The Level-of-Analysis Problem in International Relations

J. David Singer, World Politics, 14(1), 1961.

The requirements of an analytical model

• Accurate Description: A successful analytical model must provide a highly accurate and complete depiction of the phenomenon under study. This depiction should correlate with objective reality and empirical referents, offering a faithful representation.
• Completeness and Undistorted Picture: The scheme or model must present the phenomena in as complete and undistorted a manner as possible. This suggests the importance of capturing all relevant factors and dynamics.
• Correlation with Objective Reality: The analytical model should correlate with objective reality. In other words, it should accurately reflect the real-world dynamics and relationships that exist.
• Coincidence with Empirical Referents: The model should coincide with empirical referents to the highest possible degree. This means that the theoretical constructs and concepts used in the model should align with observable and measurable aspects of reality.
• Challenges in Accurate Representation: The text acknowledges that achieving an accurate representation of complex phenomena is difficult due to inherent complexities and intricacies.
• Balancing Distortion and Accuracy: Given the challenges, it's important to determine where distortion is least dysfunctional and where accuracy is absolutely essential. This suggests a need to strike a balance between achieving complete accuracy and practicality. 
• Capacity to Explain Relationships: The analytical model should possess the capability to explain the relationships that exist among the phenomena under investigation. This emphasis is on the model's explanatory power rather than solely on the accuracy of its description.
• Focus on Validity of Explanation: The primary concern shifts from the accuracy of description to the validity of explanation. In other words, the model's ability to provide meaningful and well-supported explanations becomes crucial.
  
• Valid, Thorough, and Parsimonious Causal Relationships: The model should be able to treat causal relationships as valid, thorough, and parsimonious. This means that the relationships established within the model should be logically sound, comprehensive, and free from unnecessary complexity.
  
• Priority of Explanation over Description: The theory's main focus is on explanation. In cases where there is a conflict between the requirements of accurate description and explanatory power, the latter should be given higher priority.
  
• Balancing Distortion and Detail: Given the complexities of representation, it's important to find a balance between accuracy and practicality. Recognizing where distortion is least dysfunctional and where accuracy is essential ensures a balanced approach.

• Reliable Prediction: An effective model should exhibit the potential to predict future outcomes reliably. This predictive capacity enhances the practical applicability of the model and its utility in forecasting trends.