The Effects of Consensus Messages and the Importance of a Pre-Test: A Comment on Chinn and Hart (2021)
Chinn and Hart (2021) argue that their experiment on the effects of communicating the scientific consensus on climate change revealed “mixed” and “inconsistent findings”. We note that Chinn and Hart (2021) provide clear and consistent evidence that the scientific consensus message has positive indirect effects on climate beliefs, attitudes, and support for public action, and that these effects are more pronounced among conservatives (as predicted). Importantly, however, the authors’ claim that these positive findings might simply be the result of an experimental design choice: the use of a pre-test of the dependent variables. Here we argue that no convincing evidence is provided for the conclusion that pre-tests should not be used. In fact, contrary to the authors’ recommendation, we make the case that the inclusion of a pre-test in randomized controlled designs increases power and precision. Furthermore, we point to its central importance for the ability to capture and evaluate the Gateway Belief Model’s (GBM) key predictions about changes in public attitudes.