Convergent and discriminant validation by the multitrait-multimethod matrix.¶
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discriminant validation by the multitrait-multimethod matrix. Psychological Bulletin, 56(2), 81-105.
Cited by¶
6 citations across 6 artifacts.
Each citation links to the sentence it supports in the citing article.
Primes¶
- Construct Validity
- Convergent and discriminant validity together approximate a small factor-analytic argument: if everything correlates with everything, the proxy is not isolating the construct; if nothing correlates with anything, the proxy is not measuring a real construct at all.
This sourceIntroduces the multitrait-multimethod matrix and the convergent/discriminant distinction for locating method-variance leaks.
- Convergent and discriminant validity together approximate a small factor-analytic argument: if everything correlates with everything, the proxy is not isolating the construct; if nothing correlates with anything, the proxy is not measuring a real construct at all.
- Cross-Dimensional Leakage
- Measurement and instrumentation: a single instrument's drift, calibration error, or operator effect contaminates multiple measured quantities; multi-trait/multi-method matrix designs were built to separate trait from method variance.
This sourceIntroduces the MTMM matrix to separate trait variance from shared method variance — the canonical 'cross the channel, then partition' design, distinguishing same-method/different-trait from same-trait/different-method correlations.
- Measurement and instrumentation: a single instrument's drift, calibration error, or operator effect contaminates multiple measured quantities; multi-trait/multi-method matrix designs were built to separate trait from method variance.
- Triangulation
Domain-specific¶
- Halo Effect
- The practical consequence, documented by Campbell and Fiske's (1959) multitrait-multimethod framework, is that cross-attribute correlations in rating data are upper bounds on — and not estimates of — the underlying trait correlations; a substantial portion of the apparent correlation is judge-side method variance, not joint trait variance
This sourceThe multitrait-multimethod framework, which shows that a test score's systematic variance mixes trait content with measurement-method variance, so cross-trait correlations from one method are inflated and cannot be read as trait correlations.
Supported in partVerified against a saved copy of the source
“The systematic variance among test scores can be due to responses to the measurement features as well as responses to the trait content.”
- The practical consequence, documented by Campbell and Fiske's (1959) multitrait-multimethod framework, is that cross-attribute correlations in rating data are upper bounds on — and not estimates of — the underlying trait correlations; a substantial portion of the apparent correlation is judge-side method variance, not joint trait variance
Mechanisms¶
- Multitrait-Multimethod Matrix
- Its strength is that it needs no statistical model and exposes method variance a single-method study cannot see: it is the cleanest way to answer "would these constructs still look related if a different instrument had measured them?"
This sourceUses multiple traits and independent methods to reveal method variance and test whether relationships persist across instruments without requiring a fitted latent model.
- Its strength is that it needs no statistical model and exposes method variance a single-method study cannot see: it is the cleanest way to answer "would these constructs still look related if a different instrument had measured them?"
Verification¶
Does it exist? Confirmed. This work's DOI resolves to a registered record, which fixes its identity. That is all it fixes.
Does it back the claim? Read against the text for 1 of 6 citations: 1 supported in part. Each verdict is shown under its citation below, with what in the work backs the sentence.
Was it audited? Yes. A second, independent pass read the citation against the article text and recorded a verdict.
Support is checked per citation rather than per work — the same source can be cited soundly in one article and wrongly in another. Per-citation recording began recently, so a citation with no recorded check is a gap in the record rather than evidence it went unchecked.
See how references were verified.
Registry ID ref:352626f18e80 · see in the full table