Reliability & Validity

Question 1

Self - report measured on two occasions, interval long enough to minimize respondents remembering how they filled it out at Time 1. Correlation calculated between scores.

Answer 1

Test - Retest Reliability

Question 2

Rating scales. How well do two different clinicians rate the same patient on the test. Agreement between raters measured with ____ correlations.

Answer 2

Inter - Rater Reliability & Intra - Class

Question 3

Internal consistency, average of all possible split halves.

Answer 3

Cronbach’s Alpha

Question 4

Multiple versions of a test matched for content and difficulty administered to the same sample and their scores calculated.

Answer 4

Parallel Tests Reliability

Question 5

The ability of tests to predict future criterion. Cross-sectional studies.

Answer 5

Predictive Validity

Question 6

Factor Analysis coupled with evidence to show a test measures are robust theoretical construct.

Answer 6

Construct Validity

Question 7

Test systematically examine representative range of all the salient behaviours that constitute the trait being measured.

Answer 7

Content Validity

Question 8

Convergent & discriminant validity demonstrating a close statistical relationship with another variable. Correlations with other similar tests longitudinal studies.

Answer 8

Concurrent Validity

Question 9

Evidence for the validity of a test by demonstrating that the test agrees closely correlates well with a separate objective criterion of the construct of interest.

Answer 9

Criterion Validity

Question 10

Whether a test looks and feels like it is measuring what it claims to. How relevant and realistic is it in the eyes of the test taker.

Answer 10

Face Validity

Question 11

A measure of reliability used for interpreting individual test scores.

Answer 11

Standard Error of Measurement

Question 12

Observed score = True score + Error, standard deviation of normal distribution of test scores obtained by one person on equivalent tests.

Answer 12

Standardized Error of Measurement

Question 13

Test calculating Pearson ‘r’ for internal consistency. Half of the test is summed & correlated with the other half. Split randomly into odd & even items.

Answer 13

Split - half Reliability

Question 14

Measure linked to the population to which one wants to apply the measure. Related to scores not the test itself.

Answer 14

Reliability