Psychometric Properties and Norms Flashcards

The WAIS-IV Technical and Interpretive Manual reports an average internal consistency reliability coefficient of .98 for the Full Scale IQ (FSIQ). Which of the following is the most accurate interpretation of this statistic?

Answer: 98% of the variance in FSIQ scores is attributable to true score variance.

A reliability coefficient indicates the proportion of total observed score variance that is due to true differences in the ability being measured, as opposed to random measurement error. A coefficient of .98 means that 98% of the score differences among examinees reflect true differences in their cognitive abilities, and only 2% is due to error.

An examinee obtains a Perceptual Reasoning Index (PRI) score of 112. Using the standard error of measurement (SEM), the examiner determines the 95% confidence interval is 105-118. What is the correct way to interpret this confidence interval?

Answer: The examiner can be 95% confident that the examinee's true PRI score falls within the range of 105 to 118.

A confidence interval provides a range of scores within which an individual's true score is likely to fall, accounting for the inherent error in any test score. A 95% confidence interval means that we can be 95% confident that the person's true, error-free score lies somewhere within that specific range.

Factor analysis of the WAIS subtests consistently supports a hierarchical model with a general intelligence factor ('g') at the top and more specific abilities (the indices) below. This type of statistical evidence primarily supports which psychometric property of the test?

Answer: Construct validity

Construct validity is the extent to which a test measures the theoretical construct it purports to measure. Factor analysis is a key method for establishing construct validity. The finding that WAIS subtests cluster into theoretically meaningful factors provides strong evidence that the test is measuring the constructs of general and specific cognitive abilities as intended by its design.

Which of the following is the most critical feature of the WAIS standardization sample for ensuring an individual's scores are accurately interpreted relative to their peers?

Answer: The sample was stratified to match recent U.S. Census data on key demographic variables.

For norms to be meaningful, the standardization sample must accurately represent the population. The WAIS normative sample is carefully stratified based on U.S. Census data for variables like age, sex, race/ethnicity, geographic region, and education level. This ensures that an individual's performance is compared to a truly representative group of their peers.

A clinician finds a 20-point difference between an examinee's Verbal Comprehension Index (VCI) and Processing Speed Index (PSI). To determine if this discrepancy is not just statistically significant but also clinically meaningful, the clinician should consult the WAIS manual to determine the:

Answer: base rate of this difference in the normative sample.

A statistically significant difference simply means the difference is unlikely due to chance or measurement error. However, its clinical importance is best understood by its frequency in the general population. Base rate tables show how often a specific score difference occurred in the standardization sample. A large difference that is very rare (i.e., has a low base rate) is more likely to be clinically meaningful.

In the standard scoring convention for the WAIS, what are the respective mean (M) and standard deviation (SD) for individual subtest scaled scores and for the Full Scale IQ (FSIQ)?

Answer: M=10, SD=3 for subtests; M=100, SD=15 for FSIQ.

The WAIS uses two different metrics. Performance on individual subtests is converted to scaled scores, which have a mean of 10 and a standard deviation of 3. Composite scores, such as the Index Scores and the FSIQ, are on a different scale with a mean of 100 and a standard deviation of 15.