DEVELOPING INSTRUMENTS TO

ASSESS PERSONALITY
 Tool such as logic, theory and data reduction
methods (such as factor analysis) are frequently used
in the process of developing personality tests.

Another tool in the test development process may be

a criterion group.
LOGIC AND REASON
There is a place for logic and reason in psychology,
at least when it comes to writing items for a
personality test.
Logic and reason may dictate what content is
covered by the items. Indeed, the use of logic and
reason in the development of test item is sometimes
referred to as the content and content-oriented
approach to test development,
 As an example of the content approach to test
development, suppose we create a new test, for the
purpose of identifying young people who are at risk of
developing chronic insomnia.
The test may be called “ A Test to Identify Risk of
Chronic Insomnia among Young People” or TIRCIYP.
Logically, the content of the test items would relate to
what is known about this sleeping disorder. In writing
the items for the test, one may rely on what he/she
knows about the disorder from reading of psychological
and medical books, personal experience, and accounts of
people who suffer from it.
 The test may end up a yes-no questions, a sample
might be like these:
1. Is your current sleeping pattern not more than 4
hours of?
2. Do you fear not being able to sleep at all?
3. Do you perceive your sleeping pattern abnormal
in any way?
 The items of the sample test are probably based on
DSM criteria for sleeping disorders. Whether or not
this content-oriented test ever enjoys widespread use
will depend on the number of factors, not at least of
which is how well it measures chronic sleeping
problem or what it is that it purports to measure.
 Efforts to develop such content-oriented, face-
valid items can be traced at least as far back as an
instrument used to screen World War 1 recruits for
personality and adjustment problems.
The Personal Data Sheet (Woodworth, 1917)
later known as the Woodworth Psychoneurotic
Inventory, contained items designed to elicit self-
report of fears, sleep disorders, and other problems
deemed symptomatic of psychoneuroticism. The
greater the number of problems reported, the more
psychoneurotic the respondent was presumed to be.
 A great deal of clinically actionable information
can be collected in relatively little time using such
self-report instruments—provided of course, that the
test-taker has the requisite insight and responds with
candor.
Typical companions to logic and reason, and
intuition in item development are research and
clinical experience. Another possible aid in the test
development process is correspondence with experts
on the subject matter of the test. And yet another
tool, even the guiding force is Psychological Theory.
THEORY
As noted, personality measures differ in the extent to
which they rely on a particular theory of personality in their
development and interpretation.
For example, in Psychoanalytic theory can be a guiding
force for the test example for Chronic Insomnia, then the
test items could be quite different. So based on
psychoanalytic notion people who experience chronic un-
sleepiness are unconsciously disturbed by many things in
their life that they could not even get to sleep.
DATA REDUCTION METHODS
Date reduction methods represent another class of
widely used tools in contemporary test development.
Data reduction methods include several types of
statistical techniques collectively known as factor
analysis or cluster analysis.
 One use of data reduction methods in the design of
personality measures is “ to aid in the identification
of the minimum number of variables or factors that
account for the inter-correlations in observed
phenomena.
 As a result of a pioneering research program in the
1940s, Raymond Bernard Cattells’ answer to this question
was “16”.
Cattell reviewed previous research by Allport abd Odbert
(1936), which suggested that there were more than 18,000
personality trait names and terms in the English language.
Of these, however, only about a quarter were “real traits of
personality” or words and terms that designated
“generalized and personalized determining tendencies—
consistent and stable modes of an individual’s adjustment to
his environment… and not merely temporary and specific
behavior.” (Allport, 1937).
 Cattell added to the list some trait names and
terms employed in the professional psychology and
psychiatric literature and then had judges rate “just
distinguishable” differences between all the words.
The result was a reduction in the size of the list to
171 trait names and terms. College students were
then asked to rate their friends with respect to these
trait names and terms, and the factor-analyzed results
of that rating further reduced the number of names
and terms to 36, which Cattell referred to as “surface
traits”
 Still more research indicated that 16 basic
dimentions or “source traits” could be distilled. In
1949, Cattell’s research culminated in the
publication of a test called the Sixteen Personality
Factors (16 PF) Questionnaire.
Revisions of the test were published in 1956,
1968, 1993. In the year 2002, supplemental updated
norms were published.
CRITERION GROUPS
A criterion may be defined as a standard n which a
judgment or decision can be made. With regard to scale
development, a criterion group is a reference group of test-
takers who share specific characteristics and whose
responses to test items serve as a standard according to
which items will be included in or discarded from the final
version of a scale.
 The process of using criterion groups to develop
test items is referred as empirical criterion keying
because the scoring or keying of items has been
demonstrated empirically to differentiate among
groups of test-takers.
The shared characteristics of the criterion group to
be researched– a psychiatric diagnosis, a unique skill
or ability, a genetic aberration, or whatever– will vary
as a function of the nature and scope of the test.
 Development of a test by means of empirical
criterion keying may be summed up as :
1. Create a large, preliminary pool of test items from
which the test items for the final form of the test will
be selected.
2. Administer the preliminary pool of items to at least
two groups of people.
* group 1 : a criterion group composed of people
known to possess the trait being measured.
* group 2 : a randomly selected group of people
(who may or may not possess the trait being measured).
 3. Conduct an item analysis to select items
indicative of membership in the criterion group.
Items in the preliminary pool that discriminate
between membership in the two groups in a
statistically significant fashion will be retained and
incorporated in the final form of the test.

4. Obtain data on test performance from a

standardization sample of test-takers who are
representative of the population from which future
test-takers will take.
 The test performance data for Group 2 members on
items incorporated into the final form of the test may
be used for this purpose if deemed appropriate. The
performance of Group 2 members on the test would
then become the standard against which future test-
takers will be evaluated.
After the mean performance of Group 2 members
on the individual items (or scales) of the test has been
identified, future test-takers will be evaluated in terms
of the extent to which their scored deviate in either
direction from the Group 2 mean.