Randomization

An excellent way to avoid confounding is to assign individuals to the treatmentand control groups at random, and then administer the treatment to those whowere assigned to the treatment group. Randomization keeps the two groups similarapart from the treatment.

If you are able to randomize individuals into the treatment and control groups,you are running a randomized controlled experiment, also known as arandomized controlled trial (RCT). Sometimes, people’s responses in anexperiment are influenced by their knowing which group they are in. So you mightwant to run a blind experiment in which individuals do not know whether theyare in the treatment group or the control group. To make this work, you willhave to give the control group a placebo, which is something that looksexactly like the treatment but in fact has no effect.

Randomized controlled experiments have long been a gold standard in the medicalfield, for example in establishing whether a new drug works. They are alsobecoming more commonly used in other fields such as economics.

Example: Welfare subsidies in Mexico. In Mexican villages in the 1990’s,children in poor families were often not enrolled in school. One of the reasonswas that the older children could go to work and thus help support the family.Santiago Levy , a minister in Mexican Ministry of Finance, set out toinvestigate whether welfare programs could be used to increase school enrollmentand improve health conditions. He conducted an RCT on a set of villages,selecting some of them at random to receive a new welfare program calledPROGRESA. The program gave money to poor families if their children went toschool regularly and the family used preventive health care. More money wasgiven if the children were in secondary school than in primary school, tocompensate for the children’s lost wages, and more money was given for girlsattending school than for boys. The remaining villages did not get thistreatment, and formed the control group. Because of the randomization, therewere no confounding factors and it was possible to establish that PROGRESAincreased school enrollment. For boys, the enrollment increased from 73% in thecontrol group to 77% in the PROGRESA group. For girls, the increase was evengreater, from 67% in the control group to almost 75% in the PROGRESA group. Dueto the success of this experiment, the Mexican government supported the programunder the new name OPORTUNIDADES, as an investment in a healthy and welleducated population.

In some situations it might not be possible to carry out a randomized controlledexperiment, even when the aim is to investigate causality. For example, supposeyou want to study the effects of alcohol consumption during pregnancy, and yourandomly assign some pregnant women to your “alcohol” group. You should notexpect cooperation from them if you present them with a drink. In suchsituations you will almost invariably be conducting an observational study, notan experiment. Be alert for confounding factors.

This page was created by The Jupyter Book Community