For many
years, researchers have been conducting assessment tests, to identify different
aspects of personality. According to (Cohen & Swerdlik.
2010. p. l-20), this construct
is defined as “an individual’s unique constellation of psychological traits and
states, including aspects of values, interests, attitudes, worldview,
acculturation, sense of personal identity, sense of humor, cognitive and
behavioral styles, and related characteristics.” However, (Cloninger, Allen,
Friedman & Shustac) define personality as “the underlying causes within the
person of individual behavior and experience” (p. 2). There have also been
several individuals who have created theories which explain personality. One
particular man was Sigmund
Freud, and his highly recognized theory included 5 stages of personality
development. These were as follows:
1. Oral - The child’s focus is the
mouth which includes sucking, chewing, biting and eating.
2. Anal - During this stage, the
child’s focus will shift to the anal region.
3. Phallic - This is when the child becomes interested in sex.
4. Latency period - “libidinal energy
is submerged or expressed” (Bukatko. 2008. p. 15).
5. Genital - During the final stage
of development, mature genital satisfaction will occur.
Although,
Sigmund Freud was considered the father of psychoanalysis, and modern
personality theory, there was one individual who also provided a theory about
personality development that deviated from Freud’s original work in several
ways. His name was Erik Erikson and his theory focused on 8 stages of
development. These included the following:
1. Trust Versus Mistrust - will feel secure with others in the world or
unable to trust others.
2. Autonomy vs. Shame and Doubt - feel confident/independent or dependent
on others.
3. Initiative vs. Guilt - secure in the ability to lead others or will
lack initiative and follow.
4. Industry vs. Inferiority - begin feeling confident in abilities or
inferior if not acquired.
5. Identity vs. Role Confusion - self-identity is formed or confusion if
it’s not obtained.
6. Intimacy vs. Isolation - will explore intimate relations with others
or avoid intimacy and experience loneliness and isolation.
7. Generativity vs. Stagnation - will develop a sense of belonging by
giving back and being productive or experience stagnation if accomplishing
nothing or being nonproductive.
8. Ego Integrity vs. Despair - During this time, one develops integrity
because he or she is fulfilled with life or despair due to feelings of
dissatisfaction.
Regardless, of how one’s personality may develop, researchers
have also determined that it is very important to measure the level of
construct validity associated with a particular testing method, when attempting
to assess this construct. The reason for this is because it allows
the opportunity to analyze how well a test instrument measures this construct,
when compared to similar testing methods. If researchers are unable to support
construct validity, then the assessment method should not be used, because
final data may not be valid. Many researchers also believe that it is necessary
to gather evidence of convergent and discriminant validity when measuring construct
validity. This is because according to (Zechmeister,
Zechmeister, & Shaughnessy. 2001), “convergent validity identifies “the degree of how two measures with the
same construct go together,” while “discriminant validity determines the extent
of how the new measure correlates badly with measures that don’t have the same
construct (p. 343). With this in
mind, the purpose of this paper will be to address how I can establish
convergent and divergent validation in the test construction of my new test instrument
for personality, and how these
validation factors were explored when constructing a specific test instrument called
the Minnesota Multiphasic Personality Inventory – 2 (MMPI-2).
How can I establish
convergent validation in test construction for a new instrument?
According to (Campbell, Fiske. 1959. p. 104), one way that I can establish
convergent and discriminant validation for my new instrument that measures
personality is by “utilizing a matrix of intercorrelations
among tests representing at least two traits, each measured by at least two
methods.” This is also known as a multitrait-multimethod matrix. Therefore, the
first thing that I will do to complete this process is gather evidence which establishes
convergent validity, so I can include this in my test manual. This validation
process can also be accomplished by administering two testing methods, to 50
adults within my sample. The reason that multiple tests will be administered is
because I can acquire scores for two testing methods which are used to assess
personality, while also measuring the data in a different way. The test methods
which will be used are also as follows:
Personalities R Us Scale (PRUS) - This
is my new instrument, which is a paper and pencil method that measures
personality by scoring answers that are based on 40 true or false questions. These
questions also cover several different aspects of personality in an attempt to
identify abnormal to normal personality traits for each individual test-taker.
The test results are then determined based on a where the final score falls on
a range scale. This scale also goes from 1 to 10, with 4 through 8 being the
normal range.
Thematic
Apperception Test (TAT) - This projective test was developed by Henry Murray in
1938 and it includes 31 vague picture cards. The examiner performs this test by
asking the client to respond to each specific card. The client’s response must
also include a story which has a
beginning, middle and ending. The test results are then determined based on the
clients story.
