Can Requiring Internet Access Exclude Members of the Target Research Population?

Who Cannot Participate When Your Research Requires an Internet Connection?

Online surveys, virtual interviews, digital consent forms, and web-based research platforms can make data collection faster and less expensive. They can also create a participation requirement that researchers may overlook: participants must have sufficient internet access to engage with the study.

Consider a researcher investigating barriers to online education who distributes a questionnaire exclusively through an institutional learning platform. Students with reliable connectivity may respond easily, while those experiencing serious access difficulties may never encounter the invitation.

The resulting problem extends beyond inconvenience. If internet access influences both the opportunity to participate and the phenomenon being investigated, the study may systematically miss people whose experiences are particularly relevant to its conclusions.

Internet Requirements Can Exclude Eligible Participants and Distort Population Estimates

How Internet Requirements Affect Research Participation and Validity

What Counts as an Internet Access Requirement in Research?

An internet access requirement exists whenever participation depends on connecting to an online service or transmitting information through the internet. The requirement may be explicit, such as an eligibility criterion requiring home broadband, or implicit, such as distributing invitations only through email.

Internet dependency can occur at several stages: learning about the study, completing eligibility screening, reviewing consent information, attending interviews, submitting measurements, or participating in follow-up assessments.

A participant may be able to complete one stage but not another. For example, someone might receive a short text message containing a recruitment invitation yet be unable to open a lengthy web questionnaire.

Researchers should therefore examine the entire participation process rather than treating internet access as a single yes-or-no characteristic.

Why Is Internet Access More Complicated Than Being Online or Offline?

Internet access varies in reliability, affordability, available equipment, privacy, and technical capability. A person who occasionally accesses the internet through a shared smartphone does not necessarily have the same participation opportunities as someone with reliable home broadband and a personal computer.

The Pew Research Center's 2025 survey of United States adults illustrates why these distinctions matter. Home broadband subscription was reported by 54% of adults in households earning less than $30,000 annually, compared with 94% among those in households earning at least $100,000. These estimates describe a particular national context and should not be generalized to other countries or populations without supporting evidence.

The underlying methodological issue is broader: having some internet access does not establish that a participant can reliably complete every digital procedure required by a study.

These barriers can overlap. A participant may have a smartphone but insufficient data, or adequate connectivity but no private environment for an interview.

Which Population Groups Might Be Disproportionately Affected?

Potentially affected groups include people with limited financial resources, residents of poorly connected areas, individuals who depend on shared devices, and participants whose accessibility needs are not supported by the research platform.

Age, educational attainment, employment circumstances, and household arrangements may also be associated with differences in digital access. However, researchers should not assume that every older person lacks digital skills or that everyone in an urban area has reliable connectivity.

Population-specific evidence is preferable to demographic stereotypes. A study conducted among university students, rural households, healthcare patients, or working adults may encounter different access patterns.

The broader question of which groups may be systematically excluded by a research design should be addressed before selecting recruitment and data collection procedures.

How Does Online-Only Recruitment Create Coverage Error?

Coverage error arises when the mechanism used to identify or contact potential participants does not adequately cover the target population.

Suppose the target population includes all adults receiving a public service, but recruitment occurs only through an online client portal. People who do not use the portal have no opportunity to receive the invitation through that channel.

Even if every portal user has an equal probability of selection, the sampling process cannot select people outside the portal-based frame.

It is useful to distinguish coverage error from nonresponse. Coverage concerns whether members of the target population have an opportunity to be selected or contacted. Nonresponse concerns eligible people who are reached or selected but do not provide the requested data.

An online-only study can experience all three problems. A participant may be unreachable through digital recruitment, may decline after receiving an invitation, or may misunderstand a question displayed poorly on a mobile device.

When Does Internet-Related Exclusion Produce Selection Bias?

Internet-related exclusion does not automatically make every estimate biased. The consequences depend on how internet access relates to participation and the variables being investigated.

For descriptive research, bias may arise when the outcome differs between people who can participate online and those who cannot. For example, an online-only survey may overestimate satisfaction with digital public services if people who experience the greatest access barriers are missing.

For analytical research, selection can distort estimated associations under particular causal structures. The direction and magnitude of the distortion cannot be determined from the existence of an internet requirement alone.

For qualitative research, online recruitment may exclude experiences necessary to understand a phenomenon, even when statistical population estimation is not the objective.

Researchers should identify the intended inference and examine whether internet-related participation mechanisms threaten that inference.

When Is Requiring Internet Access Scientifically Justified?

Internet access may be integral to the research question. A study investigating advanced features of a web-based platform among its active users may appropriately recruit people who can use that platform.

However, the population boundary must be explicit. Findings about active users should not automatically be extended to all people who might benefit from the service, including those who cannot access it.

