▼ The present study explored the effects of age, decision styles, and processing strategies on using choice deferral. Thirty-three younger adults and 30 older adults finished a decision task of purchasing a stereo system on the computer. The MouselabWEB software was used to trace participants’ decision making processes. Results indicated that individuals who used more alternative-based processing patterns and took more time to decide were more likely to use choice deferral. On the other hand, individuals who used more attribute-based processing patterns and took less time to decide were more likely to use choice deferral. Further, older adults who were higher on maximization scores were more likely to use choice deferral; whereas, younger adults who were lower on maximization scores were more likely to use choice deferral. Advisors/Committee Members: Boudreau, Nancy.

▼ Latent Curve Analysis (LCA) is a statistical technique used for longitudinal studies that combines the methods of factor analysis with linear models. The purpose of LCA is to allow the modeler to combine common variables into a smaller number of random variables. The random variables are then regressed linearly to allow the modeler to express how the variables change over time. Due to the complexity of LCA, a researcher is forced to choose between a one growth curve model, or spend countless hours attempted to locate a better model. The purpose of this study is to create general methods that will allow future researchers the ability to create more adequate LCA models than the one growth curve model in a timelier manner. The data used for the thesis came from a longitudinal study dated 2000 to 2004. The study was a positive orientation survey that measured seven characteristics of positive orientation for 45 males and 81 female participants over three time periods. Using the program LISREL, a one growth curve LCA model was first analyzed by a group of researchers named Guido Alessandri, Gian Vittorio Caprara, and John Tisak. Using a variety approaches, a total four techniques were created in effort to solve the problem. Using a chi square test, two of the techniques resulted in models that were statistically significantly better than the one LCA model. These two models are discussed in detail within the paper. Though only two of the four techniques produced a better model than the one growth curve LCA model, each of the four methods could prove helpful for future researchers attempting to solve this problem. Advisors/Committee Members: Boudreau, Nancy.

▼ Unreplicated fractional factorial experiments are sometimes used as screening experiments whenever replication is not practical due to cost or time constraints. It is then important to be prudent when choosing which factors are to be included in the model. This study suggests a method in identifying significant location and dispersion effects using generalized linear modeling approach. A critical values table was generated and the difference in the minus twice the log-likelihood of two different joint models is compared against the critical values to determine significance. Advisors/Committee Members: McGrath, Richard.

▼ Categorizing human remains by sex and race has long been a challenge for the medicolegal profession. Logistic regression models and multicategory logit models can be used to accurately place individuals into their respective groups using measurements of the talus bone. Successful placement was achieved in 93% of the individuals for the response variable of sex. The percentage of correct placement for race alone was 95%. The correct placement for sex with race ranged from 76% to 95% depending on the group. The goal of this paper was to show that in addition to the pelvis and skull, the talus can be used with equal or greater success rates to classify human remains. Advisors/Committee Members: McGrath, Richard.