Research Validity and Reliability
As we discussed, not all research requires undertaking an elaborate study. But even marketers conducting small, informal research should know that any type of research performed poorly will not yield relevant results. In fact, all research, no matter how well controlled, carries the potential to be wrong. There are many reasons why research may not yield good results (a full discussion being beyond the scope of this tutorial), however, most errors can be traced to problems with how data is gathered. In particular, many research mistakes occur due to problems associated with research validity and research reliability.
This problem with data gathering represents several concepts that to the non-researcher may be quite complex. But basically validity boils down to whether the research is really measuring what it claims to be measuring. For instance, if a marketer is purchasing a research report from a company claiming to measure how people prefer the marketer’s products over competitors’ products, the marketer should understand how the data was gathered to help determine if the research really captures the information the way the research company says it does.
While research validity is measured in several ways, those evaluating research results should keep asking this simple question: Is the research measuring what it is supposed to measure? If the marketer has doubts about the answer to this question then it is possible the results should also be questioned.
This problem relates to whether research results can be applied to a wider group than those who took part in a study. In other words, would similar results be obtained if another group containing different respondents or a different set of data points were used? For example, if 40 salespeople out of 2,000-person corporate sales force participate in a research study focusing on company policy, is the information obtained from these 40 people sufficient to conclude how the entire sales forces feels about company policies? What if the same study was done again with 40 different salespeople, would the responses be similar?
Reliability is chiefly concerned with making sure the method of data gathering leads to consistent results. For some types of research this can be measured by having different researchers follow the same methods to see if results can be duplicated. If results are similar then it is likely the method of data gathering is reliable. Assuring research can be replicated and can produce similar results is an important element of the scientific research method.