Collecting data samples in survey research isn’t always colored in black and white. Sometimes, members of your research population may be under-represented, which leads to what is known as undercoverage bias. 

Undercoverage bias is common in survey research as it often results from convenience sampling which a lot of researchers are guilty of. Like many other pitfalls in survey research and data collection, in general, undercoverage bias can hugely alter your survey results and affect the validity of your research. 

What is Undercoverage Bias? 

Undercoverage bias is a type of sampling bias that occurs when some parts of your research population are not adequately represented in your survey sample. In other words, undercoverage happens when a significant entity in your research population has an almost-zero probability of getting selected into the research sample. 

For example, let’s say you’re conducting a product evaluation survey via Formplus to find out what users think about a product. To accurately gather data for this research, you’ll need to collect feedback from both new and existing users of the product. If any of these groups are excluded or poorly represented in your data sample, then your survey will suffer from undercoverage bias. 

The whole idea behind conducting a survey is to arrive at findings that represent a clear approximation of what is obtainable in the research population. Undercoverage bias makes it difficult for you to achieve this in your research or systematic investigation.

Causes of Undercoverage Bias in Surveys 

There are many reasons undercoverage bias may be recorded in a survey. While some may result from the researcher’s error, others are beyond the direct control of the researcher in question. 

Understanding the causes of undercoverage bias brings you a step closer to mitigating this crisis. Let’s discuss some of these causes.

• Convenience Sampling

This is, perhaps, the most common cause of undercoverage in research. Convenience sampling occurs when the researcher only retrieves data samples from sources that are easily accessible, without trying to get data from members of the research population who are difficult to find. 

It is often described as a non-probability sampling technique because it depends on proximity and accessibility to gather data. Convenience sampling can cause undercoverage because most times, the research group that is easy to access may not be a fair representation of the research population. 

For instance, if you want to research food price inflation and only collect data from individuals in a city mall at a particular time of the day, you would have excluded a large chunk of the research population. Hence, the result of this research cannot be termed valid. 

• Poor Knowledge of the Research Population

Without having a clear knowledge of your research population, you are quite likely to exclude certain groups from your data sample. This is why we advise that you carefully map out your research population and target audience before setting out to administer your survey. 

In other words, you need to know who you would be gathering data from and also know where you can find them. For example, when researching the experiences of People of Colour in the U.S., your research population and data sample should have a fair representation of all People of Colour in the U.S.  

• Lack of Resources

Research surveys are sometimes capital-intensive一especially if you have to pay respondents to fill out your survey. Due to lack or inadequate resources, a researcher may be unable to adequately gather data samples from some parts of the research population and this leads to undercoverage bias. 

Let’s say your survey involves traveling to multiple countries to collect data samples from migrants. Without a deep purse; that is sponsorship, grants, or personal finances, you may be unable to fund these trips and this can force you to leave out the experiences of some parts of the research population. 

• Survey Design

Sometimes, your survey design may discourage research participants from providing data samples for your systematic investigation. Your survey should be relatable to the different groups in your research population and this should reflect in your use of language, survey PR channels, and survey design. 

If you want to gather data from an elderly population, asking them to complete a survey via email invitation may not be the best way to go. For this age demography, paper forms may work better and be more effective. The key is to create a survey that suits everyone to a large extent. 

• Time-Constraint

Timing is a huge deal when it comes to research and systematic investigation. In many cases, research processes are time-bound which means the researcher may be unable to gather all the data samples he or she needs from all the groups in the research population. 

Time-constraint forces the researcher to rely only on the data he or she can have access to with little or no hassles. To beat this, be sure to do enough work in the background and draw up a near-perfect schedule of how long it would take to execute different aspects of your systematic investigation; especially data sample collection. 

Examples of Undercoverage Bias 

In Elections

Consider a country with limited access to the internet deciding to adopt electronic voting as the means of casting ballots for preferred candidates in an election. This form of survey sampling will suffer from undercoverage affecting the following groups: 

1. Individuals who do not have access to the internet due to one reason or the other.
2. Individuals who are not computer-literate.

This method of convenience sampling excludes key groups in your research population, which can adversely affect the results of the election. 

In Statistics

To get the number of yellow Volkswagen in a country, a researcher decides to stand at one of the busiest roads in his neighborhood and count the yellow Volkswagen as they drive past. 

This method of sampling will lead to undercoverage bias as it does not account for the following:

1. Yellow Volkswagen that will be driven on other roads (major and minor) in the country.
2. Yellow Volkswagen that will not be driven at all (parked in home garages)
3. Yellow Volkswagen undergoing repair.
4. Yellow Volkswagen still in the sales shop.

