Researchers often treat four short scales of “conspiracy mentality” as substitutes for one another. All four do reflect one common trait, which accounts for about 60 percent of the variance in belief in 21 specific conspiracy theories. But two of the scales carry something more, so they are not strictly interchangeable.
A scale can stand in for the others only if it reflects the common trait, carries no extra content that shapes answers to specific conspiracy questions, and scores people the way the common trait does. Hover over or tap a mark for the evidence.
Each of the 21 conspiracy theories in the BCTI regressed on the common factor and the four scale-specific factors. “All factors” shows the five estimates for each theory together. Choose a factor to bring it forward with its 95% interval; the others stay in place, faded. The common factor predicts every theory. The GCB-5 leftover is largest for the two alien theories. The GMC leftover predicts many theories a little and the four partisan-coded ones the most. These estimates use factor scores, so they run smaller than the paper’s latent-model estimates (.34 here for the alien items, against .52 and .53).
Single items should be read with care. Outside the GMC, the scale-specific loadings were weak and inconsistent in sign, so a large value on one item, such as CMQ.2, does not mean the scale carries coherent extra content. The scale-level results above are the ones the paper’s conclusions rest on.