attention will
now be turned.

Stereotypes and stereotyping

When people deal with everyday objects in their lives, they rely
heavily on their ability to categorize. Categories refer to sets of
objects which we group together. Most of the time these objects are
not identical, but by giving them a common name, storing
information about what they have in common, and treating them
in roughly the same way, people are able to make life infinitely
easier than it would otherwise be. Consider for a moment what the
consequences would be if categories for common objects like chairs
and tables were not used. Every time someone wanted to sit down
they would have to consider the characteristics of every available
object to see whether it would support their weight, be of roughly
the right size, etc. But armed with the category ‘chair’, all they have
to do is recognize that an object is a member of the chair category
to know that they will be able to use it to sit on. So once someone
sees that an object has some or all of the features common to chairs
(a flat and horizontal platform around 12 inches in diameter, four
legs, a back, etc.), they are able to recognize it as a chair and
behave towards it in a similar way to any other object recognized as
a chair.
Categories are used not only for everyday objects, but also for
people. At work reference is made to categories of job roles such
cleaners, middle managers, chief executive officers, accountants,
and also to other categories of people not associated with specific
roles, such as women, the middle-aged, the disabled, Asians,
Australians, ambitious people and so on. Recognizing that an
object is a car gives us information about something that is not
directly observable (that it has an engine, that it can transport
someone from one place to another, that it requires some kind of
fuel to do so, etc.), and allows people to orient their behaviour
around it (if it does not belong to us we should not touch it, but if it
does we can get into it and go somewhere). In just the same way,
recognizing that someone is a member of a category of people such
as doctors allows an individual, at least in principle, to tell certain
things about them (for example, that they are probably middle-
class, reasonably wealthy, intelligent, knowledgeable about
medical matters, etc.) and organize their behaviour around them
(for example, tell them about our symptoms if we are feeling ill).
When categories refer to people rather than objects, the
information about these categories is often referred to as a
stereotype. So there are stereotypes of cleaners, accountants,
Australians, and the middle-aged, and armed with these
stereotypes people are in a position to know some useful things
about someone just by recognizing to which category they belong.
Clearly these stereotypes have some potential advantages for those

Birkbeck University of London © 2009 5

Postgraduate studies with the Department of Organizational Psychology

involved in selection. If by classifying someone as a middle-aged

woman from an ethnic minority a person is able to infer that she
has a variety of physical and psychological characteristics, and
these may be of considerable help in deciding upon her suitability
for the job for which she is applying.
This will of course only work if the stereotypes being drawn on
here, the stereotypes of women, the middle-aged, and people from
ethnic minorities, are accurate descriptions of all, or almost all, of
the people to whom they refer. If part of the stereotype that
someone holds about women is that they are poor at taking tough
business decisions, this will influence that person in deciding
whether a man or a woman would be suitable for a job that will
involve making such tough business decisions. However, this will
only work as long as the stereotype is accurate. If, in reality there is
a considerable amount of variation with regard to the extent to
which women are able to take tough decisions, or if the proportion
of women who are able to take tough decisions is no different from
the proportion of men who are able to do so, applying the
stereotype to a particular woman is likely to lead to a poor selection
decision.
Certainly there is clear evidence that people do have stereotypes of
social groups. The first study of the content of stereotypes was
carried out 70 years ago by Katz and Braly (1933). They gave North
American college students a list of ten national/ethnic groups (for
example, German, Turkish, Jewish) and a list of 84 adjectives. The
students were required to select the five adjectives which they
thought best characterized each of the ten groups. The researchers
found that there was a considerable amount of agreement between
the students about the adjectives describing the groups. For
example:
„ Italians were described as artistic (53%), impulsive (44%),
passionate (37%), quick-tempered (35%) musical (32%), and
imaginative (30%)
„ Americans as industrious (48%), intelligent (47%), materialistic
(33%), and ambitious (33%)
„ the English as sportsmanlike (53%), intelligent (46%),
conventional (34%), tradition-loving (31%), and conservative
(30%)
„ Negroes (sic) as superstitious (84%), lazy (75%), happy-go-
lucky (38%) and ignorant (34%)
„ Germans as scientifically-minded (78%), industrious (65%),
stolid (44%) and intelligent (32%).
Although the content of these stereotypes may have changed
somewhat over the years, there is little doubt that stereotypes still
exist. For example, in a study of the gender stereotypes held by
college students Nesbitt and Penn (2000) found that male and
female participants showed very high agreement about the typical
characteristics of men and women.
Research on occupational stereotypes has largely focused on the
extent to which they are gender-typed, and there is a wealth of
evidence that occupations are associated with males or females
(Kulik, 1999; Truss, Goffee, & Jones, 1995), and even that children
as young as four years old are aware of the gendered nature of

