Theories of Personality (PSY 225-A01)/Chapter 2

Objective vs. subjective assessments. Subjective experiments have apparent flaws, such as conflicting perspectives. However, we can't be 100% objective when it comes to personality. For example, certain behaviours exhibited by sexually abused children may be 'subjective' (interpreted by the exterminator), but if consistent across multiple experiments, may prove to be sound & [evidently] strongly supported. Subjective tests do provide more depth and exploring into one's personality (ink blot test).
 * Objective assessments - Doesn't force the experimenter to interpret the subject's actions (pushing a button when they hear a beep).
 * Subjective assessments - Force the experimenter to measure the subject's actions to his own interpretation (Pirate Test, ink blot test).

2.1 Measuring Personality
How do we measure a person's charisma & predict his likelihood of becoming a leader?


 * Affective Communication Test (ACT) - Look at multiple statements and on a scale of -4 to 4, you rate your relativeness (-4 - no relation; 4 - relation to the xtreme)

Reliability

 * Error variance - Variances that come about through chance.

Internal Consistency Reliability
Reliability is measured in that equivalent parts of a test have high correlation. For example, if the ACT was split into 2 parts - one that scores high on the 1st section should score high on the 2nd section. This is measured by the internal consistency reliability, meaning repeated measures need to yield consistency in order to prove reliability. Split reliability = split assessment in half, give the split assessment to subjects, compare the results. The "items" on the test should relate to each other (charisma test).

It is important not to add so much to the point that it becomes unnecessarily repetitive. The general coefficient stays at .80.

Interrater Reliability
Measures the overall degree of agreeableness of independent examiners. Look at the correlation among these ratings. If consistency? Good interrater reliability! No consistency? Is the measure reliable then?

Test–Retest Reliability
This way tests the reliability of the test by administrating the test on two different occasions. We expect a stable test to result in similar outcomes throughout time (ex, personality test).

How can we have a reliable (stable) assessment of personality if personality may change?

 * Personality is made up of dynamic patterns (ex, situational) - Allport
 * Only stable through a short period of time, not over a lifetime

2.1.2: Construct Validity
Are these tests measuring what they are supposed to be measuring?


 * Construct validity - Extent on which a test is truly measuring a 'theoretical construct'. This is proven by seeing if the behavior measured in a specific test will be replicated in different scenarios. For example, those that score high on the ACT test should be extroverted in other situations. Face validity is not statistical: does the test measure what it is suppose to measure at face value? Convergent validation shows that the assessment is related theoretically to what it needs to be attached to [charisma --> extroversion], while discriminant validation is the opposite (If the ACT is related to extroversion scales (as it should be), then that is evidence for convergent validity. If the ACT were related to intelligence (as it should not be), then that would be evidence that it is lacking in discriminant validity).
 * Criterion-related validation - A measure predicts a criteria (ACT --> extrovertedness --> being a leader | SAT --> intelligence --> college success).
 * Multitrait-multimethod perspective - Proper test validation assesses multiple traits through different assessments/tests.

''In gathering items for the ACT, we collected a wide range of characteristics that seemed relevant to having a highly expressive style. Unreliable items and ambiguous items were discarded, but the final scale includes items about nonverbal expressions (such as touching and laughing and facial expression), about acting, about social relations (such as behavior at parties), and about interpersonal expressive communication (such as giving a seductive glance). The ACT tries to capture many of the content dimensions of personal charisma.''

Ethnic Bias
The test in question fails to account for cultural differences or values (ex, comparing the test results from one culture to another without accounting for the differences). Better to study within a culture than between cultures.

Gender Bias
Adjusting test results to match gender biases (forcefully equating men and women's extroversion scores to keep them the same).

Response Sets
Response sets are 'responses unrelated to the personality characteristic being tested'. This includes acquiescence bias (respondents have a disproportionate favor to agree with a statement/question), social desirability (have you ever considered sexually assaulting a child?), random answers ("let me get this over with") and lying.


 * Reverse scoring: To address acquiescence response set (tendency to say yes), as people may agree with a certain wording.
 * Social desirability response set: To answer in a way that paints them as favorable.

