Viewing society from space: Image-based sociocultural prediction models

Report an error

Summary

AI-generated

This summary is written by a language model reading the source document. It is not the publisher's words and is not a substitute for the original.

Learn more about AI enrichment

The research demonstrates that automated processing of satellite imagery can predict local infrastructure access and certain socioeconomic attitudes in Botswana, Kenya, and Zimbabwe by correlating physical observables with Afrobarometer survey data.

Key insights

AI-generated

These insights are written by a language model reading the source document. They are not the publisher's words and are not a substitute for the original.

Learn more about AI enrichment
  • The study successfully developed models that predict people's attitudes regarding security, leadership, social involvement, and the economy using only imagery-derived information.
  • Model performance is strongest when predicting access to physical infrastructure, such as electricity, clean water, and sewage systems, compared to social attitudes which often perform only slightly better than chance.
  • In Botswana, the models achieved high accuracy for individual responses regarding infrastructure, specifically 94.1% for piped water access and 91.9% for electricity access.
  • Aggregating survey data to the village level generally results in more accurate predictions than classifying at the level of individual respondents.
  • Certain image features are critical for predictions, including spectral features for vegetation health, spatial features indicating urbanization, and indicators of social connectivity like road density.
  • The model's ability to predict cell phone access in Kenya was very low (28%) because the extensive nature of the network makes urban-versus-rural location data an insufficient predictor.
  • The researchers identified that the current use of one-kilometer square image chips lacks the necessary spatial context for larger urban areas, where local neighborhood differences significantly impact socioeconomic status.

Cite the original document

APA
Irvine, J. M., Wood, R. J., & McBee, P. (2017). Viewing society from space: Image-based sociocultural prediction models. Afrobarometer. https://www.afrobarometer.org/wp-content/uploads/2022/02/afropaperno178_viewing_society_from_space.pdf
Chicago
Irvine, John M., Richard J. Wood, and Payden McBee. Viewing society from space: Image-based sociocultural prediction models. Afrobarometer, 2017. https://www.afrobarometer.org/wp-content/uploads/2022/02/afropaperno178_viewing_society_from_space.pdf.
Wikipedia
{{cite report |last1=Irvine |first1=John M. |last2=Wood |first2=Richard J. |last3=McBee |first3=Payden |title=Viewing society from space: Image-based sociocultural prediction models |publisher=Afrobarometer |date=December 2017 |url=https://www.afrobarometer.org/wp-content/uploads/2022/02/afropaperno178_viewing_society_from_space.pdf |access-date=5 November 2026 |via=Climate Insights Directory (document afbarom-000525)}}
BibTeX
@techreport{irvine2017viewing, author = {Irvine, John M. and Wood, Richard J. and McBee, Payden}, title = {{Viewing society from space: Image-based sociocultural prediction models}}, institution = {Afrobarometer}, year = {2017}, month = dec, url = {https://www.afrobarometer.org/wp-content/uploads/2022/02/afropaperno178_viewing_society_from_space.pdf}, urldate = {2026-11-05}, note = {Indexed by Climate Insights Directory, document afbarom-000525} }

Full text

Collected · Record updated