6 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

occupations (Trice & Rush, 1995). Certainly there is evidence that

gender stereotypes of occupational types do exist. For example,
Beggs and Doolittle (1993) examined college students’ perceptions
of whether 129 different occupations were associated with men or
women. They found the students viewed occupations on a
continuum from the most male to the most female, although
women tended to perceive jobs as being more neutral than men
did.
There is also evidence that stereotypes can influence the job
selection process. Vanvianen and Willemsen (1992) studied the
influence of sex stereotypes on the evaluation of job applicants for
higher-level technical and academic jobs in the Netherlands. They
found that the applicant most similar in traits to the ideal applicant
was hired for each job, and that job interviewers viewed ideal job
applicants as tending to have masculine rather than feminine
traits. Consistent with this, when males and females were regarded
as having the same qualifications for the job, the male applicants
were accepted more often. A later study by Vanvianen and
Vanschie (1995) showed that when males and females are
interviewed they are assessed not only on the basis of stereotypes
based on their gender, but also on their behaviour. Specifically,
they found that women who demonstrated both stereotypically
masculine and feminine behaviours were more likely to be
accepted as an eligible candidate by a selection board than women
who demonstrated only stereotypically feminine behaviours.
There is also evidence that race and ethnic stereotypes can
influence selection. King et al. (2006) presented fictions resumes
(CVs) to 155 white Americans and asked them to judge the suitably
of the people described for jobs. They manipulated the race of the
fictitious applicant (Asian, Black, Hispanic, White) the quality of
their resume (high quality versus low quality), and the status of the
job they were applying for (high versus low). The White and
Hispanic applicants were found to be more likely to be evaluated
positively if they had a high quality resume, but Asian Americans
were evaluated highly for high status jobs irrespective of the
quality of their resume, and Black applicants were evaluated
negatively irrespective of resume quality.
Research by Davison and Burke (2000) and Finkelstein, Burke et
al. (1995) suggests that the effect of stereotypes on the selection
decision may depend on the amount of job-relevant information
provided. Davison and Burke examined various factors affecting
sex discrimination in simulated employment settings, and found
that while both female and male applicants received lower ratings
when being considered for an opposite-sex-type job, the difference
between ratings of males and females decreased as more job-
relevant information was provided. Finkelstein, Burke et al. found
that younger raters tended to give less favourable ratings to older
workers when they were not provided with job-relevant
information. They also showed that the extent to which stereotypes
are used may depend partly on whether those involved in selection
are choosing between applicants with contrasting backgrounds,
because the raters also gave less favourable ratings to older
workers when they concurrently rated old and young workers.

Birkbeck University of London © 2009 7

Postgraduate studies with the Department of Organizational Psychology

Perry, Kulik et al. (1996) have also carried out research on the
factors that inhibit or facilitate the use of stereotypes in a selection
context. They carried out a laboratory study on stereotypes held
about older workers. Undergraduate students were asked to
evaluate applicants in a simulated employment context. Results
indicated that the extent to which the age of the applicant played a
role in the selection decisions depended on whether the raters were
biased against older workers, whether or not they were cognitively
busy and thus had the cognitive resources to inhibit the use of
age-associated stereotypes, and whether or not the applicants were
applying for jobs congruent with their age.
This research seems to suggest that people involved in selection
can, in the right circumstances, suppress any unhelpful stereotypes
of applicants, and base their decisions on job-relevant information.
Such a conclusion is also supported by research by Frazer and
Wiersma (2001) who studied the effects of applicant race and
applicant quality on both employment interview decision making
and on the way in which participants later recalled information
about the applicants. Undergraduate students in the US reviewed
paper credentials and then interviewed a high- or low-quality black
or white applicant in a controlled experimental setting. The results
indicated that although the interviewers hired the black and white
applicants in equal proportion, one week later they recalled the
answers to the interview questions given by the black applicants as
having been significantly less intelligent than those of whites –
when in fact the actual interview content in the two race conditions
was identical. This suggests that the students’ stereotypes of black
applicants as less intelligent than whites was suppressed when they
made the selection decision, but nevertheless influenced the
information they later recalled about the applicants.

Attribution theory
When a person involved in selection reviews information about a
job applicant they often need to make inferences about its
implications. For example, imagine that you receive an application
form from a recent school leaver. You notice that they live locally
and that they live in a poor and deprived part of town. The school
they attended has a poor reputation, and the examination grades
they obtained were average. What are you to make of this? On one
hand you could conclude that the mediocre examination results
indicate that this person is of no more than average ability, and
place their application form in the ‘no’ pile. On the other hand, you
could decide that the examination grades were only average
because this person was disadvantaged by being brought up in a
deprived area and attending a poor school, and that you should at
least give them the benefit of an interview.
If you decide that the reason why the applicant received mediocre
exam grades is because they are only of average ability, you are
making an internal attribution. That is, you are attributing the
cause of their mediocre performance to factors internal to them,
such as their average academic ability, or their less than impressive
level of motivation to study. If, on the other hand, you decide that
the reason they performed at only an average level is their deprived

