Chapter 4 Personality Assessment (Personality Psychology: Foundations and Findings

How can response sets be reduced?

- Reverse item scoring - Infrequency scale (lie scale) - Q-sort (item use limitation) - Anonymity assurance - Crowne-Marlow Social Desirability Scale

What are the pros of self-reports?

- provides access to thoughts and experiences - facilitates ease of administration and scoring.

Psychologists generally divide personality tests into two kinds.

1) Self-report personality tests (objective tests) 2) Performance-based personality tests (projective tests)

Response set (noncontent responding)

A habitual way of responding to a personality assessment that is not directly related to the questions asked; noncontent responding. When people have a set way that they tend to respond to self-report questions - either deliberately or unknowingly. Response sets can lead to false results. 1) faking good 2) faking bad 3) socially desirable responding 4) acquiescent responding 5) reactant responding 6) extreme responding 7) moderate responding 8) patterned responding 9) carelessness

Cronbach's alpha (a)

A measure of internal consistency reliability; the average correlation among all possible combinations of test items taking them half at a time. Take the correlation between the scores of two halves of a test. Calculate the average correlation of all possible halves of the test. That is what Cronbach's alpha reliability estimates: the generalizability of the score from one set of items to another. Most researchers use this statistic. They want the measures to have an alpha of .70 to .80.

Split-half reliability

A measure of internal consistency reliability; when each half of a test gives consistent results. Splitting a test in half and see if test-taker's scores on one half correlated with scores on the other half.

Parallel-forms reliability

A measure of internal consistency reliability; when two or more versions of a test give consistent results. Two versions test. Psychological testing developers would make up two versions of a test that were comparable and checked to see that the scores on the parallel forms of the test were similar.

Test-retest reliability

A measure of temporal consistency; when a test gives a consistent result from one point in time to a later point in time. Make sure that participants are not merely remembering what they originally said in the first test-taking session and eliminate practice effects where participants perform better merely because they've seen the test before. Careful for memorization and practice effects on doing this test.

Infrequency scale (lie scale)

A scale inserted into a personality assessment to identify people who may be using a response set. Helps identify pattern responding. Includes extremely rare questions. Often called lie scales because they are designed to identify a person who is lying.

Construction techniques

A type of performance-based test in which respondents create a story or a piece of artwork in response to an ambiguous stimulus. Ex: Thematic Apperception Test (TAT); Goodenough-Harris Draw-A-Person test (DAP)

Expression techniques

A type of performance-based test in which respondents express their thoughts and feelings through creative play or artwork. Ex: creative doll or puppet play; artwork

Completion techniques

A type of performance-based test in which respondents fill in the blanks with their own responses. Ex: sentence-completion tests

Arrangement techniques, or selection of stimuli

A type of performance-based test in which respondents move objects around or choose their favorites using ambiguous stimuli. Ex: pick your favorite color, picture, or other stimuli

Association techniques

A type of performance-based test in which respondents report their reactions to ambiguous stimuli. Ex: Word Association Test; Rorschach Inkblot Test

Overt/Clear purpose integrity tests

A type of personality assessment, often used during the hiring process, to test the honesty of job candidates in a way that is clearly stated and obvious to the test taker. Often has two parts: 1) One assesses attitudes towards dishonest behaviors 2) The second asks about theft and other illegal activities, such as drug use and gambling. Responses on overt tests are easy to distort to create a favorable impression or fake good on. A more subtle way of assessing integrity is to use disguised purposed integrity tests.

Disguised purpose integrity tests

A type of personality assessment, often used during the hiring process, to test the honesty of job candidates in a way that is hidden from the test taker. These are used because overt tests are easy to distort. Personality tests can assess characteristics related to a range of behaviors including disciplinary problems, violence on the job, absenteeism, tardiness, drug abuse, and theft.

Acquiescent responding

A type of response set in which respondents agree with nearly every question. Yea-saying - inflates test scores

Patterned responding

A type of response set in which respondents answer by making patterns on their response sheet. Ex: Make a pattern by marking their answers in a physical pattern like 1-2-3-4-5-4-3-2-1 or circling all 3s down the page.

Moderate responding

A type of response set in which respondents avoid the ends of a scale, choosing answers in the middle.

Extreme responding

A type of response set in which respondents avoid the middle of a scale, choosing answers on either end. Ex: Choosing mostly 1s and 5s on a 5-point Likert scale.

Reactant responding

A type of response set in which respondents disagree or say no to nearly every question. Nay-saying - deflates test scores

Counterproductive work behaviors

Actions that make it difficult or impossible for people to function in their jobs, including absenteeism, tardiness, turnover, accidents, and stealing.

