This week I gave a research seminar at the Department of Economics at Cornell University, on the impact of introducing Harriet, our ECONS101 tutor, on student performance and experiences in ECONS101. This was a mildly updated version of talks I gave at the EduLearn Conference and at the University of Exeter earlier this year. The audience was small, but very engaged. There is more to come on this project, as once I have collated the 2026 data, I'll be writing it up for publication.
Aside from that, here's what caught my eye in research over the past week:
- Süer et al. (open access) find that, when presented with a social dilemma game (like a probabilistic version of the prisoners' dilemma), women's behaviour is more strongly affected by a recommendation on socially optimal behaviour than is men's, which they interpret as women being more 'compliant' than men
- Brodeur et al. (with ungated earlier version here) map the evolution of economics research from 1998 to 2019, showing that awareness of reproducibility has increased modestly, sample sizes have increased, and econometric methods and causal inference practices have become more sophisticated, but pre-analysis plans, and power tests to accompany null results remain uncommon
- Harvey (with ungated earlier version here) finds that, in the 1707 Scottish parliament vote for union with England, eligibility for compensation increased the probability of a parliamentarian voting for union by over 30 percent
- Brülhart et al. (with ungated version here) find that housing rents within 0.7km of an active asylum seeker hosting centre in Switzerland are on average 3.8 percent lower than in the control group, and that the effect is significantly more pronounced when centres host a higher share of asylum seekers from sub-Saharan Africa
- Barrett and Tan (with ungated earlier version here) find that doubling the rate of immigration at the city level in the US reduces inflation by about 0.1 to 0.2 percentage points, and that the effects are larger for working-age immigrants
- Dickinson (open access) finds limited evidence that there is a difference in moral choices between daytime (3pm-5pm) and night-time (3am-5am) for choices that harm another participant, but that at night-time there is more dishonesty in a task that doesn't harm other participants
No comments:
Post a Comment