Thursday, July 11, 2019

Ethics Classes Can Influence Student Behavior: Students Purchase Less Meat after Discussing Arguments for Vegetarianism

by Eric Schwitzgebel, Bradford Cokelet, and Peter Singer

[poster presentation for the Society for Philosophy and Psychology, 2019 meeting]

All data and analyses are tentative, pending final checks and peer review.


Work by Haidt (2012) and Schwitzgebel & Rust (2016) suggests that philosophical ethical argumentation might have little influence on real-world moral behavior.

However, to our knowledge there are no existing ecologically valid studies of the influence of university-level philosophy classes on non-laboratory, non-self-reported moral behavior.


1143 undergraduates in four large lower-division classes at U.C. Riverside.


Half of students were required to read a philosophy article defending vegetarianism (Rachels 2004), followed by a group discussion section and an optional advocacy video.

The other half received similar materials and discussion on charitable giving.

Opinion Measure

Later that week, students received an anonymous questionnaire asking their opinion about four moral issues.

The target question was “Eating the meat of factory farmed animals is unethical” with response options from “strongly agree” (+3) to “strongly disagree” (-3).

Behavioral Measure

We examined campus dining card purchase data for 5,981 campus food purchases from 476 students for whom data were available.

Only purchases of at least $4.99 were included.

Purchases were coded as either vegetarian or non-vegetarian.

Results: Opinion

In the meat ethics group, 43% of respondents agreed (+1 to +3) that eating the meat of factory farmed animals is unethical, compared to 29% in the charitable giving control group (z = 5.0, p < .001; mean +0.12 vs -0.46, t(1029) = 5.6, p < .001).

Responses to other ethical questions, including one about charitable giving, did not differ between the groups.

Results: Purchase Behavior

In the control group, 52% of purchases included meat, both before and after the discussion section.

In the meat ethics group, meat purchases declined from 52% to 45% (z = 3.3, p = .001).

Participant-by-participant among students with purchases both before and after the discussion:

  • The control group averaged 53% meat purchases both before and after.
  • The meat ethics group averaged 56% meat purchases before and 45% after (paired t(162) = 4.3, p < .001).

    Tyler John said...

    Cool! What was the time frame for the behavioral measure? Immediately following the class reading? Would be very interested to see follow up studies on their purchases 1 and 6 months out.

    Eric Schwitzgebel said...

    For most participants, it was a few weeks' purchases. I'm seeing if I can get access to data for a longer stretch of time.

    Zeke said...

    Fascinating, thank you for this work!

    I'm wondering about observer effects—were students aware that their meal purchase data might be examined? And why was this data available for fewer than half of the participants?

    Eric Schwitzgebel said...

    Thanks, Zeke! No, not aware other than the general knowledge that the university and card companies track their purchases. (Yes, we had IRB approval and took cautions not to know any individual's purchase behavior.) Less than half because not all students use their cards for purchases.

    Zach Freitas-Groff said...

    This is awesome, wish I'd seen it sooner! Is there any evidence on specific changes in consumption of different animal products, especially poultry and fish?