This is Part 2 of a 4-part series overview of market research, how to conduct and talk about research. It is designed for a general audience and not necessarily for trained or professional researchers. In Research 101, we talked about the basics of research and how it starts with a business decision. Today, we're talking about how important your sample is when doing research.
Census or Samples?
When we think about quantitative research methods, we're basically counting people. You ask yourself the question: who do we want to count, and can we count them all? If you can count them all, that's called a census.
Previously, we discussed that Field Agent uses store audits to count things. Let's say, you want to see the state of displays that are up in 200 retail stores. Then 200 store displays comprise the total population you are trying to count. You can go and count all 200 of those with a Field Agent audit. We're essentially trying to take a census of those displays, and we're trying to count them all.However, if you're talking about, for example, the shoppers in a retail store – switching back to the idea of counting people – there may be millions of people who shop in certain retail stores. There's definitely no practical, cost effective, or timely way you could ever count all the retail store shoppers and find out what they're doing in the store. A census just isn't feasible.
If you can't do a census because it's too expensive, it's too time consuming, or it's just not possible to do, then you take a sample of that population. And you count the people in your sample.
So if you cannot count all the people in your population, then how do you draw a sample that you can count instead? How do you know your sampling methods represent the people in your population? We want to draw a sample that is going to be a good indication of what the real, true population is like. If it's not a good sample, we get a bad measure.
In an ideal sample, every single person in the population would have an equal chance of being sampled. If you have everybody equally likely to be sampled, then that would be a truly representative sample.
Everybody would love to have the perfect sample in every single study, but it's not easy to do. It's expensive to try to include everybody in a population with an equal chance of being sampled. Sometimes there is no way to include everybody in your whole population when you draw a sample; or if there is, it could be very expensive or time consuming to do. If this is your only approach, then, as a business, you may find that you have to forgo doing a survey research study because you can’t afford it, or there might be no way to get the information from the study in time to inform your business decisions.
Frequently for business decision-making, we use what are called convenience samples. They're not strictly random. Everybody doesn't technically have an equal chance of being sampled. But they are large enough and they're recruited from a wide variety of sources so that we assume they're not biased in any particular direction.
A survey research panel is a group of people who have agreed to participate in studies. If a panel is recruited from a wide variety of sources, and if this panel is large enough, then the idea is that you can then dip out of that to get your study sample, and get a pretty good cross-section of the population without it being biased in any particular direction.
It's almost always quicker and cheaper to use a convenience sample than it is to use a strictly random sample. However, if you are going to publish findings in an academic journal, or make public policy based on survey research, or try to predict the election, you don't want to use a convenience sample. But for day-to-day business decision-making, convenience samples offer timely, cost-effective ways for businesses to understand the voice of the customer and gain greater insights into their markets.
Take Advantage of Crowdsourcing via Mobile
By harnessing the power of crowdsourcing and mobile technology, Field Agent makes it easy to use convenience samples to gather accurate data and insights. Conducting research and sampling is all possible with our Do-It-Yourself platform. Get started today.
Continuing Research Education