AN ABSTRACT OF THE THESIS OF MEGAN K. CLEARY, for the Master of Arts degree in ANTHROPOLOGY, presented on MARCH 28th at 8am, at Southern Illinois University Carbondale TITLE: SEX ESTIMATION FROM THE CLAVICLE: A DISCRIMINANT FUNCTION ANALYSIS MAJOR PROFESSOR: Dr. Gretchen R. Dabbs The development of methods for sex estimation using postcranial remains other than the os coxa is imperative for physical anthropology to improve the reliability of biological profile estimates in cases of incomplete and/or fragmentary skeletal remains. As the last skeletal element to complete fusion, the clavicle has the longest period of time to develop sexually dimorphic features, making it an ideal skeletal element for use in sex estimation. Sexual dimorphism in the clavicle was assessed using 18 measurements of the left clavicle of 265 (129 females; 136 males) individuals from the Hamann-Todd Collection. Independent samples t-tests with Bonferroni correction show males and females differ at a statistically significant level for all 18 variables with a significance level of 0.0028. Discriminate function analyses using the stepwise method (0.05 to enter, 0.10 to exit) produced a four variable model with cross-validated accuracy of 89.8%. A holdout sample from the Hamann-Todd Collection (n=30) similar in demographic character to the calibration sample was tested using the four variable model. The accuracy of the four variable model on the holdout sample was 90.0%. Additionally, four single variable models developed to accommodate fragmentary remains also have high predictive power (75.1-82.3% cross-validated calibration sample; 60.0-86.7% hold-out sample).

One of the basic goals of the physical anthropologist is to create a biological profile, consisting of sex, ancestry, age, and stature, from the skeletal material that they are presented with. This thesis seeks to explore size and shape differences related to sex and ancestry from the hyoid bones of the Robert J. Terry Anatomical Collection in order to gauge its usefulness in the process of developing a biological profile. A series of measurements were taken from 398 hyoids and analysis was conducted using a number of statistical methods. Independent samples t-tests were used to examine size differences between sexes and ancestries, while linear regression analysis and principle component analysis were used to examine shape differences. Discriminant function analysis was employed to test the ability of the hyoids to be classified by sex or ancestry. The ultimate goal of the thesis is to provide physical anthropologists with a series of discriminant function equations that can be used to estimate the sex and ancestry of a hyoid. Five equations ranging in accuracy from 83-88% were developed to determine sex of a hyoid, while four equations ranging in accuracy from 70-89% can be used to determine ancestry. In addition, the t-tests, regression analyses, and principle component analysis have identified several variations in size and shape between sexes and ancestries. These analyses have provided further knowledge as to the morphological form of the hyoid, as well as a method that can be easily used by physical anthropologists to assess sex and ancestry.

A preliminary study of the feasibility of using categorical variables in discriminant function analysis was performed. Data including both continuous and categorical variables were used and predictive results examined. The discriminant function techniques were found to be robust enough to include the use of categorical variables. Some problems were encountered with using the trace criterion for selecting the most discriminating variables when these variables are categorical. No monotonic relationship was found to exist between the trace and the number of correct predictions. This study did show that the use of categorical variables does have much potential as a statistical tool in classification procedures. (50 pages)

In stead of assuming the starting observations as zero, the cointegration test statistic derived in this paper takes the initial conditions into consideration. The statistic helps us understand how the initial conditions affect the test and helps us recognize whether the cointegration relationship exists in the data generation process or not. Beside, with this statistic derived with the concept of discriminant function and residual based test, we can simulate our own critical value table in according to what starting observations we have in hand, what significant level we want and what value of rho we meet under H_1.

With a large number of unidentified skeletal remains found in South Africa, the development of osteometric standards to determine sex from various bones is crucial. It is imperative that a forensic anthropologist have access to a variety of techniques both morphological and metric, which can be used to establish accurate demographic profiles from complete, fragmentary and/or commingled remains. Standards for South African populations are available for the cranium, humerus, pelvis, femur, tibia, calcaneus, and talus. No research has been done on bones of the forearm, even though they are known through international studies to exhibit sexual dimorphism. The purpose of this research was to develop discriminant function formulae to determine sex from the radius and ulna for a South African population. The sample consisted of 200 male and 200 female skeletons from the Pretoria Bone (University of Pretoria) and Raymond A. Dart (Witwatersrand University) collections. Sixteen standard anthropometric measurements were taken from the radius (9) and ulna (7) and subjected to stepwise and direct discriminant function analysis. Distal breadth, minimum midshaft diameter and maximum head diameter were the best discriminators of sex in the radius, while minimum midshaft diameter and olecranon breadth were selected for the ulna. Classification accuracy for the forearm ranged from 76% to 86%. In summary, the radius and ulna can be considered moderate discriminators for determining sex in a South African group. However, it is advised that whenever possible these formulae are used in conjunction with visual methods to determine sex from skeletal remains. Copyright 2007, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. Please cite as follows: Barrier, ILO 2007, Sex determination from the bones of the forearm in a modern South African sample, MSc dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-08052008-090115 / > E961/ag / Dissertation (MSc)--University of Pretoria, 2008. / Anatomy / unrestricted

