Page:What Are Conspiracy Theories? A Definitional Approach to Their Correlates, Consequences, and Communication.pdf/8

 literature—research into the causes of conspiracy belief—is marred by some problems. One of these is inconsistent findings, such as those emerging from tests of the hypothesis that lacking personal control affects belief in conspiracy theories. While some research supports this claim (e.g., Whitson & Galinsky 2008), other research does not, finding no causal effect at all (e.g., Hart & Graether 2018). Using a general measure of conspiracy belief that does not refer to specific events, Stojanov et al. (2020) found no evidence that a lack of control causes people to believe more in conspiracy theories. This is not the only inconsistency to have emerged. There are also inconsistent findings, for example, concerning the relationship between right-wing authoritarianism and conspiracy beliefs (e.g., Dyrendal et al. 2021, Wilson & Rose 2014; see Biddlestone et al. 2022 for a meta-analysis of the predictors of conspiracy beliefs).

These inconsistencies are an important obstacle to answering some of the most important questions about conspiracy beliefs. We want to understand, for example, why conspiracy theories seem to have an increasing appeal to people; we want to know what societal and psychological effects they are having; and we want to understand the relevance of conspiracy theories to applied fields such as clinical, forensic/criminological, and political psychology. A complete understanding of these questions requires us to be able to profile people who are prone to believing in conspiracy theories—a task that ought to be possible by now, given the hundreds of studies into psychological, demographic, and political risk factors (e.g., Freeman & Bentall 2017). However, inconsistent results make it very difficult to build such a profile.

Understanding the causes and consequences of conspiracy belief has also been hampered by methodological problems—especially, the dominance of cross-sectional research. This means that the consequences of particular conspiracy theories are often inferred, and reverse causal directions cannot be ruled out (e.g., do conspiracy theories about vaccines reduce vaccine intentions, or do people’s vaccine intentions inform their conspiracy beliefs?). Experimental studies of the effects of exposure to conspiracy theories remain relatively rare, and reverse causal directions are typically not tested. Longitudinal research on the effects of conspiracy theories is also sparse (for recent exceptions, see Bierwiaczonek et al. 2022, Liekefett et al. 2021).

Problems such as inconsistent and difficult-to-interpret findings have been made worse by a lack of attention to the conceptualization and measurement of belief in conspiracy theories. Stojanov et al. (2020) argued that inconsistent results related to the effects of control on conspiracy theorizing may be due to the way conspiracy belief was measured in the different studies. Indeed, as we mentioned earlier, some researchers measure belief in conspiracy theories as a combination of agreement with several common conspiracy theories (e.g., Douglas et al. 2016, Swami et al. 2014), while others tap into more general notions of conspiracy, such as the idea that governments generally conspire, rather than referring to any specific government or event (e.g., Brotherton et al. 2013). The choice of measure often seems arbitrary (Imhoff et al. 2022a), and sometimes care has not been taken to ensure that the chosen items truly are measures of belief in conspiracy theory, according to an explicit and principled definition. Seemingly little attention has been paid to capturing the features of the alleged conspiracies (e.g., their agency, malevolence), and in some cases the alleged conspiring group is not identified. For example, there is no conspiring group identified in the statement “The virus is a scaremongering tactic to prevent Brexit” (see Freeman et al. 2022). When scales are developed without reference to a stable, reasoned, and explicit definition of conspiracy theories, there will always be the risk of inconsistent measurement, and therefore inconsistent results, between studies.

Arguably, a still more serious downside of this lack of careful conceptual articulation is that strictly speaking, we know surprisingly little about the causes, consequences, and transmission of belief in conspiracy theories per se. Conspiracy theories overlap partially with many other categories of belief or representation. For example, many if not all conspiracy theories overlap with 278Douglas • Sutton