Know Thyself: Is This Possible?

Excerpt

As researchers in science teacher education, and science education research in general, we often consider the sample size of our studies and the studies of others. The ultimate concern is how much we can believe or generalize the results of a study with a small sample size. What constitutes an adequate sample size is a perennial argument and how much a small sample size can establish “believability” has just as long a history (Edgar & Billingsley, 1974; Krathwohl, 2009). Consider two common situations relative to this debate. There are 300 people in the audience of a magic show and the magician appears to pull a rabbit out of a hat that was previously empty. All 300 people have no problems testifying that the rabbit suddenly appeared. They can provide no explanation, but they have no problems agreeing on what they have seen. Alternatively, a woman makes visits her doctor because she is feeling a consistent pain in her abdomen. It is especially acute after eating a meal. The doctor checks for ulcers and runs a series of other diagnostic tests, but can find no physiological or anatomical evidence for the pain. The doctor tells his patient that nothing is wrong, but the she insists she feels the pain. Should we believe the 300 people in the magician’s audience more than the single woman, just based on sheer numbers of observations? In addition, the woman is observing something about herself, as opposed to what others have agreed they have seen external to themselves. This immediately raises questions about potential bias. Returning to science education research and research in science teacher education, small sample sizes and research on oneself raise concerns that clearly exceed what is seen in “typical” and “rigorous” traditional research designs. In some ways this harkens back to the contentious debates about the value of qualitative research (Jacob, 1987; Peshkin, 1993; Rist, 1977). …