14 November 2014

Science Increases Trust

Welcome back. Suppose you’re out buying food for your pet unicorn. You walk into my pet store, and I tell you, “A survey showed that 8 out of 10 unicorns prefer the new crunchy grass mix over the old creamy grass mix.”

Now suppose that, in addition, I show you either a graph (see figure) or formula that depicts that statistic:

Creamy Mix (20%) + Crunchy Mix (80%) = Unicorn Grass Preference (100%) 

Graph illustrating 8 out of 10
unicorns prefer crunchy grass
mix over creamy grass mix.

Would the graph or formula help you choose which grass mix to buy? Would your feelings toward science make a difference?

Last Friday’s blog post highlighted a study that found scientists, and by extension, science, may be respected but not trusted (Trust a Scientist). Here’s a new twist on trust in science.

Trivial Scientific Data

A recent report by Cornell University investigators documented how adding scientifically formatted data to a sales pitch for a product--even if it provided no new information--increased trust in the product. This isn’t the old Madison Avenue obfuscation with more or less scientific, irrelevant information; this is simply adding unnecessary, redundant detail that looks scientific.

The investigators conducted three separate studies. For the first study, 61 participants read a paragraph about a new medication. Half of the participants were also given a graph that provided no additional information. When the participants were later asked if the medication reduced illness, about 30% more said yes if they’d seen the graph.

The second study gave 56 participants either the same paragraph and graph or the paragraph without the graph but with an added sentence that repeated that the medication reduced illness by 20%. All participants were later asked (i) to estimate how much the medication reduced illness and (ii) how much they agreed with the statement, “I believe in science.”

When it came to recalling the percentage by which the medication reduced illness, there was no difference between those who had seen the graph and those who had read the added sentence. Yet the participants who saw the graph and professed a belief in science expressed the strongest confidence in the medication’s effectiveness. Scientific-looking information--the graph--was perceived as true even if it had no affect on information retention.

The third study gave 57 participants the same paragraph, but half of the participants were also given the chemical formula of the medication’s active ingredient. Although the paragraph stated that the new medication would work 5.9 hours versus the old medication’s 3.8 hours, those who saw the formula believed that the new medication would work an additional two hours longer.

Wrap Up

The investigators concluded that adding scientific-looking data increased confidence in the medication’s efficacy. Since the data didn’t improve comprehensibility or recall, the apparent reason was that the data were linked with science and that science conveys truth, especially for some people.

However you feel about science, the next time you’re out buying food for your pet unicorn, make sure to read the ingredients carefully and don’t be swayed just because something looks scientific. Grass mixes are not the same to unicorns. Thank you for shopping in my pet store. Please stop by again.

P.S.


Cornell report in Public Understanding of Science:
pus.sagepub.com/content/early/2014/09/23/0963662514549688
Press release on the report and a department summary:
www.newswise.com/articles/blinded-by-non-science-trivial-scientific-information-increases-trust-in-products
foodpsychology.cornell.edu/OP/trivial_graph
Article on the report on Science Daily website:
www.sciencedaily.com/releases/2014/10/141017101224.htm
Background on unicorn food:
www.unicorns.co.za/unicorn-facts/unicorn-food.html

No comments: