In this lesson, we will examine the personal construct
approach developed by George Kelly. George Kelly is often considered the
father or grandfather of the more modern cognitive approaches that will
be described in the subsequent chapters.
Slide two. A key assumption shared by all of the cognitivist approaches
is that our experience is a map of the world. We do not react to the
world directly but to our map or construal of the world, so by
understanding how people experience the world, we can understand their
reactions. It follows that people will react similarly to events that
they construe or interpret as similar; and people will react differently
to events that they construe as different. For example, imagine you see
two different students giving class presentations. One student is
reacting to that situation by becoming extremely anxious and nervous and
inhibited. Another student is reacting to that same situation by
becoming excited and confident and enthusiastic. The cognitive
perspective says that the reason why they’re reacting differently to the
same situation is that they’re experiencing the situation differently;
one student might be seeing it as an opportunity for shame and failure
and the other student seeing it as an opportunity for positive attention
and accolades. Now let’s go on and look specifically at George Kelly’s
approach.
Slide three. George Kelly suggested that the construals or
interpretations we have of our world are based on personal constructs,
which are bipolar distinctions, such as shy versus bold, nice versus
mean, hard versus easy. There is, for each construct, two poles. The
pole that we’re currently applying to whatever it is we’re trying to
understand is called the emergent pole. The pole that we’re not applying
is called the implicit pole. So if while watching one student give a
presentation, you think that person is shy, then the emergent pole is
shy, and the implicit pole is bold. If while watching another student
you think “that person is bold”, then the emergent pole is bold, the
implicit pole is shy. Regardless, in both cases Kelly suggests that
whenever making a judgment we are making a distinction; that whenever we
are thinking that someone is shy, we are also implicitly saying they are
not that other thing they could be, they are not bold.
Slide four. Kelly argued that each person has a unique construct system,
a unique set of bipolar distinctions that they use to understand their
world. How may people’s construct systems differ? First, they may use
different constructs. For example, imagine two students, John and Jane,
who are both attending the first day of a class taught by Professor X.
John might think “Professor X is nice” while Jane might think “Professor
X is hard”. They have used different constructs; but they may not
disagree with each other. After class, John may say to Jane “Professor X
seems nice”, and Jane may say “yeah, sure”. And Jane may say to John
“Professor X seems hard”, and John may say “yeah, sure”. So, they may
accept what the other person says, but spontaneously the distinctions or
constructs they tend to use are different; that is, John tends to use
the construct nice versus mean, and Jane tends to use the construct hard
versus easy. A second way that individuals might differ is with respect
to how they organize their constructs. For example, consider the term
intellectual, which for John and Jane is a superordinate construct
meaning that it subsumes or implies other more specific constructs.
Imagine that for Jane, intellectual implies someone who is open,
questioning and thoughtful; whereas for John the same term intellectual
implies someone who is aloof, smart and boring. So when Jane says “I
think that Professor X is an intellectual”, she is implying that
Professor X is open, questioning, thoughtful. But John would interpret
her statement as implying that Professor X is aloof, smart and boring.
So they use the same terms, but what the terms mean are quite different.
Thus, by virtue of having different construct systems, John and Jane
have different experiences and reactions; indeed, because we all have
different construct systems, we all have different experiences and
reactions. But some construct systems are more similar than others, and
people with more similar construct systems tend to have smoother
communication. For example, Jane and John by using the term intellectual
in such different ways may end up having misunderstandings that become
awkward. Indeed, research using Kelly’s personal construct approach has
shown not only that strangers with more similar construct systems are
more likely to have smoother communications in laboratory studies, but
we can actually predict who in the real world are likely to become
friends and to remain friends over long periods of time on the basis of
the degree to which they share similar construct systems. So construct
systems are important. But how do we assess construct systems?
Slide five. Well your homework assignment shows you the most popular way
for assessing construct systems--George Kelly’s Role Construct Repertory
Test, usually called the REP test for short. The REP test helps identify
constructs people spontaneously use. While most commonly used to assess
the constructs people use to construe their social world, it can be
applied to any domain, from the constructs people use to compare
professors to the constructs people use to compare shampoo bottles!
Slide six. Here is one example of a REP test grid. Remember the rep test
asks you to indicate how two things are similar and a third thing is
different and to do that repeatedly within a set of objects or
individuals. In this case, the person did it on a set of former
boyfriends (as well as one current boyfriend). These are actually the
responses of one of my former students who graciously volunteered to
share her responses; this was done using the WebGrid II program which is
freely available on-line; first, she listed 8 former boyfriends (yes,
she had a busy social life); then, the program repeatedly showed her
random sets of three boyfriends and asked her how two were similar and a
third were different, and her job was to say the constructs that
spontaneously and immediately came to her mind. So let us imagine that
it first asked her how Fred and Shane and Scott were similar or
different and she said “Shane and Scott were good natured and Fred was
hostile”; so the first construct to emerge was good natured versus
hostile. The second construct to emerge was sane versus psycho. Third
construct was treated me nice versus did not treat me nice; and so on
down the line. So these are the constructs that spontaneously occurred
to her when she thought about how her boyfriends were similar to or
different from one another; in other words, these are some of the
important distinctions that she makes among her boyfriends. One cool
thing that the program then does are cluster analyses. The cluster
analysis on the right shows which constructs are closely related to each
other; for example, it shows that if a boyfriend is described as
clueless he is also described as close-minded and typically described as
an unmotivated stay-at-home guy as well. Thus, some constructs tend to
cluster together. Similarly, some boyfriends tend to cluster together in
her mind, as shown by the cluster analysis on the bottom. For example
George and Darren were described identically. And Scott and Troy were
described identically, and Shane differed from Scott and Troy by only
one construct.