8 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

childhood, or the poor quality of the school they attended, you are
making an external attribution. Here you are attributing the
cause of their examination performance not to their internal
dispositions, but rather to external causes over which they have
little on no control. In this circumstance, and many similar
circumstances with which those involved in selection may be faced,
whether or not the cause of an event is attributed internally or
externally has clear and significant implications for selection
decisions.
The first psychologist to draw attention to the attribution process
was Heider (1958). Heider argued that people need to make sense
of the world they live in and, to the extent that they can, control the
events that take place around them. So in social interactions with
other people individuals are motivated to explain why these people
do the things they do. This is because with this information it is
possible not only to predict how others will behave in future but
also to organize our own behaviour towards them successfully. All
people, said Heider, are naïve scientists. Just as scientists seek to
understand the causes of natural phenomena so that they can
predict how these phenomena will behave in future, so ordinary
people as naïve scientists are trying to understand the causes of the
behaviour of others so that they can predict how others will behave
in future.
Heider’s initial ideas about the attribution process were developed
by Jones and Davis (1965) with their correspondence
inference theory and by Kelley (1967) with his covariation
model. These theories were concerned with explaining the
processes by which we decide how to attribute the cause of a social
event, and in particular how we decide whether to explain the
event internally (it is due to some characteristic of the person
involved) or externally (it is due to something in the environment).
However, the development of attribution theory which will be
focused upon here is probably the one most relevant to selection,
and concerns how we explain the successes and failures of others.
So, as in the example given in the introduction to attribution
theory, how is it possible to decide whether to explain the successes
or failures of a job applicant, whether in their distant past, or
during the actual selection process such as an interview or
assessment centre, as being due to internal or external causes?
This issue was addressed by Weiner (1979). Weiner suggested that
our decision to attribute the cause of a success or failure depends
on two important factors:
„ locus of causality (internal or external)
„ stability (stable or unstable over time).
Locus of causality refers to where the cause of an event is situated.
For example, does someone laugh because they are happy (the
locus of causality is internal to them) or because someone has just
told them a funny joke (the locus of causality is external)? Stability
refers to whether the event is stable over time. Someone who is
always conscientious is showing stable conscientious behaviour,
whereas someone who is only conscientiousness now and again is
not. It is the combination of these two factors, locus of causality

Birkbeck University of London © 2009 9

Postgraduate studies with the Department of Organizational Psychology

and stability which, according to Weiner, leads us to make an

internal or external attribution for failure or success.

10 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

Table 4.1 Ways of explaining success or failure

Stable Unstable
Internal Ability Effort
External Task difficulty Luck

Source: based on Weiner, 1979

Table 4.1 shows how, according to Weiner, the attribution which

somebody makes to explain someone else’s success or failure
depends on their evaluation of whether the factors leading to the
success are stable or unstable, and internal or external.
If you believe that a woman called Jane performed poorly at an
interview for stable and internal reasons, you might decide that she
was lacking in ability. If you decide that the reason for her poor
performance was unstable and internal, you might attribute her
poor performance to a lack of interest in the job. If your attribution
is a stable and external one, you might say that the questions put to
all the candidates, including Jane, are particularly difficult. Finally,
if you assume that the cause of her disappointing performance is
unstable and external, you might decide that she was unlucky –
perhaps because she was not feeling well at the time of the
interview. You should be able to think of four equivalent types of
explanation that might be made if Jane’s interview performance
had been particularly good rather than particularly bad.

Activity 4.1
Imagine that you are very impressed by the performance of a job applicant when
you interview them. Depending on whether you view their impressive performance
as arising from stable or unstable factors, and internal or external ones, you might
explain this success in four different ways:
„ stable/internal
„ unstable/internal
„ stable/external
„ unstable/external.
What type of explanation might you give for the success in each of these four
cases?

Comment
If you use a stable/internal explanation for their performance, you might say that
the person is a very able and clever job candidate. If you use an unstable/internal
explanation you might say that they are were highly motivated to do well at the
interview. If you adopted stable/external you might say that you had made the
interview questions for all the candidates very easy, and if it was unstable/external
perhaps that the interviewee had been lucky to find you in a good mood when you
carried out the interview.

Birkbeck University of London © 2009 11

Postgraduate studies with the Department of Organizational Psychology

What influences us to make particular sorts of attributions? That

is, why might you make an unstable/internal attribution for one
person’s performance, and a stable/internal attribution for that of
someone else? There is some evidence that the nature of our
attributions can be influenced by our stereotypes. So if I believe
that women are not particularly able intellectually, and I am
presented with an application form from a woman with good
examination results, I can attribute the cause of her success
externally (for example, her examiner was biased towards women,
the subject in which she was examined is easy, she got special
coaching, or she was lucky with the examination questions). If on
the other hand I am presented with a male with good examination
results I can explain this with internal and stable explanations (for
example, he is able and/or he is a hard worker). Evidence that such
differential attributions can actually take place was found by Deaux
(1976). She found that people tended to explain the success of a
man as being due to ability, whereas the success of women was
explained as being due to luck or temporary hard work.

