Disaster research cannot be prospective, forcing only correlational cross-sectional designs.

Evidence

A further limitation reflecting the inherent practical difficulties involved in research on disasters is the fact that it is simply not possible—except by chance—to carry out prospective research on reactions to an earthquake. The fact that the design was not only entirely correlational but also cross-sectional, however, is one important reason why, even though the pattern of results was largely as hypothesized, alternative explanations are still plausible. — John Drury and Rupert Brown and Roberto González and Daniel Miranda, Emergent social identity and observing…, 126

Categorization

• QUESTION: Worth noting: even well-designed disaster studies are structurally limited to correlational inference, so causal claims here depend on convergent evidence rather than design.
• corroborated: 3 independent passages support this claim — John Drury and Rupert Brown and Roberto González and Daniel Miranda, Emergent social identity and observing…, 126
• XREF: connects to: Case studies of clear emergencies lack variability needed for causal comparisons. — John Drury, The role of social identity processes…, 76

Related

• Case studies of clear emergencies lack variability needed for causal comparisons. (similarity 0.87)
• Methodological access constraints explain psychology’s neglect of emergencies (similarity 0.79)