Industry-funded media research cherry-picks designs that guarantee null findings.

Evidence

The television industry tried its hand with research of its own by sponsoring several field studies. The findings of one such study by Feshbach and Singer (1971) were of questionable reliability because of major methodological flaws (Bandura, 1973). Replication of the research without the methodological defects yielded findings that were not to the industry’s liking (Wells, 1971). The study was never published. With financial sponsorship and co-production by CBS, Milgram and Shotland (1973) conducted studies showing that exposure to modeled thievery does not lead viewers to steal money from a charity box called Project Hope, a medical charity that treats poor children worldwide. The charity box, with a dollar bill hanging from it, was mounted on a poster showing a physician treating a little girl and a picture of the hospital ship with the words “Where there is Hope there is life.” This experimental setup was analogous to demonstrating that televised modeling of theft has no adverse effects by showing that viewers will not rip off a tiny charitable contribution to Mother Teresa. — Albert Bandura, Moral Disengagement, 904

Categorization

• CANDY: The detail that a CBS-funded study tested whether people would steal from a Mother Teresa charity box is a striking, quotable example of rigged research design.
• DEEPTHINKING: The unstudied replication that contradicted the industry line and was quietly never published is a foundational example of publication suppression through funding leverage.
• XREF: connects to: Industry cherry-picking is continuous with ordinary scientific publication bias — Cailin O’Connor and James Owen Weatherall, The Misinformation Age: How False…, 359