Grandjean et al. (1998) employ a number of models, some explicit and many implicit. What do each of the following models in Grandjean et al. (1998) substitute for? For each, identify a key assumption.

Model: Hair mercury levels as a biomarker
Substitutes for:
Key assumption:

Model: Test score at high methylmercury levels
Substitutes for:
Key assumption:

Model: Score on the Boston Naming Test
Substitutes for:
Key assumption:

Model: Faroe Island residents
Substitutes for:
Key assumption: