This study tested the applicability of the Transtheoretical Model of Behavior Change in predicting early termination, appropriate termination, and ongoing treatment of Hispanic battered women residing at domestic violence shelters. Self-efficacy, decisional balance, and acculturation were examined in relation to the applicability of this model with the Hispanic women population. One hundred and eight women residing in two shelters for survivors of domestic violence, located in the Dallas/Fort Worth metroplex, were asked to provide information regarding the problems in their relationships, the pull and the strain of their relationship, their level of temptation to stay in the abusive relationship, and how confident they felt that they would not return to their abuser (The Process of Change in Abused Women Scales- PROCAWS). In addition, the women were asked to complete a questionnaire regarding their level of acculturation. This study confirmed the stage of change profiles found in a population of battered women as well as in other clinical populations and the results suggest that this model is applicable to Hispanic populations. The results indicated that the women in this sample could be meaningfully grouped according to their level of involvement in different stages of change. Furthermore, this study provided support for the validity of this theory by finding significant relationships among the profiles of change and the intervening variables that moderate movement across the stages of change. The women in this study differed with regard to their level of temptation to stay in their relationships and the amount of cons they to making changes. The findings also confirmed that the Transtheoretical Model can be used to predict termination status from domestic violence shelter programs. Although there were no significant differences in termination status among the women with different stage of change profiles, a trend existed that women in earlier stages of change terminated earlier and women in later stages of change terminated appropriately. Overall, the results of this study provide evidence for the applicability of the Transtheoretical Model and the usefulness of the PROCAWS in identifying profiles of change that can potentially guide treatment interventions and predict early termination with the Hispanic population.

The problem of this study was to isolate predictors of academic success in both self-paced classes and lecture classes in Introductory Accounting. The purposes of the study were to determine if learning style, locus of control, reading ability, age, sex, accounting work experience, and prior accounting academic experience are predictors of success in Introductory Accounting classes taught using self-paced methods of instruction and lecture methods of instruction. Another purpose was to determine if there is a difference in the set of predictors of success in the two instructional methods and to provide some direction as to determinants of success which may be addressed by counselors in advising students. The data were collected from 463 students at a suburban community college in the Southwest. Each of the variables was analyzed by a stepwise multiple regression analysis and a backward elimination regression for students grouped according to instructional method. A two-way multivariate analysis of variance was used to examine whether the distribution of scores on the potential predictor variables were equivalent for students in the two teaching methods and for successful completers of the course and noncompleters. Consideration of the data findings of this study permitted the following conclusions: 1. Age and reading ability have a positive relationship to academic success in an Introductory Accounting course taught in a lecture format. 2. Concrete learning style, as measured by the Learning Style Inventory, age, reading ability, and accounting work experience have a positive relationship to success in an Introductory Accounting course taught in a self-paced format. 3. Age, reading ability, accounting work experience, and a concrete learning style have a positive relationship to academic success in Accounting courses taught using either method. 4. There is a difference in the set of predictors of success for Accounting classes taught using the two instructional methods. 5. There are differences between completers and noncompleters of courses regardless of instructional method.

The problem of this study was performance on selected mathematics and reading assessment tests as predictors of achievement in remedial mathematics. The purpose of the study was twofold. The first was to determine the internal consistency of a locally developed remedial mathematics placement test and the mathematics section of the Pre-TASP Test. The second was to determine the predictive validity of performance on (a) the local remedial mathematics placement test, (b) the mathematics section of the Pre-TASP Test, and (c) the Descriptive Tests of Language Skills, Reading Comprehension Test in combination with demographic variables for mid-semester achievement, end-of-semester achievement, and course success in three levels of remedial mathematics at Richland College, Dallas, Texas.

This study re-examined a previously validated application blank in use for 1 year to screen applicants for the position of equipment operator with a company involved in hydrocarbon recovery. Subjects were 409 male equipment operators ranging in age from 19 to 38 years. Minorities accounted for 12% of the group, while 88% were white. Subjects were randomly divided into an even group, N = 201, and an odd group, N = 208. Multiple R's of .39 were obtained for the most significant 10 variables in each group, but these shrank considerably during cross-validation. Only 3 variables were common to both groups since the unique error variances for each group resulted in different arrangements of variables. It was concluded that the items should be re-examined for relevancy and job relatedness.

