Unlike quantitative research projects, there are few formal statistical guidelines governing the design and execution of a qualitative market research study. Best practices regarding randomness, confidence intervals, margins of error, and estimations for sample size needed are not standardized for qualitative research. Ultimately, these decisions are arrived at either by default or are simply chosen more based on intuition and past experience and not based on any known formula.
This subjective nature has led to many different opinions in the realm of correct sample size for qualitative project design. Interestingly, a common practice appears frequently in the industry and that is a sample size of 30. But how did that become the most prevalently used number and how reliable is that as a default? When looking to estimate the right sample size for your next project, consider these two different approaches. As you evaluate your needs, understanding how and why a sample size of 30 is respected as an optimal number in qualitative research can provide increased confidence as new projects are structured.
The Limitation Approach
One method often used to estimate the needed number of qualitative interviews, online bulletin boards, or focus groups is to determine a feasible number given the limitations of the project (e.g., budget, timing, audience availability, etc.) and the intended scope.
In the context of academic research, professors Patricia A. Alder and Peter Alder note how these limitations manifest themselves differently based on the academic level. For instance, they advise undergraduate students not to bite off more than they can chew, relegating themselves to around a dozen interviews. More generally, they say collectively that their best estimate “is to advise in the broad range of between a dozen and 60, with 30 being the mean.”
Outside of academia, time limitations and budgets can govern this number. Considering the labor intensive work of conducting over 100 interviews requires a protracted timeline and significant budget. In these cases, examining the needs of variance of opinion is the best starting point. For example, when studying the differing needs of car buying consumers, meaningful segmentation of the population group can be created. If five stacked ranges of income are a factor being studied, consideration should be given to evenly recruit participants from each level. Final sample number, regardless of how large, should be representative so project limitations do not inadvertently skew results by not representing a significant portion of the intended audience.
In the context of discovery work, qualitative researchers may not feel see sample size as a significant requirement for standardization. This work is typically more iterative and often follows new findings as they emerge. While the amount of empirical data may have a required critical size, the size of the sample may be irrelevant. The open-ended methodologies used in this type of research is measured by empirical saturation instead.
The Mathematical Approach
In full consideration of project limitations, Peter DePaulo offers an alternative viewpoint regarding the determination of quantitative sample size. His predominant theory is that, within a subject with a limited amount of potential viewpoints, one can hear about 30 interviews before the “saturation point” of discovery is reached.
“Our qualitative sample must be big enough to assure that we are likely to hear most or all of the perceptions that might be important,” DePaulo states. “Within a target market, different customers may have diverse perceptions. Therefore, the smaller the sample size, the narrower the range of perceptions we may hear.”
But DePaulo takes that limitation approach a step further in order to translate these concepts for greater clarity and specificity. Interestingly, he arrives at the same conclusion in terms of sample size, but based on mathematical calculations of estimating the incidence of an opinion as a percentage of the population.
For example:
If 10 percent of the minivan-owning population purchases vehicles for transporting business goods, this minority viewpoint will not be heard if the sample size is too small. In order for a research study to capture this minority viewpoint within a population with a less than five percent chance of missing it, DePaulo’s best estimate for the sample size is 30 van owners.
Finding the Right Fit
These two differing approaches shed light on how to estimate an appropriate sample size for a qualitative study. An experienced research partner can review project details and consider a meaningful sample size with more confidence and offer guidance on the total number of recruits to meet project requirements.
Regardless of the quantity, insights professionals also consider quality benchmarks to achieve reliable and representative qualitative results once the number and the required strata are determined. Research objectives can be included in screener questionnaires for greater clarification and precision. A proper recruitment partner can confidently and transparently confirm participant qualifications to ensure quality at any sample size.
Ready to set the stage for a successful project?
Let the expert team at Qlarity Access provide a detailed bid including target audience demographics, sample size, timeline, expected incidence rate, methodology, length of interview, and field time. Our research management plans address potential attrition and show-rate considerations to ensure an optimal final sample size for the program. A well-planned kick-off call and a proactive approach to project management is the foundation of communication to meet the project needs.
The success or failure of every market research engagement rests largely on the quality of your data collection strategies and field services partner. Download our FREE guide 6 Keys to Accelerate Growth with the Right Data Collection Partner to find the best fit for your next project.