Once
this step is completed, I will than calculate the scores of both testing
methods to identify any possible correlations between these testing methods. If
the testing methods identify high correlations, than I will be able to establish
and/or present evidence of convergent validation to include in my test manual.
This is because a high correlation between these methods, will confirm that
both assessment instruments appear to measure the construct of personality.
How can I
establish divergent validation in test construction for personality?
Although, I might establish convergent
validity by following the steps above, this may not be enough evidence to confirm
a high degree of construct validity when the PRUS is used to measure
personality. Therefore, I will also need
to establish evidence of discriminant validity,
which supports divergent validity. When utilizing a matrix
of intercorrelations among testing methods, which is suggested by (Campbell,
Fiske. 1959), I can establish discriminant validity that supports divergent
validity, by measuring two different constructs. Therefore, I will establish
evidence of discriminant validity for the PRUS by administering a third test to the
same adults who initially took that and the TAT for personality. For this
particular case, I will use The Emotion Regulation Questionnaire (ERQ) which
assesses the construct of mood. The reason for this choice is because it “is designed to assess individual differences in the
habitual use of two emotion regulation strategies: cognitive reappraisal and
expressive suppression” (Gross,
John. 2003. p. 1). The overall purpose of this process is to also try and
demonstrate that the PRUS does measure personality versus a variable that is
linked or similar, like mood. This can be confirmed by calculating scores on
the PRUS and ERQ which should have a very low correlation. For example, if I
discover that the correlation score between these two testing methods is .05
then I will establish evidence of discriminant validity. I will then be able to
also include this validation data in my test manual for the PRUS. Although, I
would choose the above steps to establish evidence of convergent validity and
discriminant validity which supports divergent validation, there are more ways
that this process can be completed. One particular case includes how this was
explored for the Minnesota Multiphasic Personality Inventory (MMPI-2).
How were convergent and divergent
validation explored when constructing the MMPI-2?
In order
to establish a higher degree of convergent and divergent validation to support
construct validity, test constructers first analyzed information that was
acquired via the MMPI. The MMPI-2 was then revised by updating several aspects,
which included the addition of three validity scales that identify further
traits associated with the construct of abnormal behavior. According to (Cohen, R. J., & Swerdlik. 2010.
p. 411), the MMPI-2
is still designed to assess abnormal behavior but now “contains a total of 567
true-false items, including 394 items that are identical to the original MMPI
items, 66 items that were modified or rewritten, and 107 new items.” The age
range of test takers was changed from 14 years of age to 18 years of age and
the standardization sample was also modified to be more representative. When
constructing this revised testing method, the 10 clinical scales which were
used in the original MMPI are still present, however; they are now identified
as score numbers versus the original names. These original names, along with
the new references, are as follows:
MMPI = Hypochondriasis (Hs) / MMPI-2 =
Score 1
MMPI = Depression (D) / MMPI-2 = Score 2
MMPI = Hysteria (Hy) / MMPI-2 = Score 3
MMPI = Psychopathic Deviate (Pd) / MMPI-2 =
Score 4
MMPI = Masculinity/Femininity (Mf) / MMPI-2
= Score 5
MMPI = Paranoia (Pa) / MMPI-2 = Score 6
MMPI = Psychasthenia (Pt) / MMPI-2 = Score
7
MMPI = Schizophrenia (Sc) / MMPI-2 = Score
8
MMPI = Hypomania (Ma) / MMPI-2 = Score 9
MMPI = Social Introversion (Si) / MMPI-2 =
Score 10
Although, these
10 scales remained the same other than being changed to number references,
there are several revisions that were also made to the validity scales which
were included with the original MMPI. The original validity scales which
measure things like general test-taking attitude and test truthfulness remained
the same. These are as follows:
The Lie
Scale (L) - This is used to determine what individuals deliberately lie
when taking the test. This can also determine when people may be exaggerating
themselves to appear to be better than any other person.
The
K Scale - This 30
item scale is designed to identify psychopathology in those individuals who may
have profiles that usually fall within a normal range. This includes
measurements that are acquired when assessing, family and interpersonal
relationships and self-control. Test-takers who receive high scores on this
scale are also viewed as being defensive.
The ? Scale – This is also identified as the “cannot say” scale and
its score represents how many items were left unanswered on the test. The test
manual which is associated with the MMPI also recommends that if a test has 30
or more unanswered questions, it should be declared as invalid.
