Universidad Nacional de Educación a Distancia » Docencia » Directorio Docente » Bias Research Project

Annual reports

First year report
Second year report
Third year report
Final Report

Bias in experiments

"Biases in experiments with humans in the social and biomedical sciences" is a research project funded with 30.000 EUR by the Spanish Ministry of Science (FFI2011-28835). The members of the research team are Jason Dana (UPenn), Francesco Guala (Milan), María Jiménez Buedo (UNED) and David Teira (UNED). María González Moreno (U. San Pablo-CEU), Juan Carlos Hernández Clemente (UAM) and José Lázaro (UAM) are also affiliated part-time. Alex Díaz (UNED) is our graduate research fellow.

Research statement

In this research project we want to analyse how scientists deal with the biases that arise in experiments with humans, be it on the side of the experimenter or on the side of the experimental subject. The project proceeds in three stages. In the first stage, we want to construct a general concept of experimental bias, developing an epistemic justification for the debiasing rules actually applied in experiments. Then, we want to study how these rules work in experiments with humans in the biomedical sciences (namely, clinical trials) and the social sciences (namely social psychology and experimental economics).

This comparative analyses will focus on two types of biases, the funding bias (second stage) and experimenter effects (third stage). The former arises when the results of the experiment tend to coincide with the interests of the sponsor. The latter is generated by the reaction of the participants trying to act in accordance to the purported expectations of the experimenter. We will study to what extent the debiasing rules implemented in a biomedical or social experiment can actually control for these types of biases, weighting the analogies and differences of experiments with humans across disciplines.

The main practical contributions of this project are two. First, we want to increase the awareness of researchers in experimental economics about biases, drawing on the experience of clinical trialists and social psychologists. Certain experiments with humans are usually considered the gold standard of any evidence-based policy (such as clinical trials in drug regulation): our second goal is that of showing that there are limits to their impartiality and, therefore, to the use we can make of their conclusions in the policy-making process.

You can download the full project here.

You can download the first year report here.

Event Log | Publications