Attribution errors
Having discussed the nature of the attributions that can be made
for success and failure, we will finish this section with a brief look
at two psychological phenomena which can lead to biases and
errors in the way that attributions are made. The first of these is
referred to as the fundamental attribution error. The
fundamental attribution error is ‘the tendency for attributers to
underestimate the impact of situational factors and to
underestimate the role of dispositional factors in controlling
behaviour (L. Ross, 1977). This is best explained with an example.
Imagine that you are interviewing someone for a job, and you
notice that at times they mumble and seem incoherent. Why has
this happened? Is it because the person is nervous (a situational
attribution) or because they are generally unable to express
themselves clearly (a dispositional attribution)? The fundamental
attribution error, for which there is a great deal of experimental
evidence, suggests that you will tend to underestimate the impact
of the situation in causing the person’s behaviour (for example, the
person is not being very clear because they are nervous in the
formal interview situation) and overestimate the role of
dispositional factors (you will tend to see their incoherence as
being due to personal characteristics such as poor communication
ability).
Another phenomenon that can bias attributions is the false
consensus effect. This is the tendency for people to overestimate
how common their own attitudes, beliefs and opinions are. In a
classic study by Ross, Greene et al. (1977)passers-by at a university
were asked to walk around for 30 minutes wearing a sandwich
board which read ‘Eat at Joe’s’. Of 80 people asked to do this, 48
agreed and 32 declined. After they had indicated whether or not
they would carry the board, they were asked whether they thought
other people would do so. Of those that had agreed to carry the
board, about two-thirds thought that others would do likewise.
Similarly, about two-thirds of those who had declined to wear the
board thought that others would say no. Basically, therefore, these
people seemed to believe that most people would behave in the way

12 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

that they themselves chose to behave. Further studies have

demonstrated that we also tend to overestimate the extent to which
others have opinions similar to our own. The false consensus effect
could in principle affect the outcome of various aspects of the
selection process. For example, if you are interviewing someone
who expresses an opinion different from your own you may
wrongly assume that your own opinion is considerably more
common than it actually is (and that conversely the opinion held
by the person being interviewed is rarer than it is in reality).
In conclusion, there are a variety of stages during the selection
process at which attributions play an important part. Selection is
not a purely mechanical and technical process but rather a process
in which subjective judgements, and attributions, play a significant
role. Focusing on the nature of the attributional process, and the
biases which can occur during attribution, provides us with helpful
insights into the psychology of selection.

Combining information
As a result of non-verbal information, the use of stereotypes, or the
active attribution of the causes of behaviour, the person involved in
selection may infer that a job applicant has certain characteristics
relevant to whether or not they should be appointed. How do they
combine this information to arrive at an overall impression of the
job applicant? For example, if they infer that an applicant lacks
ambition and is over-sensitive, but is also warm, and intelligent,
how do they put this information together so as to arrive at a
decision about whether or not they should be appointed?

Asch’s research on information combination

Classic research on the question of how people combine
information to form an impression was carried out by Asch (1946).
Asch gave participants in his study a list of seven personality traits,
and asked them to form an impression of a fictitious ‘target-person’
who had these characteristics. He found that not only were they
able to do this with little difficulty, but that they were able to
generate other characteristics that the target-person probably had,
and guess what the target-person looked like and what job they
did! This suggested that people are very good at combining
information to create an overall impression of someone else.
By varying the presentation of the words in the list that he gave his
participants, Asch discovered several additional phenomena. First,
he changed some of the words in the list of personality traits, and
measured the effect that this had on the impression formed by the
participants. In one case he presented the description ‘intelligent,
skilled, industrious, warm, determined, practical, cautious’ to one
group of participants, and to another he gave the same list except
that the word ‘warm’ was replaced by the word ‘cold’. The
substitution of this one word had a very substantial impact on the
impression that the two groups formed. Participants given the list
with the word ‘warm’ indicated that the target-person was likely to
be generous, happy, good-natured, and sociable, whereas the group
given the list including the word ‘cold’ described them as
‘ungenerous, unhappy, irritable, and unsociable’. However, when

Birkbeck University of London © 2009 13

Postgraduate studies with the Department of Organizational Psychology

Asch substituted other words (for example, he substituted the word

‘polite’ in a list with the word ‘blunt’), this had a far less significant
impact on the impression that the participants formed of the
target-person. Asch concluded that some traits are relatively
central in the impression formation process, whereas others are
relatively peripheral. Central traits, such as ‘warm’ and ‘cold’, have
a particularly marked influence on the impressions, whereas the
effect of peripheral ones such as ‘polite’ and ‘blunt’ is less marked.
He also claimed that the meaning given to a particular trait when
people apply it to somebody else is influenced by the other traits
applied to them. So the meaning of the trait ‘calm’ when presented
in the list ‘kind, wise, honest, calm, strong’ is different from the
meaning of the same word when it is presented in the list ‘cold,
shrewd, unscrupulous, calm, strong’.
Finally, Asch examined the effect of varying the order in which the
traits were presented. In one case he presented the list ‘intelligent,
industrious, impulsive, critical, sullen, envious’ to one group of
participants, and to another the same list in reverse order:
‘envious, sullen, critical, impulsive, industrious, intelligent’.
Although the same traits were used in both lists, the order of
presentation had a marked effect on the impression that
participants formed of the target-person. When the more positive
traits were presented first, a more favourable impression of the
target-person was formed than when the negative traits were
presented first. Asch referred to this as the primacy effect, and
believed that as the participants read through the list the meaning
that the traits were given depended on those that had come before
them. So the meaning of the negative traits in the list in which they
appeared first were given a different meaning by the participants
from the same negative traits which were presented after the
positive ones.
Anderson (1965)challenged Asch’s belief that when an impression
of someone is formed people build up a holistic picture of them,
with the meaning of one trait dependent on the meaning of the
other traits we attribute to them. Anderson suggested that the way
traits are combined to form an impression can be described by a
fairly simple mathematical model which he referred to as the
weighted averages model.

