Psychological Testing (Psych 362)

Psychological Testing (Psych 362)

Spring, 2012

Review Sheet for Test 3

(Ability and Achievement Testing, Test Bias)

Readings

·  Gregory (Reader): 40-46, 50-53, 151-156, 163-166, 172-176, 178-194, 206-212, 214-216, 221-225, 232-239, 245-255, 439-447 (neuropsych), 475-476, 542-549 (review)

·  F&B Ch 11 (Test Bias)

·  UC Readings, WFU readings, Sackett article

In class, we focused on in-depth discussions and examples with regard to a few key tests. The textbook supplied additional important information across a slightly broader range of tests.

NOTE: DUE TO THE ABBREVIATED TIME WE WERE ABLE TO DEDICATE TO THESE TOPICS, THE “NEUROPSYCHOLOGICAL TESTING” MATERIAL WILL NOT BE INCLUDED ON THE TEST.

Intelligence testing

Conceptual background

·  Origins of ability testing

o  Testing in ancient China (purposes of testing>

o  Physiognomy and phrenology – what are they? Validity?

o  Experimental psychology and the “brass instruments” era

·  Fechner, Wundt, Galton, J. M. Cattell, Binet – who are they, in terms of testing?

·  What does the term “brass instruments era” refer to?

o  Binet - Mental level, Intelligence quotient, chronological age, mental age. Be able define each and their links.

·  Expert and Layperson definitions of intelligence, in general

o  Core themes and differences

·  Specific theories of intelligence

o  Importance of factor analysis

o  Galton (sensory keenness, reaction time –movement time – be familiar with these and with their links to intelligence, or lack thereof).

o  Spearman’s theory of intelligence (with attention to class discussion). Implications for correlations among specific tests of ability? Why would some tests be highly correlated, why would other be weakly correlated?

o  Thurstone and the use of factor analysis – differences/convergence with Spearman’s view

o  Gardner - Multiple Intelligences (with attention to class discussion). Conceptual basis, empirical evidence?

o  Sternberg – Triarchic theory (be familiar with broad components of it). Why is Sternberg so critical of most intelligence tests?

Specific “Traditional” Intelligence tests

·  Wechsler tests

o  Assumption of IQ constancy (what is it?)

o  Know the structure of the WAIS – subtests, composite/index scores, full-scale – how are these levels related to each other?

o  Be familiar with the types of subtests and indexes on the WAIS-IV in particular. Eg, be able to see/read about a task and identify which subtest it would reflect. Know the four index scores – what they are, and which subtests go with which indexes.

o  Format of test – types of subtests, individual or group, etc.

o  Standardization procedures

o  Psychometric quality –

·  Dimensionality – evidence in favor or against?

o  Reliability – general levels of reliability (internal consistency and test-retest/stability)? Is test reliable for cognitively-impaired subgroups? Is reliability better or worse for subtests (as compared to index scores or full-scale score?) and what are the implications of this?

o  Validity – general perspective on quality? Content, criterion, convergent, internal structure (ie, dimensionality)?

·  CFIT and RPM

o  Format and logic of test, Connection to Spearman, reliability, validity, is it truly “culture-fair”?

·  Wonderlic

·  Design/structure

·  Where is it applied most typically?

·  Differences from WAISIV (Structure, length, format, etc)?

·  Reliability and Validity evidence? Eg, convergence with WAIS?

Emotional Intelligence (Discussed in class)

·  testing through self-report or through “test” – correlation between these methods? Which method is more strongly related to personality variables? To actual interpersonal behavior?

Achievement testing

Conceptual background

·  Differences among ability, achievement, and aptitude (see p. 215)

Specific tests

·  SAT & ACT (with greater emphasis on the SAT as discussed in detail in class)

o  Format, Sections & subtests,

o  Psychometric quality (know general levels)

·  Reliability – internal consistency and stability. Implications for “why should one morning’s performance have such a strong effect?” criticism?

·  Validity (predictive and incremental),

o  Be familiar with the UC Report, using the questions I handed out in class to guide your reading.

o  Be familiar with discussion of “case study” of achievement testing as part of the admissions process at Wake Forest, as discussed in class – criticisms of SAT, empirical evidence, preference for interviewing (and related problem).

o  Be able to interpret statistical results in terms of making conclusions about predictive validity, incremental validity, and reliability. (eg, results akin to those presented in class and/or in the UC Report)

·  GRE, MCAT, & LSAT,

o  General info about purposes, reliability, validity

o  Problems with trying to evaluate the predictive validity of the GRE

·  Effects of coaching on test performance

o  Evidence for effect on SAT and GRE

Sackett et al. (2008) – addressing common criticisms of achievement tests.

·  What are the basic criticisms/assertions?

·  Be in-depth familiar with at least four, and be able to explain the nature of the criticism, the types of empirical data that are available, and the take-home messages from those data

·  This document also applies to the “Test Bias” section (see below)

Learning Disabilities

·  Definition and operationalization, in terms of discrepancy

o  How is it assessed, under the discrepancy model?

o  What is a “severe discrepancy”?

o  Problem with the discrepancy model?

·  NJCLD definition

o  Important difference with discrepancy model?

o  Intraindividual weakness – what does this mean? How is is related to LD?

·  The “RTI” approach

·  Features and two current categories of LD?

·  Percentage of diagnosis

·  Causes of LD?

·  Why use individually-administered achievement tests?

Test bias

Common “bias-oriented” criticisms of ability, achievement, and aptitude tests

Test bias vs test fairness – definitions and differences

Technical definition of test bias

Sources/forms of test bias

·  Content bias –

o  Definition

o  Evidence about degree of content bias in well-known standardized tests

·  Predictive/Criterion bias

o  Definition,

o  Regression – homogeneous regression, intercept bias, slope bias. Be able to identify and interpret the implications of results

o  Evidence about degree of criterion/predictive bias in the SAT

·  Construct bias

o  Definition

o  What kind of evidence is relevant?

o  Evidence about degree of construct bias in well-known standardized tests. Be able to identify and interpret the implications of results, as shown in class

·  Take-home conclusion on test bias

Social Values and test fairness

·  Difference between bias and fairness

·  Ethical stances on test use and test fairness – unqualified individualism, quotas, qualified individualism

Case study of Test bias – SAT, based upon the UC Report.

·  What were the bias-related arguments raised in WFU’s criticisms of the SAT? Along what demongarphic dimensions was the SAT accused of being biased?

·  Which of these arguments are addressed in the textbook, and what information is provided by the text?

·  Which of these arguments are addressed by empirical information in the UC Report, and in what way does the UC Report data support or fail to support those arguments? Be able to read and interpret the relevant information, both at a technical level (ie in terms of the logic of multiple regression) and explain at a more intuitive, psychological level (ie, in a way that your parents or peers would understand).

Legal Issues (in employment testing) - this should be a review from the first section of the class

·  Test soundness AND test relevance,

·  Uniform Guidelines, disparate impact,

·  Civil Rights Act of 1991

·  Soraka v Dayton Hudson

See the Sackett et al. (2009) article mentioned above