<p>Air pollution and a general concern for lack of physical activity in North America have motivated governments to encourage non-motorized modes of transportation. A key infrastructure component for these forms of transportation is sidewalks. The City of Saskatoon has identified the need to formalize sidewalk management policies to demonstrate diligence for community protection regarding sidewalk safety. Prioritization of sidewalk maintenance and rehabilitation actions must be objective and minimize risk to the community. Most research on prioritization of pedestrian facilities involved new construction projects. This research proposes a decision model that prioritizes a given list of existing unsafe sidewalk locations needing maintenance or rehabilitation using a direct measure of pedestrian safety, namely, quality-adjusted life years lost per year. </p><p>A decision model was developed for prioritizing a given list of unsafe sidewalk locations, aiding maintenance and rehabilitation decisions by providing the associated risk to pedestrian safety. The model used data mostly from high quality sources that had already been collected and validated. Probabilities and estimations were used to produce value-added decision policy.</p> <p>The decision analysis framework applied probability and multi-attribute utility theories. This study differed from other research due to the inclusion of age and gender groups. Total average daily population of the city was estimated. This population was distributed to sidewalk locations using probabilities for trip purposes and a locations ability to attract people relative to the city total. Then trip injury events were predicted. Age and gender distribution and trip injury type estimations were used to determine the impact of those injuries on quality of life.</p><p>There exist much observable high quality data that can be used as indicators of unknown or unobserved events. A decision policy was developed that prioritizes unsafe sidewalk locations based on the direct safety impact on pedestrians. Results showed that quality-adjusted life years lost per year sufficiently prioritized a given list of unsafe sidewalk locations. It was demonstrated that the use of conditional probabilities (n=594) allowed for the ability to abstract data representing a different source population to another. Average daily population confined and distributed within the city boundary minimized problems of accuracy. Gender-age distribution was important for differentiating the risk at unsafe sidewalk locations. Concepts from this research provide for possible extension to the development of sidewalk service levels and sidewalk priority maps and for risk assessment of other public services.</p>

A large number and variety of genome-wide genomics and proteomics datasets are now available for model organisms. Each dataset on its own presents a distinct but noisy view of cellular state. However, collectively, these datasets embody a more comprehensive view of cell function. This motivates the prediction of function for uncharacterized genes by combining multiple datasets, in order to exploit the associations between such genes and genes of known function--all in a query-specific fashion. Commonly, heterogeneous datasets are represented as networks in order to facilitate their combination. Here, I show that it is possible to accurately predict gene function in seconds by combining multiple large-scale networks. This facilitates function prediction on-demand, allowing users to take advantage of the persistent improvement and proliferation of genomics and proteomics datasets and continuously make up-to-date predictions for large genomes such as humans. Our algorithm, GeneMANIA, uses constrained linear regression to combine multiple association networks and uses label propagation to make predictions from the combined network. I introduce extensions that result in improved predictions when the number of labeled examples for training is limited, or when an ontological structure describing a hierarchy of gene function categorization scheme is available. Further, motivated by our empirical observations on predicting node labels for general networks, I propose a new label propagation algorithm that exploits common properties of real-world networks to increase both the speed and accuracy of our predictions.

A large number and variety of genome-wide genomics and proteomics datasets are now available for model organisms. Each dataset on its own presents a distinct but noisy view of cellular state. However, collectively, these datasets embody a more comprehensive view of cell function. This motivates the prediction of function for uncharacterized genes by combining multiple datasets, in order to exploit the associations between such genes and genes of known function--all in a query-specific fashion. Commonly, heterogeneous datasets are represented as networks in order to facilitate their combination. Here, I show that it is possible to accurately predict gene function in seconds by combining multiple large-scale networks. This facilitates function prediction on-demand, allowing users to take advantage of the persistent improvement and proliferation of genomics and proteomics datasets and continuously make up-to-date predictions for large genomes such as humans. Our algorithm, GeneMANIA, uses constrained linear regression to combine multiple association networks and uses label propagation to make predictions from the combined network. I introduce extensions that result in improved predictions when the number of labeled examples for training is limited, or when an ontological structure describing a hierarchy of gene function categorization scheme is available. Further, motivated by our empirical observations on predicting node labels for general networks, I propose a new label propagation algorithm that exploits common properties of real-world networks to increase both the speed and accuracy of our predictions.

