The author reaches the verdict that a good research must have ‘a well defined question’. In this case, the paper is on a quest to define a good research and so distinguish it from bad research in the process. A good definition of the research question is a great means to this end. It helps the researcher to stay on course and only cover relevant areas. A badly structured research questions on the other hand gives a free hand to the researcher to cover a topic unrestrained (Palmquist, 2001).
Sampling cannot be trifled either. This is where the hypothesis is tested and the research question is put to a few people who are a fraction of the whole population. It is important therefore to get this right because samples carry the views of many people. Make the samples as representative as possible, but above all, do not try to force a consensus (ASPA, 2006).
Statistics used are of paramount importance too. Statistical methods used must be compatible with the kind of research at hand. Using the wrong statistical methods can lead to the wrong finding. Statistics are a distillation of the whole research too and so must be as simplified as possible (Garson, 2009).
Then, there is the issue of the findings. These must be interpreted as they are and not tweaked in any way to conform to ideas the researcher had at the start of the project. This is the rationale behind doing the research in the first place, understanding the ideas on the ground better. Even a few dissenting views must be taken on board and not merely given a mention. These views help to create a balanced research. This is because logically, there is unlikely to be a consensus on every topic of research (ASPA, 2006).