Skip to content
Back to search results

Document Analysis

Introduced by
Max Weber and Karl Mannheim
Year
1920
Type
Method
Family
data-collection
Font size

What does it do?

Document analysis is a systematic qualitative research method for examining written, visual, or audiovisual sources—such as policy documents, historical records, organizational records, media reports, emails, social media posts, photographs, or videos—to extract meaning, identify patterns, and understand social phenomena.

Developed by Weber and Mannheim in early 20th-century sociology, the method bridges historical research, content analysis, and textual interpretation. Document analysis is used across disciplines to understand organizational change, policy evolution, media representation, historical events, and cultural meaning. Documents provide evidence of what organizations, institutions, or societies value, decide, and communicate, often revealing contradictions between policy and practice.

The intuition behind the method

Documents are traces of human activity and meaning. They represent what organizations, institutions, or individuals deemed important enough to record and preserve. Unlike interviews or observations, which capture what people say or do in the research context, documents capture what people actually wrote, decided, or created for purposes often unrelated to research. This provides a less reactive data source—documents exist independently of the researcher's questions. Document analysis is particularly valuable for understanding historical change (comparing documents across time), organizational decision-making (tracing how decisions were made and justified), or institutional priorities (what gets documented and how). Analyzing what is written reveals what is valued; analyzing what is absent reveals what is ignored or suppressed.

History

Document analysis emerged in early 20th-century sociology through Weber's and Mannheim's examination of texts as sources of meaning and ideology. Scott's (1990) framework systematized document analysis by emphasizing authenticity, credibility, representativeness, and meaning—criteria for assessing document quality and interpreting findings. Prior (2003) advanced understanding of how documents shape social practices and are shaped by them. Content analysis developed in parallel (Krippendorff 1980), offering systematic coding methods applicable to document analysis. Contemporary document analysis encompasses digital documents (emails, social media), visual documents (photographs, infographics), and multimodal analysis (combining text, image, and video). The rise of big data and text mining has enabled analysis of vast document sets, though close reading and interpretation remain central to qualitative document analysis.

When to use it

Document analysis is most valuable for understanding historical change and evolution (comparing documents across time); investigating organizational decision-making and institutional priorities; analyzing policy development and implementation gaps; examining how institutions represent themselves and frame issues; studying cultural values and ideologies as expressed in public or private discourse; understanding rare or inaccessible phenomena (archived letters, historical events); or validating findings from interviews (do participants' recollections match documented records?). It is less suitable when seeking information about individuals' internal thoughts or emotions (interviews preferred), when document availability is very limited, or when the research requires direct observation of behavior.

Limitations and considerations

  • Selection bias in available documents: archival records reflect what was preserved, often biased toward powerful actors' perspectives; marginalized voices are underrepresented.
  • Incomplete record: important decisions or communications may never have been documented; absences are difficult to interpret.
  • Context loss: understanding the original production context (why the document was created, who created it, for what purpose) is often incomplete.
  • Author bias and audience: documents are written for purposes other than research; authors' intentions, perspectives, and constraints shape content.
  • Interpretation challenges: inferring meaning from text alone, without ability to ask clarifying questions as in interviews.

Source conditions do not replace independent scientific review or validation by an analysis engine.

Strengths

  • Non-reactive data source: documents exist independently of the research, uncontaminated by researcher presence or questions.
  • Historical depth: access to documents spanning years or decades enables understanding of change and development over time.
  • Rich contextual information: documents often contain detailed background and context that interviews may not capture.
  • Cost-effective: many documents are publicly available or archived; no need for participant recruitment or ethics approval if analyzing publicly available documents.
  • Reveals organizational values and priorities: what gets documented and how it is framed reveals what institutions care about.

Applications

  • Organizational research: analyzing internal documents (memos, policies, emails) to understand organizational culture, decision-making processes, or implementation of change initiatives.
  • Health policy and healthcare: examining policy documents, clinical guidelines, patient records, or media coverage to understand how health issues are framed and how policy evolves.
  • Education research: analyzing school documents (policies, curriculum materials, student records), textbooks, or educational media to understand institutional priorities and pedagogy.
  • Social and historical research: examining archival records, historical documents, oral history transcripts, or media coverage to reconstruct events and understand social change.
  • Program evaluation: reviewing program documents, reports, and communications to assess how programs are implemented, what outcomes are prioritized, and whether practice matches policy.

