Educational Experiments

If you’ve ever been cured of a stomach ulcer, you can be thankful for the work of an Australian scientist, Dr Barry Marshall. As a young researcher, he and a colleague developed the idea that stomach ulcers were not just caused by stress, but by a bacterium that somehow managed to survive the hydrochloric acid in our digestive system.

Not all doctors accepted his idea. The issue was settled, in the end, by a series of experiments. Dr Marshall identified a collection of antibiotics he thought would be effective, and tried them out on a bunch of patients. In fact, he deliberately infected himself with the bacterium in question, got sick, and cured himself. For his discovery, he was awarded the Nobel Prize in Medicine in 2005.

If you’ve never known anyone who died of smallpox, you can be thankful for the work of Louis Pasteur, who developed the first smallpox vaccine. He had noticed that people who suffered from a mild disease called cowpox seemed to be immune to smallpox. So he deliberately infected a number of people, including children, with cowpox, then waited to see what percentage would die of smallpox.

This seems shocking – however, experiments like this are now routine in medicine. Because of that, we are learning more and more about what is effective at fighting disease and what is not. Millions of lives have been saved because a few were put at (fully informed) risk.

It took a while for this to become commonplace. When systematic experimentaion in medicine was first proposed, the objection of many professionals was that it would be “unethical”.


Are we stopping ourselves from discovering the best way to educate kids?
Are we stopping ourselves from discovering the best way to educate kids? Photo Credit : See Below*

In a  properly done experiment, there is a group of people who receive the proposed treatment. There is another group who are not given the treatment. Then, you use statistics to figure out whether one group did better than the other. This means you have to deliberately deny some people a treatment which you think might be promising. Worse, you should choose those people by lottery, so your choice doesn’t blur the information you hope to get from the trial.

This step “deny some randomly selected people a potentially promising treatment” seems unethical, but, as they say, “you can’t make an omelette without cracking some eggs”. It’s only unethical to deny people a treatment that is known to be effective. However, the whole point of doing the experiment is to determine if the treatment is effective. If it’s effective, it should be promoted far and wide. If not, the researchers should move on tot he next promising idea. The first step is to do the experiment. It’s surely even more unethical to delay this first step.

Nonetheless, it took many years for a culture of properly controlled experiments to pervade medical science.

Now, there’s a similar conflict in the social sciences, including education. There are groups wanting to do experiments to see what works well and what doesn’t, and there are others who hold back good research because of misguided ideas about ethics in education. A key to being a truly excellent teacher is to continually ask, about your own teaching, “does this actually work?”, and yet, as this blog post poignantly notes, that culture has not yet permeated all of education – there are still some who feel it’s unethical to deny students a merely promising educational methodology, even if the purpose of so doing is to find out if the methodology will fulfill its promise.

The ideal educational experiment would go like this.

  • A group of students, or classes, or schools, or districts is randomly split into two (or more) groups.
  • One of the groups is treated one way, the other (the control group) is treated some other more usual way.
  • Ideally, neither the people dealing with the subjects, nor the subjects themselves, know which group they are in (making it a double blind trial).
  • The results are collected, and analysed using correct statistical formulae.
  • The results are reported, whether or not the experiment is “successful”.
  • Others do the same experiment, in the same way or slightly different ways, and also report the results, whatever they are.

If there was a culture of experimentation, so that trials as described above were commonplace, and if “successful” methodologies were adopted widely, it would be fantastic. The next generation would become the best-educated generation the world has ever seen, just as medical experimentation in past generations has made ours the healthiest.

Keep that in mind, and whenever you see signs of a culture of experimentation in the education sector, fan it into flames!

* Photo Credit :