Data integrity is a core principle of marketing research. Just as a team of scientists performs research as a basis for making an objective go/no-go decision, a CRO or manufacturer undertakes a research study to understand markets as they really exist — the real facts, free of spin and contrivance — to use as a basis for making informed, thoughtful and strategic decisions.
However, just as in science, unless marketing research is carefully controlled, it can be subtly skewed toward confirming a bias — consciously or unconsciously. It’s a researcher’s job to understand how spin can happen and to eliminate bias and preconceived notions to reach the most objective truth possible.
High integrity study design
Integrity starts by asking why the research is needed and how the results will be used. A deep understanding of the project with respect to these two questions will help ensure the most appropriate research methodology is applied in order to minimize skewing of the final results.
We begin by determining an appropriate sample size to yield the most reliable data. For example, some life science companies have only a handful of customers while others may have tens of thousands. This must be considered when determining sample size. In addition, because some market sectors are highly technical or global while others are not, it is essential to design your research study carefully so that you extract the most pertinent information from relevant survey participants and achieve the objective of the study.
We also determine the best way to conduct the research to meet the study objectives within client constraints. For example, is it better to do a focus group, one-on-one interviews or online surveys? Each approach has its advantages and disadvantages depending upon the survey objective, population, timeline and cost.
It’s a researcher’s job to eliminate bias to reach the most objective truth possible.
The survey vehicle also plays a key role in eliminating bias. Whether for an interview or survey, questions need to be worded to not lead respondents. Multiple choice questions must be complete and offer a full range of options, and open-ended questions must be clear and unambiguous. What’s important is to cover the topic from enough different perspectives with each respondent to be assured you’re getting the data you need to make informed decisions.
Recruiting participants for an interview, focus group or survey is another key aspect of study design. Focus groups and one-on-one interviews are usually recruited by the research team, and a modest stipend for participation is negotiated up front.
For online surveys or in-person surveys at an industry trade show, it’s not uncommon to attract respondents by entering them in a drawing for a prize. It’s a proven way to build the dataset. If you offer a gift card or coupon as an incentive, keep in mind to not make it so generous that people take the survey only to win the reward. Accurate results are more often obtained if the participants have an inherent interest, without an incentive. Sharing the study results may also incent individuals to participate.
Even random, anonymous respondents should be knowledgeable about the questions at hand and be able to demonstrate their qualification for participation by background or employment. For example, you can’t research a global question if all the respondents come from one country. A study founded on a skewed dataset will not reveal the objective truth we’re looking for.
Fight the bias
Finally, integrity has to extend into the analysis and reporting. These reports are often filled with facts, but far too often the true basis on which to formulate a decision is omitted — consciously or unconsciously — from this communication. It’s expected for researchers to eliminate inappropriate responses to keep them from skewing the real truth, but it is totally unacceptable to eliminate responses because they’re negative or contradict a hoped-for outcome. The purpose of market research and market researchers is to provide counsel based on the data — and I’d be remiss without adding — within the margin of statistical error. Your market researchers should be able to tell you what your margin is.
Furthermore, to ensure data integrity, you must understand that your own biases are often difficult to recognize and that there is a danger in, again, consciously or unconsciously, corroborating preconceived notions. This is why it’s especially important to rely on research conducted by an objective third party.
At the end of the day, the point of market research is to provide context and support for a conclusion to, say, make more informed decisions about the launch a new product or service, enter a new market, or evaluate reputation or brand perceptions. Third-party research offers a valuable opportunity for companies to get unbiased data and direction that can help refine and improve your marketing strategies. Only by addressing bias and data integrity at every stage of the research can you achieve the data accuracy you need to be confident in the direction of your marketing initiatives.