Performance-based personality tests

Assessments in which people respond to unstructured stimuli, projecting their own meanings, significances, patterns, feelings, interpretations, concerns, or worldviews; sometimes called projective tests. Performance-based tests are more often used in clinical and forensic settings. Five categories of projective techniques: 1) Association techniques 2) Construction techniques 3) Completion techniques 4) Arrangement or selection of stimuli 5) Expression techniques

Self-report personality tests

Assessments in which respondents answer questions about themselves; sometimes called objective tests. Measures certain aspects of personality including well-being, values, and life goals. Self-report tests are frequently used for job selection and placement. Self-report personality measures may use: 1) Dichotomous two-choice scale 2) Likert-type rating scale 3) checklists 4) forced-choice scales 5) Visual analog scales Ex: Creative Personality Scale of the Adjective Check List, Locus of Control Scale, Machiavellianism Scale

Response sets can vary by culture

Canadian and American high school students showed more extreme responding compared to Japanese and Chinese students, a tendency that correlated with personal ratings of individualism. Asian students showed more moderate responding. Couldn't account for cultural differences in individualism-collectivism score.

Reverse scoring

Devised to identify and minimize response sets. Reverse scoring of some items prevents a person from getting a high score simply by agreeing or disagreeing with each of the items. Helps identify people engaging in acquiescent or reactant responding.

What can employers ask you and use?

Employers can use: 1) nonclinical tests 2) tests to measure the Five Factors or other personality traits

What can't employers ask of you or use?

Employers can't: 1) use clinical tests to screen 2) conduct medical exams 3) violate rights to privacy without demonstrated cause

Generalizability

Established boundaries or limitations of test. Establishes the limits of an assessment, the conditions, uses, and populations for which the measure is valid. For what purpose is this test valid? Generalizability establishes the boundaries or limitations of a test. Age, gender, race, and cultural background are just some of the population characteristics we must keep in mind when designing, validating, administering, and interpreting tests. Characteristics: - Cannot be administered for anything other than intended use. - Valid only for intended population.

Criterion validity

Establishes how good an assessment is by comparing the results to an external standard such as another personality test or some behavioral outcome.

Convergent validity

Establishes how good an assessment is by comparing the results to other tests of the same construct or to tests of related constructs in order to establish what the test measures. Converges with similar constructs (theories). Correlated with other measures of similar tests.

Discriminant validity

Establishes how good an assessment is by comparing the results to tests of theoretically unrelated constructs in order to establish what the test doesn't measure. Uncorrelated with measures of unrelated constructs. Taps a different concept entirely (specific measurements).

What are the cons of self-reports?

Faking good and faking bad. People may not be the best judges of their own skill, expertise, or knowledge, often overestimating their performance. People may also wish to present themselves in a certain light and answer in a way that jeopardizes the validity of the test.

Faking good

Falsely answering a personality assessment to appear better off in specific ways related to the outcome of the test (e.g., appearing more psychologically healthy, more qualified, more experienced, etc.)

Faking bad

Falsely answering a personality assessment to appear worse off in specific ways related to the outcome of the test (e.g., unqualified, in need of help, in need of special treatment, etc.).

Reliability

How consistent a measure is over time, items, or raters; an estimate of how consistent a test is: A good test gives consistent results over time, items, or raters. Generalizability across time, items, and raters. Reliability describes extent to which test scores are consistent and reproducible. 1) Temporal consistency reliability/ test-retest / across time 2) Internal consistency reliability / parallel forms, split-half, Cronbach's alpha / across items 3) Rater consistency reliability / interrater reliability / across raters

Validity

How well a test measures what it was designed to measure. Characteristics: - Grounded in research evidence - Correlated with some standard. 1) Construct validity 2) Face validity 3) Criterion validity 4) Convergent validity 5) Discriminant validity

Machiavellianism Scale

Named for Niccolo Machiavelli, author of The Prince, advising a fictional ruler on how best to acquire and stay in power. Machiavelli believed that most people were too trusting, not very smart, and could be taken advantage of for one's own purposes. Statements directly from his book were used by personality psychologists to design a force-based questionnaire to measure Machiavellianism.

Integrity testing

Personality assessments, often used during the hiring process, to test the honesty of job candidates (Are they likely to steal or cheat?). And to eliminate social desirable responding. 1/3 - About 1/3 of employers use some form of integrity testing. Two kinds of integrity tests: 1) Overt or clear purpose integrity tests 2) Personality-based measures Criticisms: 1) Reliability and validity data not open to scientific scrutiny. 2) Integrity compromised when test items and answers are leaked.

Psychological assessment and the business world

Personality tests can be used to aid in selection, training, and performance. They can help predict how people with counterproductive work behavior are likely to behave on average when it comes to absenteeism or any other work behavior. 1/3 - Used by about 1/3 of employers as part of hiring process. 28.5% screen for honesty/integrity 22% screen for potential for violence Predicts: 1) Openness - Likely response to training. 2) Conscientiousness - Index of basic performance (tardiness, absenteeism) 3) Friendliness, Courteousness, Responsiveness - success in customer service positions. 4) Counterproductive work behaviors.

What makes a good test?

Reliability, validity, and generalizability A good personality test must be: 1) valid 2) reliable 3) bias-free 4) theoretically sound and research-backed to confirm relationships to certain outcomes. 5) clear in specifications about conditions, populations, and cultures to which it applies.

