In recent years, many interventions have been developed to reduce people's
susceptibility to misinformation, such as media literacy programs or gamified
inoculation. Researchers have typically evaluated the effectiveness of these
interventions based on a simple discernment measure, i.e., the difference
between true and false news ratings. These measures have been shown to be
problematic, as they conflate sensitivity with response bias. Here, we will
present the results of a meta-analysis re-assessing the effectiveness of these
interventions with a Signal Detection Theory (SDT) framework. This will allow us
to differentiate between two different kinds of intervention effects. First, the
effect on sensitivity, which is the (true) ability to discriminate between true and
false news. Second, the effect on response bias, i.e., the extent to which
participants become generally more or less skeptical in their accuracy ratings,
such as rating all news as falser. We will also test potential moderators of these
effects, such as the political concordance of the headlines. We will run an
Individual Participant Data meta-analysis (IPD) based on a sample of studies
that we identified via a systematic literature review following the PRISMA
guidelines. We use a two-stage approach: First, we extract individual participant
data and run a Signal Detection Theory analysis separately for each
experiment. Second, we run a meta-analysis on the experiment-level outcomes.
Our findings will provide methodological insights to improve these interventions
as well as practical recommendations on which intervention work best and for
whom.