A selection of articles presented at the Eighth Lukacs Symposium held at the Bowling Green State University, Ohio. They discuss consistency and accuracy of the sequential bootstrap, hypothesis testing, geometry in multivariate analysis, the classical extreme value model, the analysis of cross-classified data, diffusion models for neural activity, estimation with quadratic loss, econometrics, higher order asymptotics, pre- and post-limit theorems, and more.
". . .illuminating. . .. . . .throws light on how the development of computers is going to influence future lines of research."
---Calcutta Statistical Association Bulletin
"an excellent reference. . ..expands the horizons."
Challenges for categorical data analysis in the 21st century; consistency and accuracy of the sequential bootstrap; hypothesis testing in the 20th century with special reference to testing with misspecified models; the importance of geometry in multivariate analysis and some applications; R.A. Fisher in the 21st century; students can help improve college teaching - a review and an agenda for the statistics; the classical extreme value model - mathematical results versus statistical inference; the analysis of cross-classified data - notes on a century of progress in contingency data table analysis, and some comments on its prehistory and its future; diffusion models for neural activity; estimation with quadratic loss; econometrics in the 21st century; on stability of large (ecological) systems; multi-scale statistical approach to critical-area analysis and modelling of watersheds and landscapes; R.A. Fisher - the founder of modern statistics; higher order asymptotics - costs and benefits; the analysis of subject-specific agreement; some important classes of problems in statistical experimental design, multivariate analysis, and sampling theory; pre-limit and post-limit theorems for statistics; regression modelling with fixed effects - missing values and related problems; some recent developments in nonparametric inference for right censored and randomly truncated data; estimation of parameters in nonlinear regression models.