Heuristics

What it is

The earlier account, from L. J. Shrum, Robert S. Wyer and Thomas C. O'Guinn in 1998, was memory-based: television stocks the set of examples available when somebody is asked a question about the world and the error is a failure to discount where those examples came from, which is a heuristic process and should therefore weaken when the person thinks carefully, so attention at the moment of judgement is corrective. The 2005 paper tested whether the same applied to a value rather than to an estimate of how common something is, using data from both a survey and an experiment. In the authors' words, television cultivates materialism through an online process in which television's influence is enhanced by active rather than heuristic processing during viewing, and this stands in contrast to judgements of the prevalence of societal affluence, which are attenuated by active processing during judgement elicitation. So two things differ at once, because the outcome is a value rather than a frequency estimate and the processing that matters happens during viewing rather than at the moment of being asked.

In effect

The practical consequence is that attention is not a general antidote, because for a prevalence judgement thinking harder helps, the correction being to notice that your examples came from a screen, while for a value thinking harder during viewing is the thing that installs it. That also makes the two outcomes non-interchangeable in any argument, since a claim that heavy viewers overestimate how many people are rich and a claim that heavy viewers are more materialistic are different claims with different mechanisms, and evidence for one is not evidence for the other. It is a useful constraint on media-literacy advice too, because telling somebody to watch attentively and critically is good advice against one of these mechanisms and, on this result, the wrong advice against the other.

What it does not say

It does not show that viewing causes either outcome, because the 2005 paper includes an experiment and what the experiments in this line of work manipulate is processing and priming rather than how much television anybody watches, so nobody was assigned to watch more or less. The details are not available here, which limits how far the result can be pushed: no sample size for either the survey or the experiment, no description of the experimental manipulation and no effect size was retrievable, so the direction of the reversal is reported and its magnitude is not. It is one paper on two outcomes rather than a mapped boundary, because where exactly a judgement stops behaving like a frequency estimate and starts behaving like a value is not established by it, and nothing here says the pattern generalises to other values or other estimates. And it should not be read as an account of heuristics in general, since the parent idea is much broader than this result, which concerns one specific contrast inside one media literature.

---