PSYC 421 Week 1 DB Forum 1 Example

Discussion Board Forum 1: Thread and Two Replies

PSYC 421 Psychological Measurement, Discussion Board Forum 1 (thread and two replies)

Thread by [Student Name]

The prompt and the classmates' posts are composites standing in for the course's own forum; the textbook and the journal article are real and described from their published sources; the requirements follow Liberty's public course guide for PSYC 421.

Thread

Over a holiday dinner, my uncle heard that his daughter had completed a depression questionnaire at her doctor's office. "That's just guesswork," he said. "A blood test tells you something. A questionnaire tells you what you want to hear." Many people share his view that psychological tests are soft compared with medical tests. This module's reading and a major review suggest the comparison is less lopsided than he thinks.

Cohen and Swerdlik (2018) distinguish psychological testing, the process of measuring psychology-related variables with devices or procedures designed to obtain a sample of behavior, from psychological assessment, the gathering and integration of data to make an evaluation, often in answer to a referral question. A questionnaire score is testing. What a clinician does with that score, alongside an interview, history and observation, is assessment. My cousin's questionnaire was meant to start a conversation, not to deliver a diagnosis.

Meyer et al. (2001) addressed my uncle's doubt directly. Drawing on more than 125 meta-analyses of test validity and 800 samples examining multimethod assessment, they reached four conclusions. Psychological test validity is strong and compelling, it is comparable to the validity of many medical tests, different assessment methods provide unique information, and clinicians who rely only on interviews tend to form incomplete understandings. Their comparisons included medical procedures that patients rarely question, and many of those showed validity coefficients in the same range as well-established psychological measures.

Their third and fourth conclusions may matter most in practice. A self-report questionnaire, a structured interview and a family member's report each capture part of a person's experience, and they often disagree. Rather than treating disagreement as error, the authors argued that a skilled clinician can use a battery of methods to build a fuller picture than any single method provides. That is the difference between testing and assessment in action.

None of this means tests are infallible. Validity depends on using a test for the purpose and population for which it was developed, and scores always include error. But "just guesswork" is not a fair description of instruments that have been studied as carefully as many medical tests.

Classmates: what would it take to persuade a skeptical family member that a test was worth taking seriously?

Reply to Renee

Renee, you raised an important limit on my argument: a test may be valid for one group and not for another. Your example of an English-language achievement test given to a student who had just arrived from Guatemala makes the point clearly, since the score would measure English proficiency as much as achievement. Cohen and Swerdlik (2018) discuss this concern in their treatment of culture and assessment and of legal and ethical issues, noting that test users are responsible for considering whether a test is appropriate for the person being assessed, including language and cultural background. Meyer et al. (2001) summarized validity evidence across many studies, but those averages do not guarantee that a test works equally well for every population. That is another reason their call for multiple methods makes sense. An interview in the student's first language, observation in class and records from her previous school could correct for what the test cannot capture. Do you think schools have the resources to assess students that thoroughly?

Reply to Kevin

Kevin, you asked whether a validity coefficient of .30, which Meyer et al. (2001) report is common for both psychological and medical tests, is really impressive, since it means the test explains only about 9% of the variance in the outcome. That is a fair question, and it surprised me too. One way to think about it is in terms of decisions. Cohen and Swerdlik (2018) explain that validity coefficients describe how well scores predict a criterion across a group, not how accurate a decision is for one person. When a test with a modest correlation is used to make many decisions, such as screening thousands of patients, it can still improve accuracy substantially compared with chance. A correlation of .30 can mean noticeably more correct decisions than guessing, especially when combined with other information. So the coefficient looks small as a percentage of variance but can be large in practical effect. Would you be more comfortable with a modest test if it were one part of a larger assessment?

References

Cohen, R. J., & Swerdlik, M. E. (2018). Psychological testing and assessment: An introduction to tests and measurement (9th ed.). McGraw-Hill Education.

Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. R., Eisman, E. J., Kubiszyn, T. W., & Reed, G. M. (2001). Psychological testing and psychological assessment: A review of evidence and issues. American Psychologist, 56(2), 128-165. https://doi.org/10.1037/0003-066X.56.2.128