Class 9 Dependent Variables, Instructions/Literature Review

Class 9Dependent Variables, Instructions/Literature Review

Chapters 13, 22 Corkin, & P&B 7 – SP 13

Dependent Variables: ContinuousPsychometric Properties

◦ Validity

◦ Reliability

2

Validity

Convergent and Discriminate Validity◦Correlation

Predictive Validity◦Correlation- Differences across groups◦Longitudinal

Construct Validity◦ Factor structure-latent constructs

3

ReliabilityThe extent to which scores show true variance in attributes across participants as opposed to error variance ◦ Assumption: Error variance is random,

therefore it does not correlate with anything

◦ More items of good quality = higher reliability

◦ One item scales have very low reliability; around r =.25

4

Types of Measures

Observational Measures

Self-Report Paper-and-Pencil Measures◦Content tests- right/wrong answers

◦Likert-scales

5

Sources of Measurement Error Specific error

◦ something unique about the instrument/observational task that differs from what the researcher intended (e.g. social desirability; reading level; idioms)

Transient error◦ some temporary factor that affects

the measurement (e.g. order of instruments; historical events; noise while observations occur; tiredness)

6

Sources of Measurement Error

Scoring/Recording error◦ Tend to be random (self-report/raters)

Inter-rater disagreement/observer drift ◦ in observation studies, what initially

may appear as antisocial behavior may later be judged more normative

7

Types of Reliability Inter-Scorer Test / Re-test

Alternate Forms

Internal Consistency◦how items correlate with each other

8

Internal Consistency ReliabilitySplit-Half

Kuder-Richardson◦Dichotomous items- right /wrong

Chronbach Alpha ◦Average correlation of all possible

split-half reliabilities

9

Parenting Skills Measure: Q 2a-c

a. 12-item measure rx= .70 to rx= .95 among White samples

b. Expected reliability estimates among the Latina sample

c. Three-item measure: rx=.40 to rx=.45

10

Parenting Skills Measure: Q 2a

a. 12-item measure: internal consistency reliability coefficients rx= .70 to rx= .95 among White participants

70% to 95% of variability in respondents’ scores is due to

and the rest is due to .

11

Reliability Estimates: Q 2aExtent to which variability is

due to true variation versus error ◦ Cronbach alpha =. 70;

70% of variation in scores is due to true differences in parenting skills & 30% is due to error

◦ Cronbach alpha =. 95; 95% of variation in scores is due to true

differences in parenting skills & 5% is due to error

12

Parenting Skills Measure: Q 2b

a. 12-item measure rx= .70 to rx= .95 among White samples

b.Expected reliability estimates among the Latina sample

13

Parenting Skills Measure: Q 2b

a. 12-item measure rx= .70 to rx= .95 among White samples

b. Expected reliability coefficient among Latina sample =>.70

Yes/No ?

Reliability refers to scores with specific and

. It’s a property of not of instruments.

14

Parenting Skills Measure: Q 2b

a. 12-item measure rx= .70 to rx= .95 among White samples

b.Expected reliability coefficient among Latina sample =>.70

No. Reliability refers to scores with specific respondents and conditions

It’s a property of scores, not of instruments 15

Parenting Skills Measure: Q 2c

a. 12-item measure rx= .70 to rx= .95 among White samples

b. Reliability estimates among the Latina sample

c. Three-item measure: rx=.40 to rx=.45. How to improve the scale scores’ reliability.

16

Parenting Skills Measure: Q 2c

c. Three-item measure: rx=.40 to rx=.45 –

New items increase reliability only if they are of good quality: validity and reliability

17

Reliability and Correlation: Q 5In correlational research, how does the reliability of two scores affect the probability that the observed correlation coefficient between the two scores approximates the “true” correlation coefficient of the constructs?

18

Internal Reliability and Correlation

Parenting Skills 1 PSM1 rx = .45

Parenting Skills 2 PSM2 rx = .90

Self-Esteem SE rx = .85

PSM1 (rx = .45) correl. Self Esteem PSM2 (rx = .90) correl. Self Esteem

19

Reliability vs. ValidityObserved correlation coefficient

will be smaller and less accurate with the less reliable measure

Correlations between constructs are attenuated by the reliability of the measures

The reliability of a measure puts a ceiling on its validity

20

Instruments

1. Description of Measure

2. Validity Estimates

a. Instrument name a. Convergent/Discriminate Validity

b. Acronym b. Validation Sample(s)

c. Authors

d. Key References 3. Reliability Estimatese. Brief description of

construct(s)a. Chronbach’s alpha

coefficient f. Type of measure (e.g self-

report) Previous and Current studies

g. Number of items b. Test/Re-Test

h. Example of items

i. Items response options

j. Factors or subscales

k. Scoring options and direction

21