Psychology Of Human Differences

Front 1

Traits

Back 1

-generalized tendency to behave
-can be found in personality interests, cognitive abilities, creativity, attitudes, values, and emotions
-construct
-stable
-have individual differences
-constructed from a series of similar behaviors
-typically divided into 3 different categories (construct, content, and criterion validity - all related to each other)

Front 2

Psychological Construct

Back 2

putting all parts of creativity together creates a psychological construct

Front 3

How do we measure traits?

Back 3

-follow a person and observe their behavior
-could ask people how they behave and what their reactions are

Front 4

problems with measuring traits

Back 4

time consuming, not ethical, not appropriate for large numbers of people, people may lie, etc.

Front 5

What variable are we measuring with self report measures or ability tests?

Back 5

A response variable - the measure is a stimulus and we are directly assessing the response to that stimulus

Front 6

Given the problems with self-report measures, why do we bother with them?

Back 6

they are usually found to be related to some behavior of interest, for example...self esteem predicts school performance

Front 7

External (Empirical) Approach

Back 7

1. General pool of items
2. Items are selected on the basis of a correlation

Front 8

Inductive (Internal) Approach

Back 8

1. General pool of items
2. Items are selected on the basis of psychometric characteristics like reliability (take question out and see if correlation goes up or down)

Front 9

Deductive (Rational) Approach

Back 9

1. Start with a logical definition of the trait you want to measure
2. Items are written on a rational basis by the psychologist

Front 10

We create a test that we think measures creativity. How can we be sure.

Back 10

-we have to establish the v alidity of the test
-ask the question, "Does the measure really measure what we think it measures?"

Front 11

Construct Validity

Back 11

-process of determining whether or not our measure measure what we think it measures
-this process never ends
-experimental process
-approach suggested by cronbach and meehl (1995)

Front 12

Nomological Network

Back 12

your test...has a positive correlation with...hours of sleep from the night before (example)

Front 13

Criterion Validity

Back 13

-a specialized form of construct validity
-Is a measure correlated with what it should be and not correlated with what it shouldn't be?

Front 14

Convergent Validity

Back 14

-type of criterion validity
-measure should be correlated with what we expect
-can be a positive correlation

Front 15

Discriminant Validity

Back 15

-test does not correlate with measure we think are unrelated
-expecting to find a zero correlation

Front 16

Content Validity Evidence

Back 16

-specialized form of construct validity
-is the measure covering the content domain we want it to investigate?
-do we have enough questions covering every part of what we are measuring?
-typically involves subjective opinions by expert judges

Front 17

Face Validity

Back 17

-similar to content validity
-no data collect, so not really "validity"
-a subjective decision made by those taking a test

Front 18

Benefit of Face Validity

Back 18

establishes rapport with the person taking the test

Front 19

Problem with Face Validity

Back 19

-questions are easier to fake
-just because a test has high face validity does not meam it has high construct validity

Front 20

Experimental Evidence

Back 20

-2 groups of participants
-Example: Group 1 is sleep restricted, group 2 is limited to 4 hours of sleep

Front 21

Reliability Evidence

Back 21

-If i measure something several times, will I get the same answer each time? ( measuring height in a span of minutes)

Front 22

What Does a Reliability Coefficient Mean?

Back 22

-smallest to biggest
-reliability formula- x=Tte
x=observed score
T=true score
t=error - stray away from what you actually know (guessing)

Front 23

Observed Score

Back 23

made up of true score plus error

Front 24

Reliability Coefficient

Back 24

-the higher this is, the less error there is
-less than .70 is bad
-.70 squared is .49 which is only 49% of variability around the mean is due to true score
-reliability of .70 to .90 is ok
-reliability of .90 to 1.0 is excellent

Front 25

Inter-Rater Evidence

Back 25

-similarity of the ratings of 2 or more judges
-Example: Q. If we have low inter-rater reliability, what does this say about the construct validity?
-A. Judges are interpreting what they see differently
-problem for construct validity

Front 26

Test-Retest Evidence

Back 26

-measure is taken once and then again at a later time
-correlation between 2 times indicates reliability

Front 27

Problem with test-retest evidence

Back 27

-knowledge about first assessment may bias the subjects
-due to test knowledge

Front 28

Equivalent Forms Evidence

Back 28

-2 similar measures of the same construct
-the higher the correlation between the 2 tests, the more reliable both of them are

Front 29

Benefit of equivalent forms evidence

Back 29

you do not have interference by people

Front 30

Internal Consistency Evidence

Back 30

the way people answer questions in one part of a test should be the same as they do in another part

Front 31

Split-Half Reliability

Back 31

-break the test into 2 parts (odd and even items)
-coefficient alpha: average

Front 32

If your question has low internal consistency, what does this say about the construct validity of the measure?

Back 32

there is a problem

Front 33

Data Reduction Techniques

Back 33

-designed to help you make sense of a large number of variables
-large number of variables are reduced to a smaller number of categories

Front 34

Factor Analysis

Back 34

-creating categories on how things are related to each other
-categories are based on patterns of correlations
-correlation between 5 tests is factor analysis, two basic clusters of correlation (1,2,3 and 3,4, and 5)`

Front 35

Multidimensional Scaling

Back 35

-correlations are treated as distances
-more positive correlations meaning closer distances
-more negative correlations meaning farther away

Front 36

The Evidence for a Genetic Influence on Behavior - 3 Sources of Info?

Back 36

1. Extent to which Monozygotic twins resemble each other more closely than dizygotic twins

2. Extent to which adopted children resemble their adoptive relatives and their biological relatives

3. Extent to which monozygotic twins separated at birth develop greater similarities than they would by chance

Front 37

Twin Study: Finding and Problem

Back 37

Finding: Monozygotic twins are more similar in their behaviors than dizygotic twins
Problem: Monozygotic twins are often treated more similarly than dizygotic twins

Front 38

Adoptive Study: Finding and Problem

Back 38

Finding: Having a biological parent with this disorder increases the chances the adopted child will develop the disorder
Problem: Often difficult to track the biological parent

Front 39

Twins Separated at Birth: Finding and Problem

Back 39

Finding: Twins separated at birth have far more behavioral similarities than what would be expected by pairing 2 people by chance
Problem: The major problem with this method is that there are not many twins separated at birth

Front 40

General Problems in Human Differences

Back 40

1. You must use large samples
2.You often must use self-report measures
3. There are few "scientific" names for constructs or universally accepted means for measuring a construct