作者: Mark P. Mattson , Terry Phillips , Christopher H. Morrell , Veena Shetty , Ruiqian Wan
DOI:
关键词: Delta method 、 Linear regression 、 Correlation 、 Product (mathematics) 、 Standard error 、 Multivariate statistics 、 Repeated measures design 、 Econometrics 、 Mathematics 、 Nonlinear system 、 Statistics
摘要: A study was conducted of the relationships among neuroprotective factors and cytokines in brain tissue mice at different ages that were examined on effect dietary restriction protection after experimentally induced stroke. It interest to assess whether cross-product slopes pairs variables vs. age positive or negative. To accomplish this, product estimated tested determine if it is significantly from zero. Since measurements are taken same animals, models used must account for non-independence within animals. number approaches illustrated. First a multivariate multiple regression model employed. we interested nonlinear function parameters (the product) delta method obtain standard error estimate product. Second, linear mixed-effects fit allows specification an appropriate correlation structure repeated measurements. The again error. Finally, non-linear approach linear-mixed-effects conduct test. simulation investigates properties procedure.