Topics in Current Status Data

摘要: This dissertation considers topics in current status data, a type of survival data where the only available information on time is whether or not event has occurred before examination time. We introduce concept and give some motivating examples to highlight many areas which this naturally occur practice. discuss well known widely used methods for analyzing along with more recent developments area, provide appropriate references these previously examined methods. Within dissertation, we add existing literature area by developing ideas addressed from perspective.We describe simple method nonparametric estimation distribution function based observations are subject (known) misclassification. Nonparametric maximum likelihood techniques obtained through use straightforward set adjustments familiar pool-adjacent violators algorithm, generally when misclassification assumed absent. The extended allow rates that vary over time, particularly most likely close true failure event. Using binary generalized linear models outcomes consider regression underlying motivated applied an example human papillomavirus (HPV) infection amongst women San Francisco. Additional applications breastfeeding behaviors menopausal also presented. As extension group testing presence Group combines samples, such as blood urine, number individuals tests sample disease interest instead each individual sample. examine can be reduce costs incurred large but improve efficiency estimating function. seek determine optimal size function, under various scenarios. Regression approach briefly considered.We perspective counting processes. relationship between Specifically multistate model defined two times one observes exceed common observed monitoring interested first subsequent estimate. For both single multiple scenarios, fully setting, cannot naive estimator, using only, smooth functionals firstevent (van der Laan Jewell (2003)). therefore improving estimator parametric assumptions about waiting events made. situations modifiable design, length cumulative hazard past. simultaneous accurate diluted HIV test data.