Skip to content
Back to search results

Framing Analysis

Introduced by
Robert M. Entman (synthesis); roots in Goffman, Tuchman, Gitlin
Year
1993
Type
Interpretive-quantitative analysis of how messages select and emphasize aspects of reality
Family
Framing and emphasis research
Font size

What does it do?

Framing analysis is a communication research method for studying how messages select certain aspects of a perceived reality and make them more salient — promoting a particular problem definition, causal interpretation, moral evaluation, and treatment recommendation.

Building on Robert Entman's influential 1993 synthesis, it moves beyond counting what is present to reconstructing the organizing ideas, or frames, that give media coverage its meaning and persuasive shape.

The intuition behind the method

Two news stories can report the same facts about a protest yet feel completely different: one frames it as a 'riot' threatening order, the other as a 'demonstration' demanding justice. The facts overlap, but the selection of details, the labels, the sources quoted, and the implied solution differ — and those choices steer how audiences understand the event. Framing analysis treats these recurring bundles of selection and emphasis as the unit of interest. Rather than asking only 'how often does X appear?', it asks 'what coherent way of seeing the issue is being offered, and through which textual devices?'

History

The concept of framing traces to Erving Goffman's frame analysis and to sociologists like Gaye Tuchman and Todd Gitlin who studied how news organizes reality. Robert Entman's 1993 essay in the Journal of Communication consolidated a 'fractured paradigm' by defining framing as selection and salience serving four functions, and it became one of the most cited articles in communication. Subsequent scholars — including Baldwin Van Gorp, Dietram Scheufele, and Stephen Reese — developed inductive and deductive procedures, integrated framing with agenda setting and priming, and extended it to visual and computational analysis.

When to use it

Use framing analysis when your question is not merely whether a topic is covered but how it is rendered meaningful — which aspects are selected and made salient, and what interpretations they invite. It suits studies of contested issues where competing frames compete for public understanding, and it pairs naturally with agenda-setting work by addressing the 'how' to agenda setting's 'what.' It assumes that frames are real, recurring organizing principles detectable through textual devices, and that coders can be trained to recognize them reliably. It is less suited when the goal is a simple inventory of topics, when the corpus is too small to reveal recurring frames, or when frames are so contested that no reliable coding scheme can be agreed — situations better served by manifest content analysis or open-ended discourse analysis.

Limitations and considerations

  • Frame identification involves interpretive judgment, so different analysts may extract different frame sets from the same corpus.
  • Distinguishing a 'frame' from a mere topic, issue, or argument is conceptually slippery and inconsistently operationalized across studies.
  • Establishing reliable coding for latent frames is harder than for manifest features and often requires extensive coder training.
  • Demonstrating that a media frame actually shapes audience thinking requires separate effects research; content alone cannot prove influence.

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

Strengths

  • Captures meaning and emphasis that manifest counting misses, linking textual choices to interpretation and persuasion.
  • Bridges quantitative and qualitative traditions: frames are interpreted yet coded and reliability-checked.
  • Entman's four-function definition gives a portable, theory-grounded template for specifying frames across issues.
  • Connects message analysis to downstream questions of source power, public opinion, and policy through a well-developed theory.

Applications

  • Comparing how news outlets frame immigration, climate change, or public-health risks and how those frames shift over time.
  • Analyzing the frame contests between social movements, governments, and industry in coverage of policy debates.
  • Studying how strategic actors (campaigns, PR firms) attempt to set and win frames in news and social media.
  • Examining cross-national differences in how the same global event is framed for domestic audiences.
  • Combining frame coding with experiments to connect prevalent media frames to measured audience interpretations.

How it works

How it works

1

The analyst fixes the issue under study (e.g., immigration coverage) and assembles a corpus of relevant texts — news articles, broadcasts, posts, or speeches — sampled across outlets and time so that competing frames have a chance to appear.

2

Following Entman, each candidate frame is articulated in terms of the functions it performs: defining a problem, diagnosing causes, making moral judgments, and suggesting remedies. Frames may be deductively derived from theory or inductively discovered from a close reading of the material.

3

Frames are detected through their textual signatures — keywords, metaphors, catchphrases, exemplars, depictions, sources, and visual cues. These framing and reasoning devices are the concrete, codeable manifestations that let an abstract frame be located reliably in the text.

4

Coders apply the frame definitions and device indicators to the corpus, and intercoder reliability is established (typically with Krippendorff's alpha, observed disagreement D_o over chance disagreement D_e) so that frame assignments reflect the texts rather than individual coder judgment.

Formula
\alpha = 1 - \frac{D_o}{D_e}
5

The coded patterns are interpreted to show which frames dominate, how they vary across outlets or over time, and what problem definitions and remedies they privilege — connecting the textual analysis back to questions of power, source influence, and audience effect.

Common pitfalls

  • Conflating frames with topics, so that 'the economy frame' is really just the subject of the story rather than an organizing interpretation.
  • Building frame categories so broad or overlapping that nearly every text fits several, destroying analytic leverage and reliability.
  • Skipping device-level indicators and asking coders to judge 'the frame' holistically, which inflates disagreement.
  • Inferring audience effects directly from frame prevalence without any audience or experimental evidence.

Frequently asked questions

What is the difference between framing analysis and agenda setting?

Agenda setting concerns which issues the media make salient — the 'what to think about.' Framing concerns how an issue is presented once it is on the agenda — the selection and emphasis that shape 'how to think about it.' Many scholars treat framing as a second level of agenda setting (attribute agenda setting), while others insist framing is theoretically distinct because it involves applicable considerations and interpretive packages rather than mere accessibility.

Should frames be derived deductively or inductively?

Both are legitimate. Deductive framing analysis starts from theory-specified generic frames (e.g., conflict, human-interest, economic-consequences) and codes their presence, which favors comparability and reliability. Inductive analysis reads the corpus closely to let issue-specific frames emerge, which favors sensitivity to context at some cost in replicability. Many strong studies combine the two: induce frames qualitatively, then operationalize and code them for systematic, reliability-checked analysis.

Can framing analysis be automated with computational methods?

Partly. Unsupervised methods like topic modeling and clustering, and supervised classifiers trained on human-coded frames, can scale frame detection to very large corpora, and network-based approaches map co-occurring framing devices. These tools are powerful for surfacing patterns but require careful human validation, because frames are interpretive constructs that algorithms approximate rather than directly observe.

Related methods

Sources

  1. Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51–58.

    Open DOI reference
    Copied.
  2. Krippendorff, K. (2004). Content Analysis: An Introduction to Its Methodology (2nd ed.). Thousand Oaks, CA: Sage.

    ISBN: 9780761915454
    Copied.