Anderson’s weighted averages model

According to the weighted averages model, the extent to which an
individual likes someone can be exactly predicted from a
knowledge of the traits attributed to them. The mathematical
formula for the model is as follows:

Iw0 + T1w1 + T2 w2 + ... + Tk wk

R=
w0 + w1 + w2 + ... + wk
Where I is the initial impression of the target person, w0 is the
weight given to this initial impression, T1is the first new trait
attributed to the person and w1 is the weight of this first new trait,
T2 is the next new trait and so on with Tk being the last new trait. R
is the overall degree of liking or disliking for the person after all
this information has been combined.

14 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

This is easier to understand with an example. Imagine that you

know someone called John. John works in the same organization
as you, and you have come into contact with him a few times but
do not know him well. Nevertheless, on the few occasions you have
met John he has seemed quite a nice person. If the scale of +1 to +5
for the degree to which you like someone is used, and a scale of –1
to –5 for the extent to which you dislike someone is used, your
moderately favourable impression of John might be, say, +2. Now,
because this impression is based on a few brief encounters with
John rather than any opportunities to get to know him well, it
might be given a fairly small weight. If a scale of 0 to 1 for the
amount of weight you give the initial impression is used, the
weighting in the case of John might be 0.2. So your initial
impression of John is moderately positive (+2) but because you are
not yet very certain of what he is like this impression has a small
weight (0.2).
Now, imagine that you unexpectedly find yourself paired with John
on a project. You have to work with him on a daily basis for three
weeks, and during this time you learn that he is industrious and
clever, although not very perceptive and rather shallow. You are
somebody who values hard work, and so on the scale of +5 to –5
you view John’s industriousness as +4. The values you give to
John’s cleverness, perceptiveness, and shallowness are +3, –2, and
–3. Because you see a lot of John you are fairly sure that these
characteristics are all true of him, and so you give them all weights
of 0.8. If these values were entered into the equation for the
weighted averages models you would get:

(2 × 0.2) + (4 × 0.8) + (3 × 0.8) + (–2 × 0.8) + (–3 × 0.8)

R=
0.2 + 0.8 + 0.8 + 0.8 + 0.8

R = 0.59.
So Anderson’s weighted average model assumes that trait
information is combined in a rational way, with an initial
impression being modified by new information, and the degree to
which each piece of information is positive (or negative) is
combined as a function of how much importance or weight we give
to it. In the above example this means that someone who attached
the stated values and weights to the initial impression, and to each
trait, would end up with a mildly positive (0.59) view of John.

Asch or Anderson? – Pavelchak’s study

Whereas Asch believed that the meaning of a trait changed as a
function of the others with which it was compared, and that a
holistic impression was formed which was essentially different
from the sum of its parts, Anderson argued that the meaning of the
traits remained the same when they were combined and that the
final impression was formed as a result of a rational process which
could be described by a mathematical model. Over the following
years many studies were carried out to try show that one theory or
the other was the correct one. Some of these studies provided
evidence which tended to support Asch’s theory, whereas others
found evidence in support of Anderson’s model. Eventually
Pavelchak (1989) carried out a study which appeared to show that

Birkbeck University of London © 2009 15

Postgraduate studies with the Department of Organizational Psychology

Asch’s and Anderson’s models of impression formation were

actually complementary to each other.
There were two parts to Pavelchak’s study. In the first part
participants were asked to rate 35 academic subjects (such as
history, art, engineering) and 50 personality traits in terms of how
likeable they were. In the second part, which took place a few days
later, the same participants were asked to assess how likeable a
person was when they were described with four traits (for example,
bright, studious, precise, methodical) and also to guess the
academic subject they were studying. Half the participants, the
‘category’ group, were asked to guess what academic subject the
person was studying before assessing how likeable they were, and
the other ‘piecemeal’ group were asked to evaluate the person
before they guessed the academic subject they were studying.
Pavelchak argued that if, as Asch would suggest, the participants
were forming a holistic integrated impression of the person, they
would base their assessment of the person’s likeability on the
likeability of the academic major they were studying. However, if
they were building up a piecemeal image of the person, as
Anderson would suggest, their assessment of the person’s
likeability would be a function of how likeable the four traits were.
The results of the study showed that in the category condition
likeability estimates tended to be based on how likeable the
academic subject was, but in the piecemeal condition likeability
was based on how likeable the combination of four traits were.
Pavelchak concluded that the work of Asch and Anderson was
complementary rather than antithetical, that the theories that each
had developed successfully predicted how people combined
information, but that whether one theory or the other was
applicable depended on the situation with which people were
presented.

