“Research is definitely not the most researched subject by researchers” said Charles Coolidge Parlin, the father of market research. Well, rightly so. The enormity of the field along with the diversity of methodologies available in the subject based on the slightest nuances of the research objective often leaves the researcher overwhelmed. We here have hence put together for you a ready reckoner of the don’ts of research. No matter how simple or complicated the research, steer clear of the following commonly made mistakes:
- Vague Objective Assignment to the Research : Know what you want to find, how you want to find it and why. Not knowing what you are trying to analyze will only leave you with a mammoth pool of data and no conclusion that would fully resolve your objective.
- Confusing Correlation and Causation : Variables that are connected are correlated. That however does not mean that one caused the other. Assuming causation can cause a dismissal of a third or more variables or in some cases even reverse causation. The conclusion of the thesis in turn could be the exact opposite of the reality.
- Omitting or Minimalizing the Statement of the Error in the Measurement Technique : A single statement saying that the examiners were calibrated does not specify the kind of error that one may expect in the research result. The methods of training, the standards used, the frequency of re – assessment, the degree of calibration are just some of the details that would impact the research outcome and hence should be mentioned.
- Faulty or Lack of Definition of the Sample Size : An inadequate sample size can completely subvert the result of a research. For any research to give a reliable conclusion with an accepted power of the study (>80%) the definition of the sample and the size of the sample form the key criteria. A research for smart phone users done on any individual owning a mobile phone would have no meaning while a research expected to be extrapolated to a universe of a million will be of no value if executed on a sample size of a hundred.
- Inadequate or Lack of Bias Control Measurement Techniques : Bias in collecting data will ensure that the result of the research is in line, completely or partly, with the opinion of the data collection team. A credible control system at the onset of the research and through the process brings the research closer to reality. Investigators being kept blind to the subject conditions, randomization, intervention, re – assessment through random investigators, check on the subject or the investigators personal ambitions, expectation fulfillment control are just some of the methods that can be utilized in keeping a healthy Bias Control System in place.
- Failure in Highlighting Weaknesses : Since research is an aggregation of data collected on the basis of numerous definitions and in varying conditions, some subjective and some objective, the result of a research always has inherent in its very fabric several weaknesses and shortcomings. A complete gloss over of the same would suggest the result to be idealistic which is in itself inaccurate. On the other hand, the identification of these weaknesses can subvert the possibility of any false conclusions from the research thesis.