Codebook for 118 Visual Themes in 350 U.S. Climate Change News Images

A codebook containing 118 specific visual themes within 19 categories, developed for content analysis of news imagery. The themes were derived from coding 350 images associated with 200 randomly-selected U.S. print news stories about climate change, involving over 175,000 separate codings by three coders. The codebook was created by Stacy Rebich-Hespanha and colleagues, with the final categories and themes resulting from multiple codings and exhibiting excellent inter-coder reliability.

Use Cases

• Training or evaluating computer vision models to classify climate change imagery based on the 118 defined themes.
• Conducting content analysis of news media to study visual framing patterns related to climate change.
• Replicating or extending the methodology from the associated 2015 Environmental Communication research paper.
• Developing taxonomies or ontologies for annotating environmental imagery in other datasets.

Strengths

• Contains 118 specific, well-defined visual themes derived from a systematic content analysis.
• Based on a substantial foundation of 350 images and over 175,000 separate codings.
• Final codes exhibited excellent inter-coder agreement and reliability as reported in the associated peer-reviewed manuscript.

Limitations

• Column-level documentation is absent; field semantics must be inferred from the description after download.
• Row count is unknown, which may limit suitability assessment for specific modeling tasks.
• The underlying image dataset is a separate artifact; this codebook describes the annotation schema.

Provenance

Source

Stacy Rebich-Hespanha and research collaborators.

Collection Method

Developed through content analysis and multiple rounds of coding of print news imagery.

Time Range

Images are associated with news stories from an unspecified period, with the codebook published online in 2012.

Geography

Focus on U.S. print news stories.