Although, these
scales remained the same, content component scales and three more validity
scales were added to the MMPI-2. These were also added so further traits which
may be associated with abnormal behavior can be measured simultaneously. According
to (Cherry. 2012. p. 2), the new scales are as follows:
The TRIN Scale
– This “was developed to detect
patients who respond inconsistently. This section consists of 23 paired
questions that are opposite of each other.”
The VRIN Scale – This is known as the Variable Response Inconsistency Scale and it is also used to measure and/or identify responses that may be inconsistent.
The Fb Scale (Fb) – This 40 item scale is used to measure the same F scale issues, but only during the last half of the test. When and individual scores high on this scale, it can mean that he or she lost interest and simply began answering the questions randomly.
The VRIN Scale – This is known as the Variable Response Inconsistency Scale and it is also used to measure and/or identify responses that may be inconsistent.
The Fb Scale (Fb) – This 40 item scale is used to measure the same F scale issues, but only during the last half of the test. When and individual scores high on this scale, it can mean that he or she lost interest and simply began answering the questions randomly.
The
MMPI-2, it is also scored via an interpretive report, which is constructed by
the trained psychologist. These scores are then converted to normalized “T
scores” which are based on a scale that ranges from 30 to 120. The “normal” T
score range on this test should be between 50 and 65. Any scores that are above
or below this range may also be considered significant. These particular tests will
then be interpreted by the trained psychologist. Since, revisions were made by
utilizing the MMPI and new scales were added to measure further traits that may
be associated with abnormal behavior, the MMPi-2 also establishes a higher
degree of overall construct validity. This is because the test construction
process was carried out while addressing and establishing evidence of both
convergent and discriminant validation (to support divergent validity).
Therefore, when assessing the construct of abnormal behavior, this may be a
more appropriate testing method to use when attempting to measure more accurate
and/or valid results,
Summary
For many years, researchers have been
interested in the fascinating construct of personality and how it should be
defined. There are also several well-known theories concerning personality
development, by individuals like Sigmund Freud and Erik Erickson. Although,
there has been vast research conducted to address different aspects of
personality, researchers have determined that it always important to analyze
and/or measure the
degree of construct validity associated with a particular testing method. This is
to ensure that a chosen test instrument will measure the construct
of personality, at the same degree as similar testing methods. Many also
believe that evidence which supports convergent and discriminant validity
should be explored when measuring construct validity. This is because it will
identify “the degree of how two measures
with the same construct go together,” and “the extent of how the new measure
correlates badly with measures that don’t have the same construct (Zechmeister,
Zechmeister & Shaughnessy. 2001. 343). Therefore, the second goal of this
paper was to provide a step by step process to establish
convergent and divergent validation in the test construction for personality,
according to (Campbell, Fiske. 1959). Finally, the third goal of this paper was
to also discuss how these validation factors were explored, when constructing a
widely used assessment instrument called the Minnesota Multiphasic Personality
Inventory – 2 MMPI-2. This process included several revisions that were made
when compared to the original test version known as the MMPI. Since, these
revisions and/or updates were made to address construct validity, the MMPI-2
appears to be a better test method to use when assessing or measuring the
construct of abnormal behavior.
References:
Bukatko, D.
(2008) Child and adolescent development,
a chronological approach. Ohio; Cengage Learning.
Cohen, R. J., & Swerdlik, M. E. (2010). Psychological testing and assessment: An introduction to tests and measurement. Boston, MA: McGraw-Hill Higher Education.
Cloninger, C. S., Allen, P. B., Friedman, S. H., Schustack, W. M. Past & present views of personality.
Custom Edition for Kaplan University. Pearson Custom Publishing. Boston,
MA.
Zechmeister, Zechmeister, Shaughnessy (2001). Essentials of research methods in psychology.
New York, NY: The McGraw-Hill Companies, Inc.
Campbell,
D. T., Fiske, D. W. (1959). Convergent
and discriminant validation by the
multitrait-multimethod matrix. Psychological Bulletin, 56(2),
81-105. Retrieved via the Kaplan Library. Resource No. AN bul-56-2-81
Gross, J.J., & John, O.P. (2003). Individual differences in two emotion
regulation processes: Implications for affect, relationships, and well-being.
Journal of Personality and Social Psychology, 85, 348-362.
Cherry,
K. (2012). About.com.Psychology: The Minnesota Multiphasic Personality Inventory - MMPI-2. Retrieved
via the World Wide Web at http://psychology.about.com/od/psychologicaltesting/a/mmpi_3.htm