Eliminating Bias

When you are analyzing and presenting your findings, remember to work to eliminate bias by being truthful and as accurate as possible about what you found, even if it differs from what you expected to find. You should see your data points—whether they are quantitative and/or qualitative—as sources of information, just like sources you find in a library, and you should work to represent them accurately.

Remember: your findings—whatever they are—are your findings, and you need to represent them as such, even if they are unexpected or contradict prior findings or your hypotheses. Some very important discoveries have happened when the experimental data have not supported a hypothesis. For example, the failed Michelson-Morley experiment led to research that developed our understanding of special relativity and the early failures in the Geiger-Marsden experiment led to the discovery of the atom’s nucleus. Outside of a science lab, we can examine Christopher Columbus’s failed trip to India in a similar light: he sailed west to find a faster route to India and instead ran into something quite unexpected: the Americas.

Not all unexpected results, of course, end up leading us to beneficial discoveries. Sometimes a survey sample group just doesn’t represent the whole very well, or an oddity in an observational space makes it not function in the usual way. When that happens, your purpose becomes to document what you did, what you found, and what you think it might mean—all as an attempt to develop/maintain your reputation as an ethical researcher.

As you examine and incorporate your data, then, your primary ethical goal is to represent it accurately. Your job as a researcher is to present the data as it is and to offer a plausible interpretation of that data through your analysis of it.