Complex Scenes from the International Affective Picture System (IAPS): Agreement-Based Emotional Categories

Maryann Wei*, Steven Roodenrys, Leonie Miller, Emma Barkus

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

5 Citations (Scopus)
18 Downloads (Pure)


Complex scenes from standardized stimuli databases such as the International Affective Picture System (IAPS) are organized dimensionally rather than discretely. Further, the potentially unique function of socially relevant scenes is often overlooked. This study sought to identify discrete categories of complex scenes from the IAPS and to explore if there were qualitative features that make the emotional content of some social scenes identifiable with higher levels of agreement. One hundred and three participants (53.4% female, mean age 24.4) judged 118 IAPS scenes as reflecting fear, happy, sad, or neutral. A second judgment study was conducted with a separate group of participants (N = 117; 79.2% female; mean age 30.41) to further characterize valid affective scenes across the full range of basic emotions. Sixty images received agreement on their emotional category from > 70% of judges and were considered valid. IAPS identifier codes for these images are available for reference (along with the supplementary material from the second judgment study), organized by emotional and social content. An incidental observation was such that compared to nonsocial scenes, lower agreement rates were observed for social scenes across the board. Qualitative features of social scenes that were classified into emotional categories based on higher levels of agreement are discussed.

Original languageEnglish
Pages (from-to)194-201
Number of pages8
JournalExperimental Psychology
Issue number3
Publication statusPublished - 7 Sept 2020
Externally publishedYes


Dive into the research topics of 'Complex Scenes from the International Affective Picture System (IAPS): Agreement-Based Emotional Categories'. Together they form a unique fingerprint.

Cite this