How Do You Account for Missing or Unavailable Documents in Research?

What Should You Do When Important Documents Cannot Be Found or Accessed?

You are comparing university policies on generative AI. Some institutions publish comprehensive policy documents, while others provide only brief public announcements. Several universities appear to have no relevant documents on their websites.

Does that mean those institutions have no AI policies? Should you exclude them from the study, classify them as having no policy, or contact them for additional information?

Perhaps you are conducting archival research and discover that records from certain years are missing. Alternatively, access to internal institutional documents may be restricted.

These situations raise a methodological problem because missing documents may affect which institutions, events, decisions, or perspectives become visible in the analysis.

The challenge is to distinguish what the available evidence establishes from what remains unknown because documentary materials are missing, inaccessible, incomplete, or never created.

Document Missingness Explicitly and Avoid Treating Nonretrieval as Confirmed Absence

Understanding Documentary Missingness and Its Implications

What Does It Mean for a Document to Be Missing?

A document is missing relative to a defined expectation about what documentary evidence should exist or be available for the research.

For example, a study may seek all official university AI policies effective during a specified academic year.

If a policy cannot be located for one institution, the researcher must determine what that nonretrieval means.

The institution may have no formal policy, maintain internal guidance, use another document title, or publish materials in a location the search did not identify.

These possibilities have different implications.

Researchers should therefore define missingness in relation to the documentary population and retrieval procedure rather than assume that every unsuccessful search establishes absence.

Missing, Unavailable, and Nonexistent Documents Are Different

Several documentary conditions should be distinguished.

A missing document may be known or expected to exist but cannot be located.

An unavailable document may exist but be inaccessible because of restrictions, technical barriers, or institutional decisions.

A nonexistent document is one that was never created or does not exist within the relevant scope.

However, establishing nonexistence may be difficult.

For example, an institutional representative may confirm that no university-wide AI policy has been adopted, but separate departmental guidance may still exist.

Researchers should specify exactly what form of absence has been established and avoid broader conclusions than the evidence supports.

What Are Common Reasons Documents Are Missing?

The reasons for missingness should be investigated where feasible because they affect how the remaining corpus should be interpreted.

Why Missing Documents Matter Methodologically

Missing documents can alter the apparent distribution of themes, institutional practices, historical developments, or policy provisions.

For example, universities with well-developed public policy repositories may be overrepresented in a study relying exclusively on downloadable documents.

Institutions with internal or poorly indexed guidance may be underrepresented.

If public documentation practices are associated with institutional resources or governance capacity, the accessible corpus may systematically differ from the intended population.

Researchers should therefore consider whether document availability is related to the phenomenon being investigated.

Missingness is not necessarily random or inconsequential.

What Is Documentary Coverage Bias?

Documentary coverage bias occurs when the available or selected documents systematically fail to represent relevant portions of the target documentary population.

For example, an analysis of institutional AI governance based only on universities with searchable public policies may disproportionately capture institutions with stronger public communication infrastructure.

The resulting findings may accurately describe the accessible documents while providing a distorted account of the broader institutional population.

Researchers should distinguish coverage limitations from errors in coding or interpretation.

Analyzing every available document does not correct systematic gaps in the documentary frame.

Is Documentary Missingness the Same as Missing Data in Statistics?

There are conceptual similarities, but the situations are not identical.

In statistical analysis, missing data may involve absent values for known observations or variables.

In documentary research, entire documents, historical periods, institutional perspectives, or categories of records may be unavailable.

The reasons for absence may involve institutional practices, archival processes, confidentiality restrictions, or document destruction.

Statistical missing-data terminology can sometimes help clarify assumptions, but researchers should not automatically apply statistical techniques such as imputation to missing documents.

Inventing documentary content or treating modeled text as authentic evidence would be inappropriate.

Can Missingness Be Systematic?

Yes. Certain documents may be more likely to be preserved, published, or made accessible than others.

For example, institutions may publicly release successful program reports while keeping internal evaluations confidential.

Historical archives may preserve formal administrative records more consistently than informal correspondence.

These patterns can influence the apparent documentary record.

Researchers should examine whether unavailable materials are associated with document type, institutional characteristics, time period, or subject matter.

However, the existence of systematic missingness should be supported by evidence rather than presumed from an incomplete corpus alone.

How Do You Know Whether a Document Exists?

Researchers may use multiple appropriate sources to investigate documentary existence.

These can include official repositories, institutional websites, archival catalogues, administrative records, references in other documents, and direct inquiries to authorized offices.

