Response bias makes balanced surveys appear spurious multifactor structures

Evidence

By the same token, yea-saying (and naysaying) would reduce the correlations between protrait and contrait items on balanced tests (for example, +3 on a protrait makes you look high on the trait; +3 on a contrait makes you look the opposite). Thus balanced tests would look Jess internally consistent than the content really is. And the reduction in correlation between protrait and contrait items would tend to separate them onto different factors in a factor analysis, even if they were largely tapping the same thing. — Bob Altemeyer, The Authoritarian Specter, 779

Categorization

• XREF: Explains why factor analyses of personality and ideology surveys may over-report distinct dimensions when acquiescence is uncontrolled. — connects to: Treating Likert responses as interval data distorts estimates when aggregating across disparate items — Ruiqian Li and Paul Froese, The Duality of American Christian…, 96