<p> Application of four statistical techniques of discrimination is made to a set of multivariate data. The techniques, proposed by R.A . Fisher [6], C.R. Rao Q4] , D.F. Andrews [l] and H. Chernoff [4], are reviewed, applied and criticized in an intercomparison of the four methods. Graphic illustrations are also utilized to aid in the classification of sampling units. </p> / Thesis / Master of Science (MSc)

The postsecondary enrollment of student veterans has increased with the troop draw down in Iraq and Afghanistan as well as the generous amendments made to the Post 9/11 GI Bill. Acquired disabilities remain a reality for this population as they transition into the civilian world; consequently, previous literature cites the role of disabilities amongst student veterans. Also, prior research often aggregates these two groups without a thorough understanding of the ways in which they differ. This study compared student veterans with disabilities to student veterans without disabilities in order to understand the enrollment and demographic factors on which they differed, if any. Using a secondary data analysis of the 2007-2008 National Postsecondary Student Aid Survey, univariate tests of significance, a logistic regression, and a discriminant function analysis examined the relationship between disability status and seven predictor variables: age, gender, GPA, major, risk index, degree program type, and whether or not a student was exclusively a distance learner. These seven variables as a whole were not significant predictors of disability status; however, the models provided valuable insight into the similarities and characteristics shared within this population. Univariate tests of significance revealed that students with disabilities had a significantly lower mean GPA, were more often male, tended to favor certain academic majors over others, more often enrolled in bachelor’s degree versus associate and certificate programs, and had a lower risk of attrition based on their index of risk. Major, degree program type, and risk index proved to be the most significant predictors of disability status in LR and DFA. A student veteran’s age and whether they were a distance learner had no significant bearing on disability status indicating that student veterans enroll in distance learning or campus-based programs without influence from an orthopedic or mobility impairment, the most common type of disability amongst student veterans. This study offers a full description of student veterans with disabilities including the specific types of disabilities with which this population enters higher education.

This paper presents a description of gray-headed junco habitat in the form of a model based on discriminant function analysis. Junco nests were found by searching randomly located grids on a 7.8 km2 study area in central Utah. Vegetation data was gathered on 500 m2 circular plots surrounding nests and contrasted with similar data from searched grids where nests were not found. The model explains 28 percent of the between-groups variance and correctly classifies 68 percent of the plots. Plant cover types are good predictors of areas where gray-headed juncos nest, while plant community type descriptions do not distinguish between utilized and unutilized areas. The use of quantitative models of wildlife habitat with data derived from various kinds of information systems is discussed.

Behavioral characterization of animal models for Alzheimer's Disease is critical for the development of potential therapeutics and treatments against the disease. While there are several known animal models of AD, three current models--APPsw, P301L, and APPsw+P301L--have not been well characterized, if at all. This study, therefore, aimed to perform a full behavioral characterization of these three models in order to better understand the impairments associated with each one. Between 5 and 8.5 months of age, animals were behaviorally tested in a variety of sensorimotor, anxiety, and cognitive tasks. The number of tau+ neurons in the forebrains of P301L mice was then compared to their behavioral performance. Results of this study indicate that retinal degeneration (rd) seriously impairs the performance of mice in behavioral tasks. Animals that carry the homozygous allele of this mutation must, therefore, be eliminated from any such study requiring visual acuity. After this elimination, my findings indicate that APP mice are impaired in several cognitive tasks (including platform recognition, Morris maze, Y-maze, and radial-arm water maze) at a young early age (5 to 8.5 months of age). These mice have fairly normal sensorimotor function, showing significant impairment only in balance beam performance starting at 5 months. Although P301L mutant Tau mice, as a group, did not have significant impairments in any sensorimotor or cognitive task, correlation analysis revealed that higher numbers of tau+ neurons in cortex and hippocampus were associated with poorer cognitive performance. Finally, discriminant function analysis (DFA) appears able to accurately discriminate between the three transgenic groups of mice using only an 8-measure data set. In conclusion, this study provides the first comprehensive, multiple task behavioral assessment of the APPsw and P301L animal models of AD, indicating that APPsw mice are cognitively impaired at an early age while P301L mice are largely unimpaired through 8.5 months. Nonetheless, correlational analysis implicates the formation of neurofibrillary tangles in the onset of cognitive impairments. Finally, my findings recommend the continued use of DFA to determine if groups of animals, based on different transgenicity or therapeutic treatment, could be discriminated between from their behavior alone.

AdaBoost.M1 has been successfully applied to improve the accuracy of a learning algorithm for multi-class classification problems. However, it assumes that the performance of each base classifier must be better than 1/2, and this may be hard to achieve in practice for a multi-class problem. A new algorithm called AdaBoost.MK only requiring base classifiers better than a random guessing (1/k) is thus designed. Early SVM-based multi-class classification algorithms work by splitting the original problem into a set of two-class sub-problems. The time and space required by these algorithms are very demanding. In order to have low time and space complexities, we develop a base classifier that integrates one-class SVMs with discriminant functions. In this study, a hybrid method that integrates AdaBoost.MK and one-class SVMs with improved discriminant functions as the base classifiers is proposed to solve a multi-class classification problem. Experimental results on data sets from UCI and Statlog show that the proposed approach outperforms many popular multi-class algorithms including support vector clustering and AdaBoost.M1 with one-class SVMs as the base classifiers.