# Data Sampling Techniques Used in Market Research

Page content

Market research is productive only when the organization has a clear idea about the respondents to be included in the survey. Surveying the entire population may not be possible due to the large scale or due to related costs involved in conducting these surveys. Thus, market research deploys sampling techniques to identify the prospective respondents for the research from the total population.

There are a lot of different sampling techniques available. These are divided into two main categories based on their approach – probability sampling and non-probability sampling.

## Probability Sampling Techniques

All sampling techniques that fall under this category follow a mathematical procedure to determine the respective respondents. The added efforts required for using these techniques get rewarded in the form of much more reliable and accurate data than data obtained through non-probability sampling.

Here are some of the common probability techniques that are widely accepted and used by most modern organizations.

Simple Random Sampling

As the name indicates, this technique for sampling involves selecting the respondents, at random, from a list of the population. For instance – if the population constitutes all the students of a college, the respondents can be selected at random from the list of roll numbers available with the college office.

Stratified Random Sampling

This technique requires that the population is divided into two or more subgroups – based on a certain variable. The main motive of using this technique is to ensure that respondents from all exclusive subgroups are included at the time of the survey. These subgroups are often referred to as ‘strata’ and a main feature of these subgroups is that each of them is unique and non-overlapping. For instance, the population may be divided into subgroups based on the gender and an equal number of male as well as female respondents are included in the research.

Cluster Sampling

This technique involves dividing the population into clusters and selecting only a few clusters from the population to participate in the survey. The respondents are selected from within these selected clusters. There are two prime reasons why this technique is used – there is not much information available about the population or the cost and feasibility of using any other sampling technique poses a problem owing to the geographic distance that separates the respondents in the population.

Systematic Random Sampling

This is a slightly more methodical variant of the simple random techniques and involves some specific techniques to identify the respondents from the population – like every fifth person on the population list. For instance – if the population constitutes a list of 100 names, the respondents may be selected as every fifth person on the list.

## Non-Probability Sampling Techniques

These sampling techniques are more convenient than the techniques discussed on the first page. However, one must note that the data thus obtained is not as reliable or as accurate as the data obtained through probability sampling. In most cases, these techniques are put to use in large scale surveys where it is difficult to use any of the probability sampling techniques, due to unavailability of accurate information about the population. Also, it is important to note here that with these non-probability techniques for sampling, it is quite likely that some elements of the population may not get selected as respondents.

There are again four commonly used non-probability techniques, which are discusses below.

Availability Sampling

As simple as it sounds, the criteria for selecting the respondents is entirely based on who are readily available or are easy to contact. This technique is so convenient that at times, this is referred to as a convenience technique. Since, this technique involves surveying only those respondents who are easily available; it makes the survey fairly easy and economical to conduct. However, the convenience seriously jeopardizes the reliability of the data obtained.

Quota Sampling

This is an improvised version of the above technique, which involves fixing up quotas so that the respondents bear some specific characteristics – based on their ratio of existence within the population. For instance – if the population consists of 70% college students and 30% young executives, then the respondents should be picked up in the same ratio from the population.

Purposive or Dimensional Sampling

A further extension of the quota sampling is the purposive technique, which is also commonly referred to as dimensional sampling. It takes into account more than one characteristic and requires that at least some specific percentage of the respondents represent all of these characteristics. For instance – if a survey sets up a 30% quota for respondents who are male, in the age group of 20-25 and unemployed, then if 100 respondents are to be selected for a survey at least 30% of them must hold true to these three characteristics.

Snowball Sampling

This technique works in a very unique way and the sampling begins with the selection of one respondent from the population, based on certain characteristics. This person is then asked to identify the respondents that should be included in the survey. The procedure may be repeated until the desired numbers of respondents have been surveyed.

The researcher has a wide range of sampling techniques to choose from, and the best way to decide on which technique should be used is to keep the desired outcome of the survey in mind.

## References:

Image Credits:

Image by: Sidharth Thakur

MorgueFile/xololounge