In Research

A researcher wants to find out how many people in a specific country watch television. To get the data she needs for the study, she decides to go to a popular supermarket and collect data from shoppers. This is a form of convenience sampling and it is likely to suffer from undercoverage of the following groups:

1. People were absent from the supermarket during the research period.
2. People who do not watch television at all.

Disadvantages of Undercoverage Bias 

1. Undercoverage bias affects the validity of your research and alters your research outcomes. This is because your research results will not be a true representation of what is obtainable in the research context.
2. Undercoverage bias leads to increased variability which also affects the validity of your research findings. In survey research, variability is determined by the standard deviation of the research population so the larger your standard deviation, the less accurate your research findings will be.
3. Undercoverage bias can result in voluntary response bias. This means that you are at risk of collecting data from a small number of people who feel strongly about the research subject. The result is that your data samples will be skewed toward the opinions of a minute part of your research population.
4. It can lead to researcher bias in the systematic investigation.

Read Also – Response vs Non Response Bias in Surveys + [Examples]

How Formplus Can Help you avoid Undercoverage Bias in Surveys 

• Email Invitation

Formplus allows you to send out email invitations to respondents to gather data for your surveys and track responses. As we’ve mentioned earlier, the lack of accessibility of some groups in the research population can force the researcher to carry out convenience sampling which leads to undercoverage bias. 

Sending out email invitations to respondents makes it easy for you to access different members of your research population. All you need to do is properly map out the research population, collect their email addresses, and invite them to complete your survey via email. 

The Formplus email invitation feature also allows you to monitor responses. You would know when a respondent is yet to complete the survey and you can easily reach out to assist them with any challenges they may be facing with your survey; thus limiting your survey dropout rates. 

You can also prevent multiple submissions from a single respondent in your survey. Sending out email invitations means that you can track responses and prevent respondents from having further access to your Formplus survey once they successfully fill out and submit it. 

• Offline Form Capability

Imagine you have to collect survey responses via online forms from a part of your research population who have no internet access. Or a respondent is completing your survey and suddenly loses an internet connection? These are common scenarios in survey research that can lead to undercoverage bias. 

With Formplus, however, you can avoid these scenarios and gather data effectively from all the members of your research population; with or without internet access. Formplus allows for surveys and questionnaires to be completed offline; that is with poor or no internet connectivity. 

Form respondents can fill in data in remote locations without internet access and this data will be automatically synced with Formplus servers when the internet connection is restored. With Formplus, you can avoid undercoverage bias and collect data from anyone, anywhere, and anytime. 

• Conditional Logic

With Formplus, you can create custom surveys that appeal to different groups in your survey population. Many times, your survey design can lead to undercoverage bias; especially if it is structured in a way that leaves out some members of your research population. 

Conditional logic allows you to improve the validity of responses provided in your survey. It works by hiding or showing form fields and form pages based on the responses already provided by the person filling out your survey or questionnaire. 

For instance, let’s say you are surveying the experiences of the different races in America. Conditional logic allows you to present unique questions bothering on the unique experiences of each race to respondents in particular groups. 

So, when a respondent indicates that he or she is Hispanic, the conditional logic feature automatically shows fields with questions that are unique to Hispanic experiences while hiding fields that are irrelevant. This way, you can gather quality data from every group in your research sample and avoid undercoverage bias. 

• Mobile Forms

With Formplus mobile forms, your survey can be filled out on any internet-enabled device including mobile phones on the go. This helps you to deal with the survey accessibility problem that often leads to undercoverage bias in any systematic investigation. 

Formplus forms adapt to any internet-enabled device and they can be viewed and filled out conveniently without the respondent having to pinch out or zoom in on the form. 

• Multiple Form-Sharing Options

In a bid to solve the undercoverage bias problem resulting from the lack of accessibility to the survey, Formplus offers numerous form-sharing options so that you can easily share your online consent form with target respondents. 

For instance, you can directly embed the form on your organization’s website or add it to your social media pages with direct sharing buttons. You can also share your survey via a unique QR code that can be downloaded and printed out on banners and business cards so that respondents only need to scan to fill. 

Conclusion

In this article, we’ve discussed undercoverage bias and examined several practical examples of undercoverage bias in the systematic investigation. As we’ve shared, undercoverage bias results from convenience sampling, lack of knowledge of your target audience, as well as other factors that we’ve listed in this piece. 

Every researcher must be conscious of undercoverage bias and other research/sampling biases that can greatly alter the findings of your systematic investigation. This means that you must understand these biases, know how to identify them, and more importantly, leverage data collection tools like Formplus, to help you avoid them.