HOW TO DEFINE & ASSESS PERSONALITY

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 (F^b) – 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