Innovation approaches and analytical thinking are key in maintaining research integrity.
Many of us are grappling with the question of what is happening with research in these uncertain times. As we know, research is an illumination of reality, but can never fully reflect the absolute truth. Considering this, we need to find the best possible ways to help us shine a light in the dark. In our current times, many of our tried-and-tested options may not be possible or viable. It is all about adaptability.
Interestingly, prior to 2020, researchers had already started looking for new and more agile ways of conducting research, and started implementing these where appropriate. The new realities we have had to face have created an opportunity for us to explore alternatives and be nudged out of our comfort zones. This has resulted in many creative and innovative ways of conducting research, with technology opening a whole new world to researchers around the globe.
We now find ourselves in a space that can and will enable us to achieve most of our research goals, but to do so we need to rethink our approach to research and balance what is needed with what is possible. One small prediction I will make is that in the future we will not be conducting research in the same way we have, as alternative methodologies will become more prevalent and comfortable.
This will enable us to do more agile research with quicker turnaround times and possibly lesser costs. The vital consideration, though, is that the quality of data must never be compromised for expedience or convenience, particularly in the world of media currency research.
Face value
We have many options we can consider apart from conducting face-to-face research. These include telephonic interviews (CATI), online panels, instant voice recognition (IVR), emailing, video interviewing, access to various client databases and so on. Each of these methodologies has advantages that can be effectively harnessed to meet specific research objectives. We as researchers now need to assess these and make decisions on the optimal approach for any project at hand, rather than adopt a ‘one-size-fits-all’ mindset.
Let’s discuss face-to-face surveys in a little more detail. These have traditionally been the gold standard for many research endeavours. They are seen in this light particularly for media currency research, where randomness and representation are core. Without the strict rigour that is required to conduct currency research [1] we cannot go into field half cocked; we need to be sure that the methodology selected provides what is needed for the end user.
Face-to-face, in-home interviewing has always been considered the best way of achieving this, and it still would be, if things had not changed in the world. But at this stage of the game, it is not an option in many markets and for many clients, so viable alternatives must be found.
The reality is that for some time we have been facing challenges conducting face-to-face surveys in our unique South African environment. It has become increasingly difficult to achieve randomness and representation, particularly in the more upper-end areas and those that have a high prevalence of complex living. There is also a general aversion to letting interviewers into homes.
Ways have been found to get around this – for example, technology has allowed us to use CATI effectively by being able to select mobile numbers randomly, which then goes a long way towards ensuring representation. Some research houses are conducting hybrid methodologies, opening alternative ways to collect and analyse data. It’s a case of ‘horses for courses’.
The most exciting thing we are seeing is the implementation of fusion techniques, enabling us to combine multiple datasets to get richer and more extensive information. Such a volume of data would be exceedingly difficult to achieve with one single dataset, and the possibilities of this are endless.
No substitute for analytical thinking
The important point is that selection of a particular methodology should be based on a full understanding of what the chosen methodology offers and what skews or biases are inherent. These exist in all research methodologies, and such limitations do not imply that the data is not valid or credible.
Rather, the critical issue is to understand the skews and their impact on the results. It is imperative that we understand how the sampling and selection of respondents is derived, and how the weighting is conducted. We must now think more analytically.
We can still learn from the ‘old school’ way of approaching research. That is, we must still apply the 5W plan when approaching a piece of research, either as a researcher or user of the data. The what, where, why, who and when of the dataset must still be considered. If we can fully understand all of these and know exactly what we are dealing with, then we can feel comfortable with the data.
Ironically, now that research methodologies may seem easier and more agile, understanding the data and where it comes from will require more PT and discerning thought. We must never fall into the trap of accepting that ‘data is data’ without fully understanding what it is about. This is what leads to fake news and incorrect decision making, which we can never afford in the media industry.
Going forward, we can avail ourselves of all the new approaches that technology enables, but we must stick to the classic methodological pointers. We cannot keep benchmarking data on historical face-to-face results as if it was the definitive truth. We must embrace the fact that each methodology will give us a slightly different-looking lens on the same subject matter.
With 30 years in the research industry, Lauren Shapiro is director of Kuper Research. She is extensively involved in the technical oversight and scrutiny of the audience research ROAD with the Out of Home Measurement Council. She is also a contracted consultant to Nielsen involved with TAMS. Shapiro is one of the independent researchers responsible for the Futurefact survey that monitors the minds and moods of South Africans.
