Skip to content
Back to search results

Visual Analysis — Visual Analysis in Qualitative Research

Introduced by
Roots in art history and semiotics (Panofsky, Barthes); social science applications developed by Gillian Rose and Marcus Banks
Year
Formalized in social sciences from the 1980s–2000s
Type
Qualitative research approach
Family
Qualitative design / analysis
Font size

What does it do?

Visual analysis is a qualitative research approach that systematically examines visual materials — such as photographs, films, artworks, advertisements, and diagrams — to understand how meaning is produced, communicated, and interpreted.

Drawing on traditions from art history, semiotics, and social science, it treats visual objects as data that carry social, cultural, and ideological significance. Multiple frameworks exist, from formal compositional analysis to discourse-based and audience-reception approaches.

The intuition behind the method

When a researcher studies how refugee experiences are represented in news photography, or how gender roles are constructed in advertising imagery, quantitative tools fall short — you cannot count what an image means. Visual analysis provides structured ways to 'read' images as texts. Instead of asking how often an image appears, it asks what the image says, how it says it, and in what social context it operates. The analyst examines composition, symbolism, colour, spatial arrangement, and the conditions under which an image was produced and consumed.

History

Visual analysis has roots in art history, particularly Erwin Panofsky's three-level iconographic method developed in the 1930s–1950s. Roland Barthes' semiotic analyses of photography and popular imagery in Mythologies (1957) and Camera Lucida (1980) were foundational for cultural and media studies. In social science, visual methods gained momentum from the 1980s with the work of John Collier on photo elicitation, and were systematized for qualitative research by Gillian Rose in Visual Methodologies (2001) and Marcus Banks in Using Visual Data in Qualitative Research (2007).

When to use it

Visual analysis is appropriate when the research question concerns how meaning is constructed, communicated, or negotiated through visual materials, and when those materials are central rather than supplementary evidence. It suits research on media representation, identity, cultural practices, political imagery, advertising, art, and online visual culture. Choose visual analysis when verbal data alone cannot answer the question — for example, studying how authority is performed in institutional photography or how social media images construct health ideals. It is not appropriate when visual materials are tangential to the research question, when no clear interpretive framework can be specified, or when the researcher lacks the time and skills to engage systematically with a visual corpus. It is also not a substitute for ethnographic observation when context is the primary object of study.

Limitations and considerations

  • Interpretations are inevitably shaped by the researcher's cultural background and theoretical framework, raising questions of positionality and validity.
  • Findings are context-specific and not statistically generalizable to broader populations.
  • The diversity of visual analysis frameworks (semiotic, compositional, discourse-based, etc.) can make methodological decisions confusing for new researchers.
  • Large visual corpora are time-intensive to analyze rigorously; sampling strategies must be carefully justified.

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

Strengths

  • Enables rigorous investigation of visual culture, media, and representation that text-based methods cannot access.
  • Theoretically flexible — applicable across semiotic, discourse-analytic, iconographic, and psychoanalytic frameworks.
  • Treats visual materials as primary evidence rather than illustrations, elevating their analytical status.
  • Particularly powerful for studying power, ideology, identity, and representation in mediated and everyday contexts.
  • Compatible with mixed or multi-method designs — visual analysis can complement interviews or ethnography.

Applications

  • Examining how gender, race, and class are represented in magazine advertising or social media imagery.
  • Analyzing the visual rhetoric of political campaigns, propaganda, or public health communications.
  • Studying how institutions (hospitals, schools, corporations) construct authority and identity through official photography.
  • Investigating how online communities use memes, infographics, or selfies to negotiate identity and belonging.
  • Exploring how artworks or documentary photography represent marginalized communities or historical trauma.
  • Researching children's drawings or patient-produced images as data in health or educational research.

How it works

How it works

1

The researcher defines a corpus of visual materials relevant to the research question — photographs, film stills, advertising images, artworks, social media visuals, or archival documents. Selection criteria are established (purposive, representative, or comprehensive) and documented to make the data boundaries transparent and defensible.

2

Each visual item is subjected to a formal or compositional analysis: describing what is literally visible — subjects, objects, colours, spatial arrangement, lighting, text, and compositional elements. This descriptive pass, analogous to Panofsky's pre-iconographic level, provides an objective baseline before interpretive claims are made.

3

The researcher moves from description to interpretation, examining what the image signifies, how it constructs meaning, and what symbolic or ideological work it performs. Depending on the chosen framework — semiotic, discourse-analytic, iconographic, or psychoanalytic — different analytic tools are applied to uncover connotative and culturally embedded meanings.

4

Meanings are situated within their production and reception contexts: who made the image, under what conditions, for what audience, and how it has been circulated and used. This step prevents decontextualized readings and connects the visual material to broader social, historical, or institutional structures.

5

Interpretations across the corpus are integrated into coherent findings — identifying patterns, recurring motifs, contradictions, or discourses that run across images. The synthesis is presented as a theoretically grounded account of what the visual materials reveal about the phenomenon under study.

Common pitfalls

  • Failing to specify an analytic framework — describing images without a principled interpretive framework produces impressionistic commentary rather than systematic analysis.
  • Decontextualized reading — interpreting images without attending to their production context, circulation, and intended audience leads to misleading conclusions.
  • Corpus bias — selecting only images that confirm the researcher's hypothesis rather than building a corpus on principled and transparent criteria.
  • Collapsing description and interpretation — blending what is literally depicted with what it means obscures the analytic process and makes claims harder to evaluate.
  • Neglecting researcher reflexivity — not acknowledging how the researcher's own cultural position shapes their reading of visual materials undermines credibility.

Frequently asked questions

What is the difference between content analysis and visual analysis?

Quantitative content analysis counts the frequency of predetermined categories in images (e.g., how often women are shown in passive roles), producing numerical summaries. Visual analysis — in the qualitative sense — interprets how meaning is constructed in images, attending to composition, symbolism, context, and ideology. They answer different questions: frequency versus meaning. Some studies combine both.

Do I need to be an expert in art history to use visual analysis?

No, but you do need to select and learn an appropriate analytic framework before beginning. Gillian Rose's Visual Methodologies provides an accessible overview of the main frameworks used in social science. Domain knowledge about the visual tradition you are studying (e.g., photojournalism conventions, advertising codes) strengthens the analysis but can be developed through focused reading.

How large should my corpus of images be?

There is no fixed rule. In-depth analysis of a small purposive corpus (10–30 images) is common when the goal is rich interpretation. Larger corpora may be appropriate for identifying patterns across a dataset, but the trade-off is between breadth and depth of analysis. The key is that corpus size is justified by the research question and that the sampling strategy is transparent.

Can I use visual analysis alongside interviews?

Yes — combining visual analysis with participant interviews is a well-established strategy. Photo elicitation interviews, for example, use images as prompts to generate richer verbal accounts than conventional interviews alone. Triangulating visual and verbal data can strengthen both the interpretive claims and the overall credibility of the study.

How do I handle researcher subjectivity in visual analysis?

Reflexivity is essential: document the framework you are using, explain why you selected it, and acknowledge how your own cultural and theoretical position shapes your readings. Seeking interpretive feedback from colleagues or, where appropriate, from image producers or subjects, and maintaining an audit trail of your analytic decisions, are recognized strategies for managing subjectivity without pretending it can be eliminated.

Related methods

Sources

  1. Rose, G. (2016). Visual Methodologies: An Introduction to Researching with Visual Materials (4th ed.). Sage.

    ISBN: 978-1473943056
    Copied.
  2. Banks, M. (2007). Using Visual Data in Qualitative Research. Sage.

    ISBN: 978-0761943754
    Copied.