The continuum model

More recently, in an effort to understand when people will base
their impressions of each other by placing them in stereotype
categories (such as woman, banker, engineer, Indian, etc.), and
when they base them on a piecemeal integration of the various
pieces of information they have learnt about them, Fiske and
Neuberg (1990) have developed what they call the continuum
model of impression formation. According to this model, when an
individual first learns about someone else, they make a rapid initial
categorization of them, often based on simple cues such as the
colour of their skin, their accent, their job, or the clothes they are
wearing.
If the person is of no future relevance to the individual, the
impression may not be developed any further than this. However,
if there is reason to be more certain about the person’s character,
the individual will check their initial categorization of them against
any further information collected. For example, if on the basis of
the initial rapid categorization an individual decides that someone
is a middle-aged accountant, a category associated with various
personality traits, the individual would compare any further
information obtained about them to that individual’s stereotype of
middle-aged accountants. After doing so they will either conclude

16 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

that the initial categorization of the person was correct (their

behaviour conforms to the stereotype) or they will re-categorize
them into a more appropriate category (they might re-categorize
the person as a ‘zany’ type if they find that they often do wild and
wacky things). Only if no categories fit the information
accumulated about the person will the individual use a piecemeal
approach to integrate the individual pieces of information into a
non-stereotyped, and unique, impression.

Prototypes
Although the research discussed in this section has focused
primarily on how an impression of the extent to which someone is
liked is built up, it clearly has implications for the question of how
someone involved in selection forms an impression of the
suitability of a job candidate. For example, to what extent do those
involved in selection draw on stereotypes when reviewing
information about job candidates, and when do they use the
cognitive effort required to build up a unique impression of the
candidate? Although there has been little research on such issues, a
study by Perry (1994) is relevant here. Perry based his research on
the idea of prototypes. Prototype theory is concerned with how
information about categories of objects and people is represented
cognitively. The idea is that people have a representation of the
best possible example of a particular category (the best possible
example of the category ‘bird’ might be a small robin-like creature
sitting on a branch). When they judge whether an object is a
member of one category or another, they compare the properties of
the object in question to the prototype. In support of prototype
theory, research has shown that that the more similar an object is
to a prototype, the more quickly people are able to decide that it is
a category member. For example, people are able to decide that a
robin is a member of the category ‘birds’ more quickly than they
can decide that a penguin is a member of this category.
Perry suggests that people have prototypes for different
occupations (a prototypical doctor, a prototypical carpenter, etc.)
which he refers to as person-in-job prototypes, and that those
involved in selection evaluate applicants by matching information
about them to the person-in-job prototype associated with the job
for which they are applying. These person-in-job prototypes
comprise features that are more or less strongly associated with the
prototype. Perry refers to the features more strongly associated
with the prototype as central, and to others as more peripheral.
Consistent with this model, a laboratory study indicated that
applicants who matched on more central features of person-in-job
prototypes were evaluated more favourably than applicants who
matched on fewer central features. In addition, applicants who
matched on age were evaluated more favourably than applicants
who did not match on age when age was a central but not a
peripheral feature of a person-in-job prototype (though this
finding was not replicated with gender).

Birkbeck University of London © 2009 17

Postgraduate studies with the Department of Organizational Psychology

Stringency and Leniency (Hawks and Doves) Effects

There is clear evidence that people differ in the extent to which

they are stringent (hawk-like) or lenient (dove-like) when they
appraise people and evaluate them. These leniency-stringency
effects have been found for a variety of rating roles, including
assessment centre raters (Bartels & Doverspike, 1997), medical
examiners (Harasym, Woloschuk, & Cunning, 2008; McManus,
Thompson, & Mollon, 2006; Raymond, Webb, & Houston, 1991),
people engaged in self-appraisal (Farh & Werbel, 1986; Fox, Caspy,
& Reisler, 1994), students rating their instructors (Greenwald &
Gillmore, 1997), real job performance appraisals (Kane, Bernardin,
Villanova, & Peyrefitte, 1995) and simulated performance
appraisals (Villanova, Bernardin, Dahmus, & Sims, 1993).

The tendency to be relatively stringent or lenient when evaluating

others appears to be quite stable across time, different rating
systems, and different rating formats (Kane et al., 1995).
Furthermore, there is evidence that the extent to which ratings are
influenced by the extent to which a rater tends to be a hawk or a
dove is substantial. For example, large-scale studies of medical
examiners McManus et al. (2006) found that 12% of the variation
in examination marks was associated with the tendency of
examiners to be “hawks” or “doves”, and Harasym (2008) f0und
the variation in marks given was considerably more to do with who
the examiner was than with the assessed ability of the candidate.
Clearly a systematic tendency for assessors to be lenient or
stringent has important implications for selection and assessment.
To take just one example, it is likely that good candidates will be
wrongly rejected by over-stringent selectors, and that poor
candidates will be accepted by overly-lenient ones.
Why is it that some assessors tend to be more lenient that others?
There is some evidence that the personality of the rater may be
important. In a study using the 16PF personality measure to
examine the stringency of the marks given by raters, Bartels and
Doverspike (1997) found that relatively “tenderminded” and
“warmhearted” raters tended to be more lenient. Similar findings
were found in a study by Bernadin, Cooke and Villanova (2000)
which looked at the relationships between scores on the Big 5
personality factors and leniency. This found that relatively lenient
raters tended to have low scores on conscientiousness and high
scores on agreeability.
If rating leniency is a result of stable individual differnerces in the
form of personality this raises some important issues for selection
and assessment. One such implication is that it may be difficult to
train people to be less stringent or less lenient, and another is that
steps should be taken to mitigate the effects of stringeny-leniency
effects, perhaps by systematically adjusting evaluations to
compensate, perhaps by selecting raters who are not likely to be
prone to either extreme leniency or extreme stringency in the way
they evaluate, or perhaps by ensuring that ratees or job candidates

18 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

are seen by assessors who are mixed in relation to their tendency to

be lenient or stringent in assessment.

