Inoculation Theory

What it is

The original restriction mattered: inoculation was for beliefs that had never been challenged, on the reasoning that an untested belief is the vulnerable one.

The modern programme extends it past that restriction, arguing that incubation periods vary and people are at different stages of exposure, so a prophylactic intervention can be given to a population regardless of what they already believe. Sander van der Linden's Foolproof sets out that extension, and it is the book's real theoretical contribution.

Its honest corollary, which the book itself supplies, is that in real life you cannot know anybody's infection status, which makes the extension harder to operationalise than it is to state.

In practice the delivery has been browser games, Bad News and Go Viral, in which the player takes the role of a disinformation merchant and learns the techniques by using them. Fifteen minutes, dozens of languages, adopted by governments and platforms.

In effect

The evaluations compared how participants rated true and fake news items before and after playing, usually against a control group playing Tetris or doing nothing, and reported the fall in ratings of the fake items.

Ariana Modirrousta-Galian and Philip Higham pointed out that this cannot distinguish two different outcomes, and reanalysed the existing studies with receiver operating characteristic analysis. The games did not improve discrimination between true and fake news. What they moved was response bias: people came out more sceptical of everything.

A later study by Tina Seabrooke with the same authors re-examined the Bad News game using Indian headlines and reported no evidence of improved discrimination.

What it does not say

It does not say inoculation theory is wrong. The theory is far larger than two games and nothing here speaks to its other applications.

It does not say the games are worthless, only that the effect is not the advertised one.

It is one research group, which is the same limitation the critics identify in the positive literature.

And the cost of raising scepticism indiscriminately has not been measured by anybody in this exchange. If an intervention makes people distrust accurate reporting as readily as fabricated reporting, that is a harm, and the net effect is an open question.

---