How To Determine Sample Size

How To Determine Sample Size

This complimentary excerpt from the new Principles Express courseSampling in Market Research, authored by Keith Phillips, teaches you how to determine sample size in a valid way that will determine representativeness.

Determining our Sample Size

One of the most common questions you will hear in the Market Research industry is “How large does my sample need to be in order to be representative?”  However, it is not the sample size that determines representativeness. It is the other elements discussed in our module “Sampling in Market Research”, such as defining the target population correctly and selecting the sample in a valid way that will determine representativeness.  For example, we could have a very large sample size, but if the sample is drawn in a way that introduces bias, even a largesample may not be representative.

The sample size is very often used interchangeably with the terms base size and the number of completed interviews.  Essentially, the sample size is the sample you end up with, that you are using to project to the larger population. So, for instance, you may see in the media that a survey was conducted among a sample of 1,000 people.  This is the sample size or base size. The actual sample that we began with for that study was likely much larger in order to produce 1,000 completed surveys. In today’s sampling world, the base size or sample size we end up with will almost always be smaller than the original sample.  This has to do with response rates (% of those responding to an invitation) and participation rates (% of those responding and willing to complete the survey), or, in other words, getting those who were chosen as part of the sample to complete the survey.

The sample size is determined by how much variance we are willing to accept in our data.  From our sample, we produce an answer that is unlikely to be the exact truth (a result that is precisely true of the entire population), but we expect that the results from our sample produce an answer that is close to the truth.  There are several factors that determine the likelihood of our survey results being close to the truth, the most important of which is the sample size.

So how do we determine what our sample size should be…

The size of the sample for our study will determine how much variance we can expect to see in our data and ultimately how confident we are in our result. These factors will influence how much sample we will need for our study, but they are not the only factors. The larger the sample size, the higher the cost and the greater the length of time needed to collect the data. We need to come up with the optimum plan, to both achieve a base size we are comfortable with and that is worth the time and money spent gathering it.

Consider a piece of research that is going to be used to decide which TV commercials to air.  If these TV ads were tested and revealed a 2% difference in preference among the core purchasers, a marketer may conclude based on prior experience with similar advertising testing that both ads scored similarly and that using either one would have a similar effect.  We would not need to sample 2,000 people for each ad, to learn that the 2% difference is real. It is not necessary. However, a 2-percentage point difference could be significant in a local political race, where having a large enough sample size of expected voters is important for advisors of the campaign.  

Now imagine one TV ad scored 10 percentage points higher than another, but our base size was so low, we could not conclude which ad was better.  A 10 percentage point difference in preference could be meaningful if it was real. A marketer would want a large enough base size to conclude which commercial is better when there is a 10 percentage point difference in preference.  To decide on a base size, we need to first understand what results will be significant to the decision and then choose a base size that makes sure that the result would be statistically significant.

We must weigh the practicality of finding the individuals in our target population.  If we are dealing with a population that is very small, then we cannot make our sample size too large.

We must also consider the type of analysis we will be doing with the data.  If we will be segmenting the data into smaller portions, such as younger female concert goers among a general population survey, our base size for the female concert goer may be small.  We must consider what this sample size will be when deciding on our overall sample. 

The information described above is just some of what is taught in our affordable online Principles Express course, Sampling in Market Research. Interested in learning more?  Principles Express courses are totally self paced, online and easy to complete in 12 hours.  For only $359, you’ll learn how to best collect quantitative research data on a global scale, and receive a digital badge from the University of Georgia for completing the course.  Click here or call +1 706 542 3537 to learn more!

We are grateful for the course to be sponsored by Full Circle, a leading global provider of seamless, productive online survey experiences. Such sponsorships have funded the development of our new line of Principles Express courses, a portfolio of $359 online courses that let you master a research skill, at your own pace, with just 9 to 14 hours of study.

Leave a Reply

Your email address will not be published. Required fields are marked *