Listening to the “noise” in the data

Abstract

The term “noise” is often applied to the seemingly random variability that always appears in human data, and which is assumed to be of no interest to the researcher. Some of this variability is unavoidably due to measurement tools or the way in which we use them, and some is due to the unstable nature of human behaviour. In such cases, we may be justified in treating the variability as irrelevant noise. However, we cannot assume that all inexplicable variation is unimportant. Using examples from earlier research, I will argue that individual variability is a phenomenon worthy of study in its own right. Not only can it help us understand the nuances of the learning process, but giving it careful consideration can be a valuable step in determining how to effectively apply research findings in pedagogy.