The score test presents a number of issues. There are four dimensions of concern: (1) Is the parameter space constrained? (2) Is observed information used, or some form of expected information? (3) Are size and power computed under the null hypothesis, a correctly specified alternative, or a misspecified alternative? (4) Are computations asymptotic or small sample?
No takes yet. Share an insight, caveat, or question.
Verbeke et al. (2007) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: