Learn More
Model selection procedures often use a fixed penalty, such as Mallows' C p , to avoid choosing a model which fits a particular data set extremely well. These procedures are often devised to give an unbiased risk estimate when a particular chosen model is used to predict future responses. As a correction for not including the variability induced in model(More)
This study investigates the performance of a touch screen, mouse and trackball in a motion environment. A Stewart motion platform was used to generate a six-degree-of-freedom motion environment. Participants were placed in an environment where vehicle vibration was simulated. Tasks were used according to Fitts' Law to obtain the movement time, error rate,(More)
We propose a new method, the leaf measuring-interval index (LMI), to estimate leaf age in morphological and physiological studies of leaves. When the plastochron, the interval between the initiation of successive leaves, is constant, the well-known leaf plastochron index (LPI) provides a robust measure of leaf age. When the duration of the plastochron is(More)
BACKGROUND Antipsychotics have been regarded as the most effective therapy for schizophrenia; however, Taiwan's non-adherence rate for prescribed antipsychotics of 50-80% is an important issue that relates directly and significantly to schizophrenia patient relapse rates. PURPOSE This study examines the correlational and predictive relationships between(More)
It has been suggested that estrogen supplementation can augment the treatment effects of typical antipsychotic medication. We report the use of estrogen supplementation for four female chronic inpatients. These patients were diagnosed with schizophrenia or schizoaffective disorders and were treated with atypical antipsychotics. Premenstrual exacerbation of(More)
A penalized likelihood (PL) method for structural equation modeling (SEM) was proposed as a methodology for exploring the underlying relations among both observed and latent variables. Compared to the usual likelihood method, PL includes a penalty term to control the complexity of the hypothesized model. When the penalty level is appropriately chosen, the(More)
  • 1