Individual Differences/Personality - Research

Lecture 6

Personality

mental concept that influences behaviour via mind-body interaction

Maltby et al. - Evaluating Personality Theories

a theory should:-

• help understanding of behaviour

• have empirical validity

• have testable concepts

• be comprehensive

• have heuristic value

• have applied value

• include cultural contexts

Intelligence

suggested intelligence involves ability to:-

• learn - acquire, retain + use knowledge

• recognise problems

• solve problems - take what learned, find useful solution

Emotional Intelligence

involves self awareness, empathy, motivation, self-regulation and social skills. 

Approaches to personality/intelligence research

idiographic → based in individuals. emphasises uniqueness. techniques e.g. case studies used.

nomothetic → aims to identify individual differences that occur consistently across groups. statistical techniques used.

Developing Items for Psychometric Tests

focus groups - assess question clarity (include reverse wording). assess for acquiescence response bias + social desirability bias.

response format - forced choice, Likert scale, open format, etc.

Psychometric tests - Factor Analysis

statistical technique for data reduction - simplify relationship between numerous variables e.g. Big Five Model (Costa & McCrae, 1992).

Test Reliability - Internal Reliability

free from error? evidence for usefulness?

• internal reliability → all items measure same thing, high correlations among items. Cronbach’s alpha for reliability.

Test Reliability - Test-Retest Reliability

conservative method of estimation. should have same score on 1st and 2nd tests.

Test Reliability - Inter-Rater Reliability

do raters/markers/examiners/peers, etc agree on scoring

Test Reliability - Split-Half Reliability

scores on second half compared with first half (is there a correlation?)

Test Validity - Convergent Validity

extent shows association w/ measures it should relate to e.g. self-esteem+depression

Test Validity - Concurrent Validity

correlations with known standard measures of the construct

Test Validity - Discriminant Validity

not related to things it shouldn’t be (independent measure)

Test Validity - Face Validity

how well does test measure what it was designed to measure - each item should contribute independently to overall measure

Test Validity - Content Validity

does test measure all aspects of construct e.g. in personality → behaviour, thoughts, feelings, etc.

Test Validity - Predictive Validity

does the measure accurately predict what it should predict?