A useful distinction is whether internet access is part of the phenomenon being studied or merely a requirement imposed by the researcher's preferred data collection method.

For example, evaluating the experience of current users of a specific online application differs from investigating barriers to accessing that application among all intended users.

In the second case, excluding nonusers because they lack internet access may remove the very population whose experiences could explain limited adoption.

Can Mixed-Mode Data Collection Reduce Digital Exclusion?

Mixed-mode data collection allows participants to provide information through more than one method, such as web questionnaires, telephone interviews, paper forms, or face-to-face administration.

The American Association for Public Opinion Research identifies mixed-mode approaches as an option for reaching people who are less accessible through a single survey mode. The appropriate combination depends on the sampling frame, budget, research objectives, and measurement requirements.

Offering several modes does not automatically eliminate selection bias. Some participants may still be unreachable, and responses may vary according to administration mode.

Where multiple modes are used, researchers should preserve equivalent question meaning, response options, and administration procedures as far as feasible. Pretesting can help identify differences that may affect interpretation.

Should Researchers Provide Devices, Data Allowances, or Technical Support?

Providing a device, connectivity support, or technical assistance may reduce barriers when digital participation is necessary. The appropriate intervention depends on which barrier prevents participation.

A data allowance may help someone who already has a compatible device but cannot afford sufficient connectivity. A loaned tablet may address device availability without solving unreliable network coverage. Technical assistance may improve usability but cannot compensate for an inaccessible platform in every circumstance.

Researchers should also consider whether equipment provision changes the phenomenon under investigation. If the study aims to estimate access to digital learning under ordinary conditions, supplying internet access before measuring that access may alter what is being observed.

Support arrangements should therefore distinguish assistance needed to participate in research from interventions that could change the outcome itself.

How Should Digital Accessibility and Privacy Be Addressed?

Internet connectivity is only one dimension of digital accessibility. Researchers should examine whether online forms can be navigated using assistive technologies, whether text is readable on small screens, and whether audiovisual materials have appropriate alternatives.

The Web Content Accessibility Guidelines developed by the World Wide Web Consortium provide an established reference for assessing web accessibility. Their application to a particular research platform requires consideration of the relevant accessibility requirements and user needs.

Researchers should also consider whether participants can complete sensitive activities privately. Shared devices, public connections, or assisted participation may introduce confidentiality concerns.

Where appropriate, accessible materials and alternative participation formats should be incorporated during protocol development rather than added only after participants encounter difficulties.

How Can Researchers Evaluate Internet-Related Exclusion Before Data Collection?

Begin by identifying the actual technical requirements of each research activity. Then examine available evidence about connectivity, device ownership, digital skills, and accessibility within the intended population.

Useful evidence may come from population surveys, administrative information, feasibility studies, or consultation with potential participants. The quality and applicability of these sources should be assessed rather than assumed.

A pilot study can test whether the proposed platform works under realistic participation conditions. Researchers should include people with different devices and connectivity circumstances where relevant.

During recruitment, monitor completion patterns, technical difficulties, and reasons for nonparticipation when such information can be collected appropriately. If certain groups remain underrepresented, consider whether additional recruitment or participation modes are warranted.

The central objective is not merely to increase the number of responses. It is to understand who can contribute evidence and whether the achieved sample supports the intended conclusions.

How an Online-Only Survey Can Misrepresent Digital Learning Barriers

In this constructed scenario, the online-only estimate is 10 percentage points higher than the population value. The difference results from the assumed relationship between digital access and satisfaction, together with the exclusion of the less-connected group.

The calculation is not an empirical estimate of bias in online learning surveys. Actual bias may be smaller, larger, absent, or in another direction.

The improved design may provide a more defensible basis for estimating satisfaction across the intended student population. It cannot guarantee the elimination of all bias, but it addresses a predictable exclusion mechanism directly related to the research question.

Common Misconceptions About Internet-Dependent Research

How to Decide Whether Internet Access Should Be Required

Before choosing an online-only design, determine whether the target population can realistically meet the digital requirements. Do not assume that the availability of an online platform makes it the most appropriate method for every participant.

Evaluate the scientific necessity of internet access, the likely coverage gap, and the consequences of excluding people who cannot participate digitally.

Where digital exclusion is likely to affect recruitment, researchers may need to balance broader participation with feasibility and participant protection. Providing alternatives may require additional resources, but failing to account for digital access can also compromise the scientific value of the study.

Check Whether Internet Requirements Could Exclude Your Participants

Frequently Asked Questions About Internet Access and Research Exclusion

Internet Access Should Not Become an Unexamined Requirement for Participation

How to Cite This Guide

This guide is intended to be read, shared, and used in research, teaching, and academic work. If you draw on its ideas, explanations, or other content, please acknowledge the source by citing the guide. Doing so gives appropriate credit and helps your readers locate the original resource.