Self-report tests

 * Self-report tests
 * MBTI (Minnesota Multiphasic Personality Inventory (MMPI) - used for diagnoses or personnel work
 * Millon Clinical Multiaxial Inventory (MCMI) - used by therapists to assess personality disorders [clinical syndromes/problematic personality patterns]
 * NEO Personality Inventory (NEO-PI) - relies heavily on factor analysis [several correlations --> condensed dimensions]. The Big Five personality traits are Conscientiousness (organized vs. careless), Extroversion (outgoing vs. reserved), Agreeableness (friendly vs. harsh), Neuroticism (nervous vs. confident), and Openness (curious vs. cautious).
 * Myers-Briggs Type Indicator
 * Q-sort tests: Self-concept (stating statements in card-from from least relatable to most relatable)
 * Rating and judgement from others: Ratings from staff members evaluated by psychologists (Genetic Studies of Genius - Lewis Terman's 1920 study)

Biological measures

 * Biological measures: reaction times (phrenology, now we know that the brain's nervous system is the factor behind personality)
 * Electroencephalography [easy to set up, can show brain waves through EEGs after a stimuli has been presented to determine the neurons' response)
 * Positron emission tomography - Records the brain's use of radioactive glucose. Nerve cells use glucose when completing a task, therefore we are able to see which sides of the brain are being used when doing a specific task.
 * Functional magnetic resonance imaging (fMRI) measure changes in oxygenated blood flow, as more blood flow = activity is highest (in the brain). Though, this doesn't prove anything directly. Hormone levels have became more prominent in recent studies (depression = lack of serotonin).
 * Diseases (brain poisoning --> certain cases of schizophrenia). See AP Psychology/Biological Bases of Behavior for more.
 * Behavioral observations: Experience samplings (cellphone texts, alarm reminders where individuals document their behavior/thoughts --> look into frequencies of certain behaviours or correlations).
 * Interviews: Type A structured review (psychotherapetic interviews --> therapist interviewing a client. probing interview [interrogation], structed interview is objective, less subjective [medical interview])
 * Expressive behavior: Speech rate; gestures (how many times do they smile?)
 * Document analysis: Dream diaries/biographies (one can develop a psychobiographical study from studying the bio of a famous psychologist)
 * Projective tests: Draw-A-Person, ink blot test, Thematic Apperception Test (picture --> make up a story --> ending of story --> scored in a certain way [unconcious feelings onto story]), Implicit Association Test (ambiguity --> individual's interpretation) | limited tests
 * Demographics and lifestyle: Age; culture group; race (divorce is common? let's see cultural/religious influences)
 * Online Internet analysis: Counting likes on FB

Research

 * Case studies - doesn't tell us about casual relationships and cannot apply to others, but provides crazy, in-depth info on a specific person.
 * Correlational studies - they don't tell us the association though (what causes it?) + correlation does NOT equal causation.


 * Experimental [correlational] vs descriptive - investigator manipulates an IV (neuroticism and divorce, manipulate neuroticism) | correlation does NOT equal causation, especially in descriptive studies. Problems with this method of research include: "First, we do not know if the finding can be generalized to others who have different characteristics from those studied. Second, and relatedly, we do not know which moderating factors may be important; for example, our intervention may work very well for older people but poorly for younger people. Third, there could have been problems (biases) in the ways our elegant design was carried out; that is, there could have been experimenter errors. Finally, we do not really know how other interventions would work, nor which is best." + So, we rely on quasi-experimental research or naturally occurring experiments to see what influences and components increase the likelihood that a charismatic leader will emerge. For example, we might compare teenagers from coed high schools to those from single-sex high schools.
 * Peer reviewed studies - Researcher has conducted his study, wrote up his results in a specificed format for a journal, send their write-up to the journal editor, the journal editor will remove the cover page and send the information out to others in those fields, those peers review the literature and confirm the comprehension and accuracy of it, peers will review it and send the information back to the editor - the editor will either reject or accept the article.