What are the cons of self-report and performance tests?

Response sets (noncontent responding)

Socially desirable responding

Some people may portray themselves in an exaggeratedly positive manner. Falsely answering a personality assessment to appear more cooperative, likable, or socially appropriate, often by not admitting to distasteful, but perfectly human tendencies such as gossiping or having sexual urges. Harder for researchers to control for social desirability. Some strategies are: 1)Sometimes a forced-choice format makes it easier for respondents to answer truthfully. 2) Another method is to give participants a scale that measures their tendency for socially desirable responding and then use their scores to statistically control for social desirability in the score of interest. The Crowne-Marlow Social Desirability Scale is one such scale. 3) Structure the testing situation to minimize the pressure for participants to look good.

Matchmaking and personality assessment

Some scales: 1) Marital Rating Scale 2) Dyadic Adjustment Scale Critics: 1) Predictive validity questioned 2) Homogeneity of comparison group noted. Some sites: match.com, scientificmatch.com, perfectmatch.com, chemistry.com, eharmony.com The principle of using psychological testing to match people is popular today. eHarmony has the only patented online matching system.

Goodenough-Harris Draw-A-Person test (DAP)

Test may be interpreted according to the presence or absence of certain elements (e.g., size of the eyes may indicate paranoia; absence of a face may indicate depression).

Legal issues associated with businesses and integrity testing.

Tests that explicitly screen for characteristics that are important for success on the job are legal. Assessment tests are legal unless they have a disparate impact on disadvantaged groups. Federal law prohibits discrimination on basis of race, color, religion, sex, or national origin. By law, integrity testing can be used during the hiring process to predict counterproductive behavior because validity is well established.

Personality assessment

The measurement of individual characteristics of a person, (often through personality tests, interviews, and other measures). The most commonly used are personality tests.

Thematic Apperception Test (TAT)

Used with construction techniques to measure motives including the need for affiliation, the need for power, and most often, the need for achievement. Respondents write a story in response to a picture. The stories are coded for certain elements indicative of achievement motivation.

Face validity

When an assessment appears to measure the theoretical concept it was designed to measure based on the kinds of questions it contains. It measures the construct of interest. High face validity - obvious what test is assessing. Low face validity - harder to see exactly what concept the test is measuring. Not the most convincing type of validity. Useful under two conditions. First, it's important for personnel testing where cooperation and motivation of the test-taker can affect the results of a test. Secondly, when researchers are developing a new measure of a concept.

Internal consistency reliability

When an assessment gives consistent results across items, demonstrated by parallel forms reliability, split-half reliability, or Cronbach's alpha reliability. See if different items of the test give similar results. 1) Parallel forms 2) Split-half 3) Cronbach's alpha

Temporal consistency reliability

When an assessment gives consistent results across time, often demonstrated by test-retest reliability. Have respondents take the test a second time to see if their scores are similar. (test-retest)

Construct validity

When an assessment successfully measures the theoretical concept (called a construct) it was designed to measure. Must demonstrate both what a test measures and what it doesn't measure.

Barnum Effect

When people falsely believe that invalid personality tests are actually good measures of personality because they contain feedback so general that it applies to many people at the same time. Invalid tests because it lacks construct validity because it distinguishes no one. Like a circus, "a little something for everybody." "There is a sucker born every minute."

Is the NEO-PI-R a good personality test?

Yes. - Good theory operationalization behind test, criterion and convergent validity. - High internal alphas; acceptable Cronbach's alpha - Valid for use in many populations, clinical settings, and some cultures.

Interrater reliability

a measure of rater consistency; when there is agreement among raters. Making sure measures are reliable across multiple raters.

Structure the testing situation

reduces response sets and controls for social desirability by structuring the testing situation to minimize the pressure for participants to look good. - assure anonymity in respondents responses - physically separate respondents from each other - have respondents seal responses in an envelope and deposit in a box when exiting testing room.

Q-sort test

reduces response sets by forcing respondents to limit the number of times each response can be used. Helps identify extreme responding. Forces participants to use the full scale and avoids many types of response sets.

Crowne-Marlow Social Desirability Scale

reduces response sets by giving participants a scale that measures their tendency for socially desirable responding and then uses their scores to statistically control for social desirability in the score of interest.

Likert-type rating scale

self-report personality test; ask participants to rate their agreement, degree, similarity, or frequency using a scale, such as a 5- or 7- point scale.

Visual analog scale

self-report personality test; different size boxes used for a rating scale.

Forced-choice format

self-report personality test; respondents must choose their answer from among a limited number of alternatives, typically two or three. Ex: Locus of Control Scale, Machiavellianism Scale

Dichotomous two-choice scale

self-report personality test; true-false, yes-no choices.

Machiavellianism

the extent to which a person believes that other people are easily manipulable. People high in Machiavellianism, High Machs, are not more hostile, vicious, or vindictive than Low Machs; they just have a cool detachment when dealing with other people, emotional issues, or potentially embarrassing situations.