For example, an institutional policy may be referenced in a faculty memorandum even when the original policy cannot be retrieved.

That reference provides evidence that a document may exist or previously existed, but it does not establish the complete contents of the missing material.

Researchers should record the evidence supporting existence claims and distinguish confirmed documents from inferred or reported ones.

What Should Researchers Do After an Unsuccessful Search?

An unsuccessful initial search should not automatically end the retrieval process when the document is important to the research question.

Researchers may examine alternative official repositories, search using different terminology, consult archival catalogues, or contact relevant institutional offices.

However, retrieval efforts should remain proportionate to the study's purpose and resources.

There is no universal number of searches or institutional contacts that guarantees completeness.

The important requirement is to establish and document a defensible retrieval procedure.

Researchers should avoid applying extensive searches to some institutions while accepting minimal searches for others without justification.

How Should Retrieval Attempts Be Recorded?

A retrieval log can document the sources searched, search terms used, dates of attempts, access conditions, and outcomes.

For example, the log may indicate that an institutional repository was searched, a policy was referenced in another document, and a request for access received no response.

The record should distinguish unsuccessful searches from confirmed institutional statements that no document exists.

Researchers may also document broken links, inaccessible files, and restricted archives.

These records support transparency and help establish the limits of the achieved corpus.

What Does a Missing-Document Register Look Like?

The register should preserve uncertainty rather than force every unavailable document into a definitive category.

Should Institutions with Missing Documents Be Excluded?

Not automatically.

The appropriate treatment depends on the research question and unit of analysis.

For example, a study of publicly available university AI policies may legitimately restrict its documentary analysis to accessible policies, provided the population is defined accordingly.

However, a study claiming to compare AI governance across all universities in a jurisdiction may require information about institutions without accessible documents.

Excluding those institutions could introduce selection bias.

Researchers should distinguish exclusion from content analysis, exclusion from the institutional sample, and classification as lacking a policy.

These decisions have different implications.

Can Missing Documents Be Coded as Zero?

Only when zero accurately represents the defined variable and the absence has been established.

For example, a verified absence of a formal university-wide AI policy might be coded as no such policy under a clearly defined classification.

However, failure to retrieve a policy should not automatically receive the same code.

A missing document may indicate unknown status rather than confirmed absence.

Researchers should distinguish zero, not applicable, not observed, and unknown where relevant.

Collapsing these categories can produce misleading prevalence estimates and institutional comparisons.

How Do Missing Documents Affect Quantitative Content Analysis?

Quantitative content analysis may estimate the prevalence of particular documentary features.

For example, a researcher may calculate the proportion of university policies containing explicit AI disclosure requirements.

If policies from some institutions are unavailable, the denominator and target population require careful definition.

The observed proportion among accessible policies does not necessarily estimate the proportion among all universities.

Researchers should report coverage and avoid treating unavailable cases as negative observations.

Where appropriate, sensitivity analyses may examine how conclusions depend on plausible assumptions about missing cases, without presenting assumed values as observed facts.

How Do Missing Documents Affect Qualitative Analysis?

Qualitative documentary analysis may be affected when missing materials exclude important perspectives, historical periods, or contrasting cases.

For example, a study of institutional policy development may have complete records of final decisions but limited evidence about earlier disagreements.

The analysis may adequately explain formal outcomes while providing insufficient evidence about the deliberative process.

Researchers should distinguish what the available corpus supports from interpretations requiring missing materials.

Where documentary gaps are substantial, the research question or scope may need refinement.

However, researchers should not assume that every missing document invalidates the entire study.

What If Missing Documents Are Concentrated in Certain Periods?

Temporal gaps can distort historical interpretations.

For example, a university archive may contain AI guidance issued during recent years but lack earlier drafts or memoranda.

The researcher might observe an apparent shift in institutional policy without adequate evidence about intermediate developments.

Researchers should map documentary coverage across relevant periods and distinguish documented changes from gaps in the record.

Absence of preserved evidence should not automatically be interpreted as absence of institutional activity.

Historical conclusions should reflect the surviving documentary record.

What If Missing Documents Are Concentrated in Certain Institutions?

Institutional differences in record preservation and public accessibility may affect comparisons.

For example, universities with comprehensive online repositories may contribute more documents than institutions with limited digital infrastructure.

These differences may reflect documentation practices rather than substantive differences in governance.

Researchers should consider whether institutional characteristics are associated with document availability.

Where the study compares institutions, missingness may need to be reported at the institutional level rather than only as a total document count.

Can Alternative Documents Replace Missing Originals?