How it works

How it works

1

Identify the universe of relevant documents. Define inclusion/exclusion criteria: document type (policy, email, minutes, news article, photo), date range, authorship, language, topic relevance. Search for documents: archival records, organizational repositories, freedom of information requests, published sources, social media, gray literature. For large document sets, use purposive or systematic sampling (e.g., every tenth document, or stratified sampling by time period). Obtain documents in their original format when possible; note metadata (author, date, source, context of production). Create a document log noting what you included and excluded and why. Ensure authenticity: original documents are preferable to summaries or interpretations. For sensitive documents, note access restrictions and how these may bias the available record.

2

Examine each document's surface features: author identity and authority, intended audience, date and historical context, purpose and genre (policy vs. opinion vs. narrative). Conduct manifest content analysis: what is explicitly stated? Code documents using qualitative coding: assign labels to meaningful segments. Use open coding (categories emerge from data) or structured coding (predetermined framework). Organize codes into categories and themes. Examine latent content: underlying meanings, ideologies, biases, or assumptions reflected in language and framing. Analyze discourse: how topics are discussed, what language is used, what is absent or hidden. Create content matrices comparing documents (themes present/absent across documents, changes over time). Track provenance: document source, version, and access context. For large document sets, consider qualitative content analysis (quantifying frequency of codes) or quantitative content analysis (counting occurrences of specific terms or themes).

3

Interpret documents within their production context: who created it, for what purpose, for what audience? Consider the author's positionality and potential biases. Distinguish between explicit content (what is stated) and implicit meaning (what is assumed or hidden). Analyze intertextually: how do documents reference each other or reflect broader discourse? Look for contradictions: between what documents say and what occurred, between different documents, or between official policy and informal practice. Assess the significance of absences: what is not documented? What voices or perspectives are missing? Use documents to construct historical narratives or organizational timelines. Triangulate documents with interviews or observations: do organizational actors recall events as documents describe them? Use findings to generate theory about organizational change, power, discourse, or social processes.

Common pitfalls

  • Taking documents at face value: failing to examine author perspective, audience, purpose, and context; accepting stated claims without critical analysis.
  • Cherry-picking quotations: selecting documents or excerpts that support preexisting views rather than systematically analyzing the full document set.
  • Ignoring metadata and provenance: not tracking where documents came from, when they were created, who wrote them, or how they were preserved.
  • Over-interpreting absences: assuming that what is not documented did not happen, when in fact important work may not be written down.
  • Treating documentary evidence as objective truth: forgetting that documents are authors' representations of reality, not transparent windows into events.

Frequently asked questions

Do I need ethics approval to analyze documents?

If analyzing publicly available documents (published reports, historical records, freely available media), ethics approval is typically not required. If analyzing confidential or sensitive documents (internal organizational files, personal records, health records), you must obtain ethics approval even if documents already exist. Document the status of each document (public vs. restricted) in your methods.

What if I have hundreds of documents to analyze?

Use systematic sampling: define inclusion/exclusion criteria, then sample documents systematically (e.g., every fifth document, or stratified by type or date). Alternatively, use qualitative content analysis: code a sample of documents thoroughly, then apply the coding framework to remaining documents, quantifying code frequencies. Consider software (NVivo, ATLAS.ti) to manage and code large document sets.

How do I interpret documents I don't have full context for?

Use available metadata (author, date, source, genre) to infer context. Research the historical or organizational period. If possible, interview people who created or used the documents (triangulation). Note assumptions and limitations in interpretation: 'Based on available information, this document appears to...' rather than claiming certainty.

Should I analyze digital documents (emails, social media) differently?

Digital documents have unique features: they are often time-stamped, may be part of a visible thread or conversation (providing context), and may reach broad audiences. Apply the same analytical steps (manifest and latent content, author/audience analysis, contextualization) but account for the digital medium's affordances and constraints (permanence, visibility, searchability).

What do I do if documents contain contradictory information?

Contradiction is analytically interesting: different authors may have different perspectives or incentives to present information differently. Analyze why contradictions exist: different interpretations of the same events, changes over time, or deliberate misrepresentation. Use contradictions to explore power and competing narratives rather than trying to resolve them to a single truth.

Related methods

Sources

  1. Scott, J. (1990). A Matter of Record: Documentary Sources in Social Research. Polity Press.

    ISBN: 978-0745608419
    Copied.
  2. Prior, L. (2003). Using Documents in Social Research. SAGE Publications.

    ISBN: 978-0761959052
    Copied.
  3. Bowen, G. A. (2009). Document analysis as a qualitative research method. Qualitative Research Journal, 9(2), 27-40.

    Open DOI reference
    Copied.
  4. Flick, U. (2014). An Introduction to Qualitative Research (5th ed.). SAGE Publications.

    ISBN: 978-1446282921
    Copied.