Sixteen Approaches for Deconstructing Theory, Research, or Literature Reviews Claiming a “No Difference” (Null) Result in the Social Sciences
Although I have been unable to locate the original source, many years back I read a newspaper editorial that cited the wisdom of a Native American chief. He said that in their system of governance, the chief had advisors whose job was to actually advise the chief, regardless of what the chief initially desired, with each person following his conscience under the guidance of the Great Spirit. He went on to say that in the White man’s system, leaders had servants rather than advisors. The goal of these servants was to tell the White leader what he wanted to hear (no matter the guidance of the Great Spirit or the actual facts) and provide justification[1] for the leader’s desires for the people under the leader. In contrast, this Chief said, the White man’s leaders call forth their doctors, lawyers, scientists, clergy, or other “experts” to justify—rather than to examine and possibly challenge or disagree with—the leader’s pre-existing opinions on the decision at hand. Today, that chief’s prophetic views are confirmed regularly, as both sides in any political debate almost always claim to be following “the science,” even if their views are incorrect scientifically. In other words, science, and especially social science, often becomes a servant to politics—that is, to politically powerful interests, often with substantial funding and numerous publication outlets—rather than playing a truly independent role. The goal of this report is to explain, in a small part, how to avoid being a mere servant, and instead take on a role that has great potential to offend almost everyone except those genuinely interested in truth. This effort will involve, as feminists love to call it, “deconstructing” common attempts to “prove” the null hypothesis, otherwise referred to as the “no difference” hypothesis, an example quite possibly of type 2 error. (Type 2 error represents the risk of finding evidence for the null hypothesis when the null is in
Categories:
