Testing & Measurement

Reynolds defined items with content bias as items

• that are more difficult for one subgroup than another, when ability is equal.

• that are based on cultural knowledge that one group may lack.

• for which the definition of the correct answer is based on the culture of the examiner.

• for which the wording may be unfamiliar for a subgroup.

One definition of an unbiased test is that in a regression equation predicting a criterion from a test score, all population subgroups should have the same

• slope.

• intercept.

• mean score.

• variance.

The Flynn Effect refers to

• increases in IQ scores over the generations.

• decreases in fluid intelligence scores as individuals age.

• increases in crystallized intelligence over the generations.

• decreases in computational skills as individuals age.

Which of the following best states our current state of knowledge regarding racial group differences in IQ?

• Racial group differences in IQ exist but the reasons for those differences are not clear.

• There are no racial group differences in IQ.

• Racial group differences in IQ are clearly due to genetic differences.

• Racial group differences in IQ can be explained by years of education completed.

One key difference between achievement and aptitude tests is that achievement tests

• assess current attainment, while aptitude tests estimate future performance.

• estimate current intellectual level, while aptitude tests assess current mastery of skills.

• focus on a broad array of intellectual abilities, while aptitude tests focus on a narrower, more focused range of abilities.

• focus on a student’s potential, while aptitude tests focus on a student’s current proficiency.

Which of the following represents a true distinction between content bias and criterion validity bias?

• Content bias is evaluated by comparing item difficulty to general ability, while criterion validity bias is evaluated by error in prediction.

• Content bias is evaluated at the test level, while predictive bias is evaluated at the item level.

• Content bias is evaluated by panels of expert judges, while predictive bias is evaluated by experimental evidence.

• Content bias is evaluated by regression analysis, while predictive bias is evaluated by panels of expert judges.

A psychologist wants to make a case that observed differences in IQ among racial groups are not due to inherited intelligence differences. If she were to draw on the evidence provided by the research of Brooks-Gunn, Klebanov, and Duncan, which argument would she make?

• After controlling for poverty and its cofactors, the observed racial differences disappear.

• Individual heritability of IQ does not mean that racial group IQ differences are based in genetics.

• Race is a social construct without biological meaning.

• Observed differences in IQ among racial groups have declined since 1970.