Correction: An Examination of the Factorial and Convergent Validity of Four Measures of Conspiracist Ideation, with Recommendations for Researchers

Abstract

Figures

Full text

After publication of this article [1], the authors discovered that due to an error in data processing, the reported fit of the 5-factor Generic Conspiracist Beliefs Scale (GBCS) model was incorrect at time of original publication.

With this Correction, the authors report the corrected statistics and path diagram as initially appeared in Fig 4. See the authors’ description of these changes below:

“Using the correct item enumerators (matching those reported in Table 6), the fit statistics for the 5-factor model are: χ²(80, N = 401) = 255.611, χ²_normed_ = 3.195, CFI =.958, RMSEA =.074 with 90% CI =.064-.084, SRMR =.048 (see Figure 1).” Please see the corrected Fig 4 here.

Please see Table 8 here showing the previously unreported total-sample bivariate correlations among the 5-factor scores of the GCBS with 9/11 conspiracist beliefs and anti-vaccination beliefs.

The 5-factor model of GBCS scores in our second split-half subsample may have better fit than we previously indicated. However, given support elsewhere (and in our first split-half subsample) for a 2-factor model [2], a 3-factor model [3], a bi-factor model [4], as well as for a correlated 5-factor model [5] of GCBS scores, it appears that the observed dimensionality of this instrument lacks consensus, is unstable (if not erratic), thus suggesting that it may well depend on: (i) sampling frame and sample characteristics, (ii) analytic approach, (iii) survey language, (iv) recent time trends (increases) in the prevalence (endorsement rate) of conspiracy beliefs in the population, or (v) any combination thereof. Accordingly, we repeat the advice from our original report: “Scholars who wish to measure generic conspiracist ideation may find it better to use the GCBS, but they should pay careful attention to (and report) its factor structure within studies. Our findings suggest the possibility that conspiracist ideation may be multi-dimensional and, as a result, scholars should not assume that the GBCS – or other measures of conspiracist ideation – are necessarily one-dimensional.”