Sometimes alternative sources can provide partial evidence about a missing document.

For example, an implementation memorandum may summarize provisions from an unavailable policy.

Meeting minutes may record approval of a document that cannot be retrieved.

However, these materials should not automatically be treated as equivalent to the original.

A summary may omit qualifications, exceptions, or detailed requirements.

Researchers should identify the alternative source and limit conclusions to what it actually establishes.

Reconstructing the full contents of an unavailable document without evidence would be methodologically inappropriate.

Can Interviews Help Address Documentary Gaps?

Interviews may provide information about whether documents exist, how they were used, or why they are unavailable.

For example, an institutional administrator may explain that AI guidance was distributed internally rather than published on the university website.

However, an interview account is not the missing document itself.

Participants may misremember provisions or interpret them differently.

Researchers should distinguish interview evidence about documentary practices from direct analysis of documentary content.

The decision about triangulating document analysis with other sources should follow the research question and evidentiary needs.

Should Researchers Impute Missing Documentary Content?

Researchers should not invent or reconstruct the contents of missing documents and present them as authentic documentary evidence.

Statistical modeling may sometimes be used to explore uncertainty in quantitative variables derived from incomplete documentary coverage.

However, such models require explicit assumptions and should not be confused with retrieving or verifying the missing source.

For qualitative interpretation, the appropriate response is generally to acknowledge documentary gaps and avoid unsupported reconstruction.

Generative AI systems should not be used to fabricate plausible policy provisions or archival records to fill missing sections of a corpus.

Analytical convenience cannot substitute for evidence.

What Is a Sensitivity Analysis for Missing Documents?

Sensitivity analysis examines whether conclusions change under alternative plausible assumptions about missing information.

For example, suppose 40 universities are included in a study, but official AI policy status can be established for only 32.

The researcher may report the observed distribution among the 32 institutions and examine how the overall proportion would vary under specified assumptions about the eight unresolved cases.

However, the assumptions should be clearly labeled and justified where possible.

Such analysis does not reveal the actual contents of missing policies.

Its purpose is to demonstrate how uncertainty affects conclusions rather than conceal that uncertainty.

When Should Missingness Change the Research Question?

Substantial documentary gaps may prevent the original question from being answered adequately.

For example, a researcher may initially seek to reconstruct the complete development of university AI governance over ten years.

If early records are largely unavailable, a comprehensive historical account may not be defensible.

The study might instead examine documented developments during the period for which adequate records exist.

However, such revisions should be transparent and justified by evidentiary limitations rather than a desire to produce more favorable findings.

The revised question should remain substantively meaningful and consistent with the available evidence.

When Do Missing Documents Make a Study Infeasible?

Documentary missingness may make a study infeasible when the unavailable materials are essential to answering the central research question and no adequate alternative evidence exists.

For example, a study seeking to compare the exact wording of confidential policies cannot proceed as intended if those policies cannot be accessed.

However, the same documentary limitations may be acceptable for a study explicitly examining publicly available policy communication.

Feasibility therefore depends on the intended inference rather than the proportion of missing documents alone.

Researchers should assess whether the remaining corpus supports a coherent and worthwhile research question.

How Should Missing Documents Be Reported?

The methods section should describe the documentary population, retrieval procedures, accessible corpus, and known or suspected coverage gaps.

Researchers should distinguish documents confirmed not to exist from those that could not be located or accessed.

Where possible, report missingness by relevant institution, period, or document type.

Explain how unavailable materials were treated during coding and analysis.

Where missingness may influence conclusions, discuss its likely direction or consequences cautiously.

Transparent reporting is preferable to presenting the accessible corpus as complete without evidence.

Handling Missing University AI Policies in a Comparative Study

The example demonstrates why missing documents should not be treated as zero observations. Different reasons for nonretrieval require different analytical treatment, and the final claims must reflect the achieved coverage.

Common Misconceptions About Missing Documentary Evidence

How to Manage Missing Documents Without Distorting the Findings

Begin by defining what documentary evidence should exist or be accessible under the study's eligibility criteria.

Then investigate nonretrieval systematically and distinguish the different statuses of unavailable materials.

Documentary gaps should be treated as part of the methodological evidence about what can and cannot be established.

The objective is not to eliminate every uncertainty, but to prevent unavailable information from being silently transformed into unsupported findings.

Check How Missing Documents Affect Your Research

Frequently Asked Questions About Missing and Unavailable Documents

Missing Documents Create Uncertainty, Not Automatic Evidence of Absence

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.