How influential are these Psychological Processes in Selection

and Assessment?

At present there is a lack of research examining where, in the

various stages of the selection process, these processes may be
operating, what their effects might be, and what, if anything, can
be done to minimize any damaging effects they have on selection.
In the absence of such research it is worthwhile considering the
stages in the selection process at which each of the processes might
operate, and what the effects of these processes might be. Such
stages include job analysis, the development of specific selection
methods for use in a particular case (for example, the development
of a biodata questionnaire), and the actual operation of the
selection system including both the people involved in carrying out
the selection, and the people being selected.

Activity 4.2
Read the following articles which you will find on Blackboard:

Morgeson, F. P., & Campion, M. A. (1997). Social and cognitive

sources of potential inaccuracy in job analysis. Journal of Applied
Psychology, 82(5), 627-655.

Although this article focuses on job analysis, the social and cognitive sources of
inaccuracy are relevant to other stages of the selection process, and with different
selection techniques. For example, consider the ways in which these processes
may affect structured and unstructured interviews.

Summary
In this chapter some psychological processes which play a part in the selection
process were discussed. These processes included the effects of non-verbal
information on judgements, stereotypes and stereotyping, attributions, the
methods by which people combine information when making judgements about
one another, protortypes and systematic differences in leniency-stringency.
Although there is a paucity of research directly examining the operation of these
processes in the applied setting of selection and assessment, the substantial
theoretical foundations and extensive laboratory-based research on them suggest
that they are likely to play a significant role at various stages of the selection
process.

Birkbeck University of London © 2009 19

Postgraduate studies with the Department of Organizational Psychology

References

Anderson, N. H. (1965). Averaging versus adding as a stimulus-

combination rule in impression formation. Journal of
Experimental Psychology, 70, 394-400.
Argyle, M. (1988). Bodily communication (2nd ed.). London:
Methuen.
Argyle, M., & McHenry. (1971). Do spectacles really affect
judgements of intelligence? British Journal of Social and
Clinical Psychology, 10, 27-29.
Asch, S. E. (1946). Forming impressions of personality. Journal of
Abnormal and Social Psychology, 41, 258-290.
Bartels, L. K., & Doverspike, D. (1997). Assessing the assessor: The
relationship of assessor personality to leniency in
assessment center ratings. Journal of Social Behavior and
Personality, 12(5), 179-190.
Beggs, J. M., & Doolittle, D. C. (1993). Perceptions now and then of
occupational sex typing - a replication of shinar 1975 study.
Journal Of Applied Social Psychology, 23(17), 1435-1453.
Bernardin, H. J., Cooke, D. K., & Villanova, P. (2000).
Conscientiousness and agreeableness as predictors of
rating leniency. Journal Of Applied Psychology, 85(2),
232-236.
Bull, R. (1979). The psychological significance of facial deformity.
In M. Cook & G. Wilson (Eds.), Love and attraction.
Oxford: Pergamon.
Davison, H. K., & Burke, M. J. (2000). Sex discrimination in
simulated employment contexts: A meta- analytic
investigation. Journal Of Vocational Behavior, 56(2), 225-
248.
Deaux, K. (1976). The behaviour of men and women. Monterey,
CA: Brooks-Cole.
Farh, J. L., & Werbel, J. D. (1986). Effects of purpose of the
appraisal and expectation of validation on self-appraisal
leniency. Journal Of Applied Psychology, 71(3), 527-529.
Finkelstein, L. M., Burke, M. J., & Raju, N. S. (1995). Age-
discrimination in simulated employment contexts - an
integrative analysis. Journal Of Applied Psychology,
80(6), 652-663.
Fiske, S. T., & Neuberg, S. L. (1990). A continuum of impression-
formation, from category-based to individuating processes
- influences of information and motivation on attention
and interpretation. Advances in Experimental Social
Psychology, 23, 1-74.
Fox, S., Caspy, T., & Reisler, A. (1994). Variables affecting leniency,
halo and validity of self-appraisal. Journal of Occupational
and Organizational Psychology, 67, 45-56.
Frazer, R. A., & Wiersma, U. J. (2001). Prejudice versus
discrimination in the employment interview: We may hire
equally, but our memories harbour prejudice. Human
Relations, 54(2), 173-191.
Greenwald, A. G., & Gillmore, G. M. (1997). Grading leniency is a
removable contaminant of student ratings. American
Psychologist, 52(11), 1209-1217.
Harasym, P., Woloschuk, W., & Cunning, L. (2008). Undesired
variance due to examiner stringency/leniency effect in

20 Birkbeck University of London © 2009

 Selection and Assessment Chapter 4

communication skill scores assessed in osces. Advances in

Health Sciences Education, 13(5), 617-632.
Heider, F. (1958). The psychology of interpersonal relations. New
York: Wiley.
Jackson, C. J. (1995). Assessing important and observable personal
qualities in the general selection interview. European
Journal of Psychological Assessment, 11(2), 75-80.
Jones, E. E., & Davis, K. E. (1965). From acts to dispositions: The
attribution process in person perception. In L. Berkowitz
(Ed.), Advances in experimental social psychology (Vol.
2nd). New York: Academic Press.
Kalick, S. M. (1988). Physical attractiveness as a status cue.
Journal Of Experimental Social Psychology, 24, 469-489.
Kane, J. S., Bernardin, H. J., Villanova, P., & Peyrefitte, J. (1995).
Stability of rater leniency - 3 studies. Academy of
Management Journal, 38(4), 1036-1051.
Katz, D., & Braly, K. W. (1933). Racial prejudice and racial
stereotypes. Journal of Abnormal and Social Psychology,
30, 175-193.
Kelley, H. H. (1967). Attribution theory in social psychology.
Nebraska Symposium on Motivation, 15, 192-238.
King, E. B., Madera, J. M., Hebl, M. R., Knight, J. L., & Mendoza,
S. A. (2006). What's in a name? A multiracial investigation
of the role of occupational stereotypes in selection
decisions. Journal Of Applied Social Psychology, 36(5),
1145-1159.
Kulik, L. (1999). Gendered personality disposition and gender role
attitudes among israeli students. Journal Of Social
Psychology, 139(6), 736-747.
McManus, I. C., Thompson, M., & Mollon, J. (2006). Assessment
of examiner leniency and stringency ('hawk-dove effect') in
the mrcp(uk) clinical examination (paces) using multi-
facet rasch modelling. BMC Med Educ, 6, 42.
Nesbitt, M. N., & Penn, N. E. (2000). Gender stereotypes after
thirty years: A replication of rosenkrantz, et al. (1968).
Psychological Reports, 87(2), 493-511.
Pavelchak, M. A. (1989). Piecemeal and category-based evaluation:
An ideographic analysis. Journal of Personality and Social
Psychology, 56, 354-363.
Perry, E. (1994). A prototype matching approach to understanding
the role of applicant gender and age in the evaluation of
job applicants. Journal Of Applied Social Psychology,
24(16), 1433-1473.
Perry, E. L., Kulik, C. T., & Bourhis, A. C. (1996). Moderating
effects of personal and contextual factors in age
discrimination. Journal Of Applied Psychology, 81(6),
628-647.
Raymond, M. R., Webb, L. C., & Houston, W. M. (1991). Correcting
performance-rating errors in oral examinations.
Evaluation & the Health Professions, 14(1), 100-122.
Ross, L. (1977). The intuitive psychologist and his shortcomings:
Distortions in the attribution process. In L. Berkowitz
(Ed.), Advances in experimental social psychology (Vol.
10, pp. 174-221).
Ross, L., Greene, D., & House, P. (1977). The false consensus
phenomenon: An attributional bias in self-perception and

Birkbeck University of London © 2009 21

Postgraduate studies with the Department of Organizational Psychology

social perception processes. Journal Of Experimental

Social Psychology, 13, 279-301.
Ryckman, R. M., Robbins, M. A., Kaczor, L. M., & Gold, J. A.
(1989). Male and female raters' stereotypes of male and
female physiques. Personality And Social Psychology
Bulletin, 15(244-251).
Secord, P. F. (1958). Facial features and inference processes in
interpersonal perception. In R. Taguiri & L. Petrullo
(Eds.), Person perception and interpersonal behavior.
Stanford, CA: Stanford University Press.
Sheldon, W. H. (1940). The varieties of human physique: An
introduction to constitutional psychology. New York:
Harper.
Trice, A. D., & Rush, K. (1995). Sex-stereotyping in 4-year-olds
occupational aspirations. Perceptual And Motor Skills,
81(2), 701-702.
Truss, C., Goffee, R., & Jones, G. (1995). Segregated occupations
and gender stereotyping - a study of secretarial work in
europe. Human Relations, 48(11), 1331-1354.
Vanvianen, A. E. M., & Vanschie, E. C. M. (1995). Assessment of
male and female behavior in the employment interview.
Journal of Community & Applied Social Psychology, 5(4),
243-257.
Vanvianen, A. E. M., & Willemsen, T. M. (1992). The employment
interview - the role of sex stereotypes in the evaluation of
male and female job applicants in the netherlands. Journal
Of Applied Social Psychology, 22(6), 471-491.
Villanova, P., Bernardin, H. J., Dahmus, S. A., & Sims, R. L. (1993).
Rater leniency and performance-appraisal discomfort.
Educational and Psychological Measurement, 53(3), 789-
799.
Watkins, L. M., & Johnston, L. (2000). Screening job applicants:
The impact of physical attractiveness and application
quality. International Journal Of Selection And
Assessment, 8(2), 76-84.
Weiner, B. (1979). A theory of motivation for some classroom
experiments. Journal of Educational Psychology, 71(3-
25).
Wilson, P. R. (1968). Perceptual distorition of height as a function
of ascribed academic status. Journal Of Social Psychology,
74(97-102).

22 Birkbeck University of London © 2009