<p>Air pollution and a general concern for lack of physical activity in North America have motivated governments to encourage non-motorized modes of transportation. A key infrastructure component for these forms of transportation is sidewalks. The City of Saskatoon has identified the need to formalize sidewalk management policies to demonstrate diligence for community protection regarding sidewalk safety. Prioritization of sidewalk maintenance and rehabilitation actions must be objective and minimize risk to the community. Most research on prioritization of pedestrian facilities involved new construction projects. This research proposes a decision model that prioritizes a given list of existing unsafe sidewalk locations needing maintenance or rehabilitation using a direct measure of pedestrian safety, namely, quality-adjusted life years lost per year. </p><p>A decision model was developed for prioritizing a given list of unsafe sidewalk locations, aiding maintenance and rehabilitation decisions by providing the associated risk to pedestrian safety. The model used data mostly from high quality sources that had already been collected and validated. Probabilities and estimations were used to produce value-added decision policy.</p> <p>The decision analysis framework applied probability and multi-attribute utility theories. This study differed from other research due to the inclusion of age and gender groups. Total average daily population of the city was estimated. This population was distributed to sidewalk locations using probabilities for trip purposes and a locations ability to attract people relative to the city total. Then trip injury events were predicted. Age and gender distribution and trip injury type estimations were used to determine the impact of those injuries on quality of life.</p><p>There exist much observable high quality data that can be used as indicators of unknown or unobserved events. A decision policy was developed that prioritizes unsafe sidewalk locations based on the direct safety impact on pedestrians. Results showed that quality-adjusted life years lost per year sufficiently prioritized a given list of unsafe sidewalk locations. It was demonstrated that the use of conditional probabilities (n=594) allowed for the ability to abstract data representing a different source population to another. Average daily population confined and distributed within the city boundary minimized problems of accuracy. Gender-age distribution was important for differentiating the risk at unsafe sidewalk locations. Concepts from this research provide for possible extension to the development of sidewalk service levels and sidewalk priority maps and for risk assessment of other public services.</p>

This research aims at providing methods in analyzing and estimating the taxi times of aircraft at airports, which are expected to be an important element for reducing taxiing delay and consequent excess fuel consumption and environmental costs. The proposed model involves a set of regression equations to model the taxi-out and taxi-in times at airports. The estimated results can be used to calculate the nominal taxi times, which are essential measures for evaluating the taxiing delays at airports. Given the outcomes of the regression model, an iterative algorithm is developed to predict taxi times. A case study at LGA shows that the proposed algorithm demonstrates higher accuracy in comparison to other algorithms in existing literature.

An increasing number of people are turning to online blogging communities devoted to self-change for smoking, shopping, and other behaviors. To understand processes underlying effective self-change, the current project tracked the language and social dynamics of a dieting blog community using computerized text analysis. Three research questions were asked: What predicts weight loss in blogs? What changes in blogging predict weight loss? Can we predict dropping out or successful weight loss based on the first two entries? A community of blogs devoted to weight loss was examined (n = 2530). Most bloggers were female, and on average, around 30 years old, approximately 200 pounds, with a goal weight of about 140 pounds. A sample of blogs by females that had blogged at least 15 entries within the first 15 weeks of blogging resulted in a total of 186 blogs, representing over 9,200 entries for analysis. Computerized text analysis was used to examine language for rates of self-focus, emotionality, cognitive processing, keeping food diaries, and social support. Rates of blogging were assessed by word counts, number of active weeks, and mean entries per week. Social support was assessed through the use of social words, the size of the social network, along with the positivity and negativity of the comments. The discrepancy between start and goal weight was also assessed. The results suggested that having larger weight loss goals and blogging about personal events was a more effective weight loss strategy than keeping an online food intake diary. The degree to which bloggers were socially integrated with the blog community was found to be a potent predictor of weight loss. Online components of behavioral treatment programs could encourage dieters to browse and comment on other dieters’ progress, and to share personal narratives rather than simply focusing on the benefits of food intake diaries, nutrition, and exercise. The current project points to the power of computerized text analytic tools to address important theoretical and practical social psychological issues that are evolving on the internet. Specifically, language analysis methods can identify which dimensions of blogging communities can help or hinder self-change processes. / text