{"query_echo":{"q":"","filters":{"analysis_type":["replication study"]},"warnings":[{"code":"facet_value_not_in_vocabulary","message":"Value 'replication study' for facet 'analysis_type' is not in the controlled vocabulary."}]},"total":{"value":1,"relation":"eq"},"page":1,"page_size":20,"self":"https://climateinsightsdirectory.org/api/v1/search?analysis_type=replication+study","stability":"beta","ai_marking":{"digital_source_type":{"uri":"http://cv.iptc.org/newscodes/digitalsourcetype/trainedAlgorithmicMedia","name":"Created using Generative AI"},"notice":"Summary and key insights are generated using artificial intelligence."},"results":[{"id":"prio-001735","url":"https://climateinsightsdirectory.org/document/prio-001735","title":"Sensitivity Analysis of Climate Variability and Civil War","publisher":{"name":"Peace Research Institute Oslo","organization":"peace-research-institute-oslo"},"document_type":"research paper","language":"en","date_published":"2010-11-17","source_url":"https://cdn.cloud.prio.org/files/3db295cf-fa98-42f8-abd4-8183ea5ae75b/Sensitivity Analysis of Climate Variability and Civil War PRIO Paper 2010 v1.pdf?inline=true","generator":"Climate Insights Directory","generated_fields":["summary","key_insights","climate_tags","geography_discussed","analysis_type"],"summary":{"text":"The paper challenges the finding that warming is robustly linked to civil war in Sub-Saharan Africa by replicating and testing the model used by Burke et al. (2009). Through influence diagnostics, alternative operationalizations of civil war, and expanded sample periods, the authors find that the reported positive effect of temperature is tenuous and often disappears. The analysis suggests that the original results were driven by a few influential observations involving foreign intervention and that climate variables provide negligible predictive power compared to country-specific fixed effects and time trends.","ai_generated":true,"unit":"prio-001735#summary","date_modified":"2026-10-23"},"key_insights":[{"text":"The reported positive link between temperature and civil war is highly sensitive to a small number of influential observations. Removing six highly influential cases—including Guinea-Bissau (1998), Sierra Leone (1998), and Chad (1987, 1990)—causes the effect of temperature to decrease dramatically.","ai_generated":true,"unit":"prio-001735#insight-1","date_modified":"2026-10-23"},{"text":"A common characteristic of the most influential cases in the baseline model is the presence of foreign intervention, a variable omitted from the original analysis. The authors note a lack of references to climate anomalies or agricultural income loss in the narratives of these specific conflicts.","ai_generated":true,"unit":"prio-001735#insight-2","date_modified":"2026-10-23"},{"text":"The positive effect of temperature disappears when the definition of civil war is adjusted. While the result is reproduced using the PRIO Battle Deaths Dataset, it vanishes when the fatality threshold is shifted to 500 or 2,000 battle-related deaths, or when continuous measures of battle deaths are used.","ai_generated":true,"unit":"prio-001735#insight-3","date_modified":"2026-10-23"},{"text":"Expanding the sample period beyond 1981\u00132002 weakens the reported correlation. When the temporal domain is extended to 1981\u00132008, the regression coefficient for temperature drops by 40% and loses statistical significance; it is also non-significant for the full post-colonial sample (1960\u00132008).","ai_generated":true,"unit":"prio-001735#insight-4","date_modified":"2026-10-23"},{"text":"Climate variables have a trivial marginal impact on the prediction of civil war. A model containing only country fixed effects and time trends produces predictions nearly identical to the baseline model, with a correlation of r=0.999.","ai_generated":true,"unit":"prio-001735#insight-5","date_modified":"2026-10-23"},{"text":"Out-of-sample prediction tests for the period 1999\u00132002 show that climate variables do not improve the model's ability to correctly predict wars. The model with climate variables produced more false positives without increasing true positives compared to the reduced model.","ai_generated":true,"unit":"prio-001735#insight-6","date_modified":"2026-10-23"}],"climate_tags":["climate variability","temperature","precipitation","environmental degradation"],"geography_discussed":["Guinea-Bissau","Sierra Leone","Chad","Congo, Republic of","Sudan","Mozambique","Uganda","Ethiopia","Angola"],"frame_perspective":[],"analysis_type":["sensitivity analysis","replication study","econometric analysis"]}]}