Chapter 5 + 6

Question 1

What are 4 kinds of reliability estimates ?

Answer 1

1. Test-retest
2. Parallel forms
3. Split-half
4. Inter-scorer

Question 2

Describe test-retest reliability estimate

Answer 2

This is a measure of the correlation between the same individuals score from one test to a second administration of the same test

Question 3

Describe parallel forms and split-half reliability estimates

Answer 3

parallel forms: Here in each form of the test the means and variances are equal

split-half: the process of equally dividing the scores of the same test and calculating Pearson r (correlation) between the scores, then spearman-brown is used to adjust the half-test reliability

Question 4

Describe a source of error for each reliability estimate

Answer 4

1. Test-retest
   - includes practice effects, test taker factors like fatigue,
2. Split-half
   - test construction, meaning inconsistencies of the items
3. Parallel forms/alternative forms
   - test construction and administration effects
4. interscorer reliability
   - Test scoring and interpretation

Question 5

What is a good reliability coefficient ?

Answer 5

Between 0.9 and 0.8

Question 6

What is the standard error or measurement ?

Answer 6

Estimates the spread of scores, measured with the same instrument, around the true score on a normal distribution. The higher the reliability the lower the SEM

Question 7

What are some types of validity and describe them

Answer 7

1. Content
   - do the items appropriately measure what the test is designed to measure
2. Construct
   - how scores on the test relate to other test scores
3. Criteria
   - Does the criteria of one test match others measuring the same thing
4. Face
   - does a test appear to measure what is claiming to measure on a surface level
5. Ecological
   - can the results be generalized into real word scenarios

Question 8

Describe content validity, and how it is ensured

Answer 8

Content validity
- Do the items of a test measure appropriately the entire behavior in questions for the test as a whole
Ex
- If I am designing a test to measure math ability, I can ensure content validity by consulting an math expert to help clearly define each mathematical item I wish to include on my test and there relevance to mathematical ability

Question 9

Describe criteria validity and how it is ensured

Answer 9

Criterion validity
- Describes how well one the scores from one measure are capable of predicting scores for another measure
Ex
- In ensuring this validity, I could take a pre-existing test that is deemed a valid criterion measure and administer both tests to my sample to compare the results

Question 10

Describe construct validity and how it is ensures

Answer 10

Construct
- Does the test measurement tool really represent the thing we are aiming to measure
Ex
- If we are measuring depression, we must do research to ensure that our items are accurately involved in the overall concept of depression