Paperonium

Nutrition · PLoS Med · 2021

The paper’s main result: store sales. The main result its trial publicly recorded: women’s diets.

The public recordfiled after the study ended, never changedThe paperMain resultSecondary resultWomen’s diet scoreat 3 monthsStore saleshealthier at 3 monthsStore salesWomen’s diet scorea borderline gain
The public recordfiled after the studyended, never changedThe paperMainresultSecondaryresultWomen’s diet scoreat 3 monthsStore saleshealthierat 3 monthsStore salesWomen’sdiet scorea borderline gain

Which result counts as the main one: the paper versus the trial’s public record

The paper says

A study gave discount supermarkets in England a healthier layout. The paper calls store sales its main result: after 3 months, fruit and vegetable sales rose and sweets sales fell. Its summary leads with these. What 150 women shoppers bought and ate is ranked secondary.

Its own data show

The trial’s public record names one main result: a score of how healthily women said they ate, at 3 months. In the paper’s Table 2, that score rose 0.29 more at new-layout stores than at comparison stores (1 = the usual spread between women). Range: 0.01 to 0.57, barely above zero.

Trials publicly record which result will decide whether they worked, normally before starting, so the choice can’t be made after seeing the results. This record was filed after the study ended, but was never changed, and the paper cites it. The paper never mentions that its main result differs, and its summary gives no diet numbers.

“The summary” means the paper’s abstract: the short summary printed at its top, which is the part most readers see.

Read the paper →Full technical write-up →

Altering product placement to create a healthier layout in supermarkets: Outcomes on store sales, customer purchasing, and diet in a prospective matched controlled cluster study · PLoS Med (2021) · doi:10.1371/journal.pmed.1003729