Slide 7. A second interesting way to organize this information is to do
a factor analysis. The factor analysis shown here portrays the same
information as in the previous slide, but in a different way—it shows us
graphically how particular constructs or individuals converge or diverge
in her construct space. Again we see that Troy and Scott end up at the
same point, and Shane ends up near them both; and those three are all in
a very different place from James and Darren and George. The graph
clearly shows how Fred occupies a unique place in her construct
space—namely, the place defined by the constructs psycho, hostile, and
treat not nice. Her experiences with Fred have had a powerful impact on
her construct space; hostile and psycho may not have been part of her
construct space before Fred, but now those constructs are an important
anchor point. Whenever she meets a new potential boyfriend, she is
likely to consider whether or not this is going to be another Fred, that
is, whether this is going to be a psycho, hostile treat not nice guy or
a good sane treat nice guys. You might not be surprised to find out that
her current boyfriend is Ed. He’s the opposite of Fred. So, what she was
looking for in her new boyfriend was someone who was not-Fred, who was
at the opposite of Fred on each of these bipolar distinctions, that is,
someone who was good-natured, sane, and would treat her nice. Okay, let
us consider a second example.
Slide 8. In this case, I listed six restaurants in Moscow, Idaho; and
the program repeatedly showed me three restaurants and asked how are two
were similar and a third different. The first construct to occur to me
was off main street versus on main street. The second was mediocre
versus tasty. The third was chain versus non-chain, and so on. The
cluster analysis on the right shows that some constructs tended to used
in similar ways; for example, slow restaurants were typically expensive
eat-in restaurants. The cluster analysis on the bottom shows that some
restaurants tended to be construed in similar ways; for example, similar
constructs were applied to Taco Time and Subway, both being fast and
cheap.
Slide 9. Here is the factor analysis of the data from the previous
slide, and shows graphically how some constructs and restaurants
converge, such as quick and inexpensive places where people often just
get their food and go. The graph shows how this individual, who happens
to be me, might construe any new restaurant. The constructs that might
spontaneously occur to me are whether it is downtown or not, whether
it’s a chain restaurant or not, whether it is expensive or not, and so
on. Other people are likely to emphasize different distinctions,
different constructs, and consequently experience the restaurants in
different ways and make different choices. This is why the business
world has been so interested in the personal construct approach. The REP
test technique has been applied to all sorts of consumer commodities.
For example, earlier I mentioned one study I saw in which people were
shown sets of shampoo bottles and asked how two were similar and
different from a third; in this way, the researchers discovered what
distinctions people made when evaluating shampoo bottles, such looking
warm or looking smooth or looking professional or whatever.
Slide 10. The broader perspective that guided George Kelly was a view of
persons as scientists. He viewed all of us as scientists, seeking to
predict and understand events. Our construct system is really our
hypotheses about what are the important distinctions and how those
distinctions are organized. And like a scientist who retains hypotheses
that work and discards those that do not work, we keep constructs that
efficiently predict important recurrent events, and discard those that
do not. Specifically, if events do not fit our construct system, for
example if a boyfriend or a restaurant confounds our predictions, we
feel anxious and threatened and seek to revise our construct system to
make sense of what is happening.
Slide 11. We can revise our constructs in several ways. One way is to
change it’s range of convenience. Range of convenience means the range
of objects or situations or events to which the construct might apply.
Returning to our earlier example of Jane making sense of her new
Professor, Jane might initially use the construct hard versus easy to
predict everything about the professor—how hard the exams will be as
well as how hard professor X will be to interact with. But over time
Jane might find that while Professor X has hard exams that is not
predictive of their social style and so Jane may narrow the range of
convenience of the construct hard versus easy and only use it to predict
the exams and not use it to predict other aspects of the professor.
Another way to revise the construct is to change it’s focus of
convenience, that is, what are the things that it predicts best. So
whereas range refers to everything the construct can predict, focus
refers to its area of expertise. For example, the construct nice/mean
and the construct hard/easy may both predict how hard a professor’s
exams are; but hard/easy predicts the exams a lot better than does
nice/mean because that is its focus of convenience. Conversely, the
construct nice/mean and the construct hard/easy may both predict whether
a professor forgives a missed assignment due to a life crisis, but the
nice/mean construct predicts this better than does the hard/easy
construct because that’s nice/mean’s focus of convenience. Finally, Jane
might find that a particular construct should be thrown out entirely;
for example, Jane may find that nice and mean is just not useful because
she never meets anyone who is simply nice or simply mean; instead, she
finds that one person might very empathic but not very generous, while
another person is very generous but not very empathic. Consequently she
decides that nice versus mean is too broad and inefficient of a
predictor, and so simply replaces it with more precise constructs like
empathy and generosity.
Slide 12. Kelly was fond of the phrase constructive alternativism. What
this means is that since we use constructs to construct our world, we
can change our world by changing our constructs. Consequently, the goal
of Kelly’s therapeutic approach, which he called fixed role therapy, was
to facilitate construct system flexibility and development by showing
clients how their existing constructs were limiting their behaviors and
how changing their constructs could open new avenues for
self-fulfillment.
