Quota sampling is a very common method of sampling particularly in online research, where online research panels or internal customer databases are utilised and is often used instead of randomised sampling methods that can be costly and time consuming to administer.
Table of contents:
What is quota sampling?
Quota sampling is a non-randomised sampling method where judgement is used to control quotas based on participants traits or characteristics such as gender or age that are representative of the population that is being researched like a 50/50 split of males and females.
Basically, this is broken down into two stages, where the first stage is developing the controlled quotas or categories of elements of the target population such as socio-economic groups or regions, while the second stage is based on judgement or convenience in selecting the sampling elements that these participants belong to.
This is the point of quota sampling that the research needs to be reflective of the target population, otherwise the research is meaningless and plus, further screening during the research of the target population makes it more accurate.
For example, a survey about how car ownership differs across the UK needs to be nationally representative, based on regions like the north, south, London, Midlands, Wales and Scotland and setting different sample proportions to reflect the population size of each region. Obviously, you cannot interview everyone but setting the actual percentage of each of the main elements of the target population within the total sample size makes it reflective, achievable and within budget. This is providing the total sample size is large enough for robust analysis to take place. More on this in the next section.
Note there are two types of quota sampling to be aware of. The first one is controlled quota sampling that limits the sample that can be used based on the tight quota restrictions, while uncontrolled quota sampling allows for the researcher to select all sample, they deem fit based on their knowledge of the target audience without any restriction.
How to do quota sampling in 4 easy steps
There are no formal rules that need to be followed with quota sampling like you do with random sampling methods. The following are 4 steps that should be taken in order to do quota sampling the right way:
Step 1: Divide the sample into sub groups
The very first step should be to divide the total population into subgroups, which ought to be mutually exclusive by their traits or characteristics such as splitting the audience of car buyers by the brand of car they mainly use, whether that is Ford, Nissan, Toyota, Mercedes, BMW and so on. Another example is dividing total population of football fans by what team they support in the top division.
Step 2: Evaluate the proportions of each of the subgroups
After the sample population has been divided into subgroups, you need to learn the key proportions of each of the subgroups, which need to be kept within the applied percentages. For example, applying percentages of the different age groups of a country’s adult population to be nationally representative for a study about personal finance. An alternative example is you find that 60% of the female population do the main household shopping, so you would apply the relevant 60% versus 40% sample split for gender.
Step 3: Choosing the correct sample size
The sample size of each of the subgroups should be reflective of the target population. For example, the total sample size of 400 respondents to be representative of the total population of 40,000 users of a certain credit card brand. There are sample calculators available like this one that you can use to ensure you calculate the correct sample size.
Step 4: Select the participants based on the set quotas
The final step is going back to when you evaluated the percentage proportions for each of the subgroups that are representative of the target audience at step 2. This now needs to be applied to the sample, where the relevant participants are selected to participate in the research.
So going back to the previous example about users of a credit card brand, there are 3 types of credit cards that the credit card brand offers. Credit Card A represents 40% of the brand’s credit card users, while Credit Card B and C are equally split with a usage of 30% each. Therefore, 40% of Credit Card A needs to be applied to the total sample size of 400 and the remaining 60% is equally split between Credit Card B and C users.
Quota sampling example
A typical standard example of quota sampling is a survey of buyers of hairdryers across the three Eastern American states of New York, New Jersey and Philadelphia to see which brands they prefer to buy with a total sample size of 900 respondents. The following is an example of how the total sample could be split amongst the basic demographic quotas below:
Gender – 300 males and 600 females
Age – 18-44 – 48% of population - 433 completes
45-64 – 35% of population - 313 completes
65+ - 17% of population - 154 completes
Region – New York – 300 completes
New Jersey – 300 completes
Philadelphia – 300 completes
Socio demographic groups – ABC1 – 450 completes
C2DE – 450 completes
The quotas can change depending on the attributes and characteristics of the target population and the judgement of the researcher, so they may or may not allocate some quotas equally. Plus, further screening can be carried out early on in a survey to go further, which will improve the accuracy of the research.
Advantages of quota sampling
The following are 4 key advantages of quota sampling:
1. Ideal when restricted by time
Quota sampling allows you to save time especially if you are in situations when time is limited and you need to quickly conduct some research to help answer a question. This is a key advantage that quota sampling has over random sampling methods that can be time consuming to manage.
2. It reduces the cost of carrying out the research
Quota sampling will reduce the cost of conducting the research as it corresponds with the time reduction advantage mentioned above as you save time in gathering all the data and information required from a representative audience, which is particularly effective for quantitative methods (surveys) as well as focus groups or depth interviews.
3. Greater convenience
As quota sampling will allow you to select elements of your target audience easily, particularly as these elements will tend to be demographics consisting of attributes like gender, age, location, income groups, that you are able to select to be reflective of your target audience. This will help to make comparisons and analyse different groups.
4. Used in the absence of a sampling frame
Quota sampling can be used when there is no sampling frame available in which the data or information can be drawn from whether that be a subset of particular types of people or businesses. So, basically this could a database of subscribers or customers of their buying habits (products purchased, quantity bought, current or lapsed customers), which is not available for you to tap into and you would therefore use quota sampling to get around this issue and screen respondents during the early part of the research.
Disadvantages of quota sampling
Below are 3 main disadvantages of quota sampling:
1. Potential for selection bias
There is potential for selection bias of participants from interviewers, who may reject potential respondents from taking part as they appear to be difficult or live in undesirable locations and likely to go to areas where participants are most likely to be found. Plus, initial non responders may not be followed up as there may be a bias that these respondents are not accessible for reasons like work commitments.
2. Unable to assess the sampling error
Like random sampling, quota sampling is subject to sampling error but you cannot calculate what it actually is especially since the sample is not randomised and is based on judgement of the researcher in selecting the quotas. So basically, as you cannot calculate the sampling error, you are unable to measure the true representativeness of the sample.
3. Potential issue of insufficient numbers for a subgroup
A practical issue of quota sampling is certain subgroups of the sample selected may not be enough to draw statically valid conclusions from the analysis to make comparisons due to the fact the sample size for that subgroup is too small such as very high-income earners. The way this is normally resolved is over sampling the subgroup such as double the sample size and applying weighting to the data collected.
Application of quota sampling
There are many occasions where quota sampling is utilised and is common place when conducting surveys with the use of survey panels. Below are just some instances where quota sampling is used.
When there are time constraints
This type of sampling is applied when there are time constraints involved, so it will allow the researcher to get a relevant set of data of the target audience based on the quotas that have been set.
Research is on a tight budget
When there is a limited budget, setting quotas of the population of interest is a great option to have to get a general and relevant outlook of the audience with regards to the research objectives that have been agreed.
Occasions where specific criteria need to be followed
There can be situations where you need to follow set criteria of characteristics or attributes for different subgroups of the research project, which quota sampling will allow you to do so with ease.
To be able to conduct comparative analysis
Quota sampling will allow you to compare more than one subgroup by determining distinctive qualities that are among the linked traits and characteristics of different subgroups.
RELATED HOW TO GUIDE POSTS BELOW FOR SURVEYS AND MARKET RESEARCH
