Observations
Birth totals, name counts, media appearances, geographical differences and timing may all be measured accurately.
Academic slop
An underconstrained narrative wearing the credentials of a constrained inference.
The observations may be real, the citations genuine, the equations valid and the prose cautious. The failure occurs when the connections among those components remain looser than the presentation implies.
Academic slop is essentially joining the dots with credentialled backing.
The diagnostic
What does the model rule out?
A model earns explanatory value by excluding possibilities. A model compatible with every outcome is adaptable, not informative.
Methods
Regression, simulation, network models and differential equations may all be implemented correctly.
Inference
Several incompatible mechanisms may reproduce the same pattern, leaving the cause unidentified.
Interactive demonstration
Build a more persuasive story
Add plausible mechanisms. Notice what happens to narrative richness and inferential constraint.
Boundary conditions
How a paper avoids becoming slop
The remedy is not less mathematics or fewer citations. It is stronger constraint.
Pre-specify comparisons
State which models will be compared and which measurements decide among them before inspecting the most convenient pattern.
Name the non-identifiability
Say when different causes generate observationally equivalent results.
Separate fit from mechanism
A model reproducing a curve does not establish that its internal story produced the curve.
Publish the null result
If ordinary population size and an unequal name distribution explain the apparent pattern, stop there.