Risk Attitudes, Randomization to Treatment, and Self-Selection into Experiments

By Glenn W. Harrison, Morten I. Lau, E Elisabet Rutström. 2005

200 citations recorded by the source corpus.

Randomization to treatment is fundamental to statistical control in the design of experiments. But randomization implies some uncertainty about treatment condition, and individuals differ in their preferences towards taking on risk. Since human subjects often volunteer for experiments, or are allowed to drop out of the experiment at any time if they want to, it is possible that the sample observed in an experiment might be biased because of the risk of randomization. On the other hand, the widespread use of a guaranteed show-up fee that is non-stochastic may generate sample selection biases of the opposite direction, encouraging more risk averse samples into experiments. We undertake a field experiment to directly test these hypotheses that risk attitudes play a role in sample selection. We follow standard procedures in the social sciences to recruit subjects to an experiment in which we measure their attitudes to risk. We exploit the fact that we know certain characteristics of the population sampled, adults in Denmark, allowing a statistical correction for sample selection bias using standard methods. We also utilize the fact that we have a complex sampling design to provide better estimates of the target population. Our results suggest that randomization bias is not a major empirical problem for field experiments of the kind we conducted if the objective is to identify marginal effects of sample characteristics. However, there is evidence that the use of show-up fees may have generated a sample that was more risk averse than would otherwise have been observed.

Subfields: General Decision Sciences, Safety Research, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/c4ba3c9f6969471006739bb09a856a1f77ddb54d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In General Decision Sciences

In Safety Research

In Statistics and Probability

Citations in the corpus

Subject pool recruitment procedures: organizing experiments with ORSEE (2015, background): “Harrison et al. (2009) varied the spread of earnings advertised in the recruitment e-mail sent to potential subjects, and found that participants who signed up in response to a recruitment e-mail advertising a lower variance were less willing to take risks in the laboratory experiment.”
Viewpoint: On the Generalizability of Lab Behaviour to the Field (2007, unknown): “In a study measuring individual risk propensities, Harrison et al. (2005) find that the use of show-up fees to increase experimental participation generates a sample of more risk-averse participants.”
Testing the Waters: Behavior across Participant Pools (2018, background): “This work shows that selection into lab experiments from broader student populations— such as those taking introductory economics—is not related to risk aversion or generosity (Harrison et al., 2009; Cleave et al., 2013; Falk et al., 2013).”
The Roles of Risk and Ambiguity in Technology Adoption (2014, background): “Harrison et al. (2005) present evidence that risk preferences among students are stable over a six month period while Love and Robison (1984) present similar evidence for Midwestern farmers over a 2 year period.”
Experimenting with Measurement Error: Techniques with Applications to the Caltech Cohort Study (2019, methodology): “This follows the same protocol as the risk MPLs described above, except participants are told that the number of red balls is uniformly drawn between 0 and 20 for the ﬁrst urn and between 0 and 30 for the second.”
Does Management Matter? Evidence from India (2011, background): “Harrison, Lau and Rustrom (2009) find that these effects are relatively small in the class of experiments they examined, while Lazear, Malmendier and Weber (2012) find stronger evidence of the effects of such self-selection into experiments measuring social preferences.”