Approximate Small-Sample Tests of Fixed Effects in Nonlinear Mixed Models

Vólaufová, Júlia

doi:10.1080/03610918.2013.835407

Cited by 1 publication

(1 citation statement)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The large-sample tests often used are the likelihood ratio test (LRT) and the Wald test. It is well documented that rejection rates of these tests under the null hypothesis are greater than the nominal level (see, e.g., [10],[11], and more recently [12] and the references therein), especially when the so called first-order approximation to the marginal likelihood is used. Thus, our main objective is to investigate and compare properties of commonly used tests when constructed from a two-stage estimation method not involving the marginal likelihood.…”

Section: Introductionmentioning

confidence: 99%

Approximate testing in two-stage nonlinear mixed models

Burton¹,

Vólaufová²

2014

Journal of Statistical Computation and Simulation

View full text Add to dashboard Cite

We investigate here small sample properties of approximate F-tests about fixed effects parameters in nonlinear mixed models. For estimation of population fixed effects parameters as well as variance components, we apply the two-stage approach. This method is useful and popular when the number of observations per sampling unit is large enough. The approximate F-test is constructed based on large sample approximation to the distribution of nonlinear least squares estimates of subject-specific parameters. We recommend a modified test statistic that takes into consideration approximation to the large sample Fisher information matrix (See [1]). Our main focus is on comparing finite sample properties of broadly used approximate tests (Wald test and likelihood ratio test) and the modified F-test under the null hypothesis, especially accuracy of p-values (See [2]). For that purpose two extensive simulation studies are conducted based on pharmacokinetic models (See [3, 4]).

show abstract

Section: Introductionmentioning

confidence: 99%