The present study was conducted to further our understanding of the correlates of and variations in perceptions of police treatment among African American adolescent offenders. Ethnic identity development can play a role in youths' sensitivity to stigma, but whether this finding applies to black youth involved in the juvenile justice system has not been explored. Although there is evidence for a robust association between perceptions of discrimination and negative psychological outcomes, there is a dearth of research that investigates a) the directional nature of these associations, and b) how associations vary as a function of perceptions of personal and group discrimination. Participants were 501 African American youth ages 14-18 who were adjudicated of a felony or serious misdemeanor in Philadelphia. Data were taken from annual interviews conducted over the course of four years. Increased ethnic identity exploration was related to the perception that police use biased behavior against people from different racial/ethnic backgrounds. Furthermore, there was a relatively stronger association between psychological distress and perceptions of police behavior among youth who reported taking an active role in making meaning of their ethnicity,. Finally, the results of this study support drawing a distinction between personal and global perceptions of discrimination, in that their links to psychological distress differed with respect to the direction of effects. Specifically, whereas negative personal encounters with the police lead to higher levels of distress, being distressed led to more negative global perceptions of the police. This study provides evidence that normative processes in adolescence, like ethnic identity development, operate much the same way among high risk youth (e.g., juvenile offenders) as in more normative samples. This is especially important given that the consideration of normative developmental processes in high-risk samples like juvenile offenders can have implications for rehabilitation efforts. Finally, the present research highlights the need for the education of law enforcement agencies regarding adolescent development and factors that might increase or decrease young people's willingness to comply with the law. / Psychology

The most recent research trend within the field of organizational justice is the study of interactional justice (Tyler & Bies, 1990; Greenberg, 1990). Most of the work conducted in this area focuses on the explanations leaders give to followers about decisions made or actions taken that frequently lead to adverse consequences or loss. These explanations are called "social accounts". While research indicates that social accounts are effective at mitigating negative reactions to adverse decisions, and improving perceptions of justice, the underlying causes remain largely unknown. This study used the Elaboration Likelihood Model (ELM) as a theoretical base to explore what factors make an account "acceptable", and the impact that an acceptable account has on a number of organizationally relevant outcomes. Specifically, this study used an experimental design in a pay-for-performance context to assess the impact of message specificity (high vs. low) and source expertise (expert vs. non-expert), under varying conditions of outcome involvement (higher vs. lower) on the acceptability of a justification for a change in a distributive criterion that resulted in a loss. Other dependent variables investigated included perceptions of justice (procedural, interactional, and distributive fairness), attitudes (satisfaction with the trainer and the task), and behaviors (commitment to the trainer and the task, and complaints). The results indicated that justifications delivered by the expert trainer were more acceptable than non-experts under conditions of lower involvement (except when a non-expert delivered a specific justification). This effect reversed itself, however, under higher involvement. There was a main effect of specificity such that specific justifications were more acceptable than vague justifications regardless of the level of involvement. Acceptability was positively related to all dependent measures with the exception of complaints, which had a negative relationship. Furthermore, acceptability fully mediated the impact of message specificity on trainer commitment, and partially on procedural fairness, distributive fairness (assessment), and task commitment. Acceptability fully mediated the negative impact of expertise under high involvement on procedural and distributive fairness (assessment), and interactional fairness. / Ph. D.

This dissertation proposes the Weaves runtime framework for the execution of large scale parallel programs over lightweight intra-process threads. The goal of the Weaves framework is to help process-based legacy parallel programs exploit the scalability of threads without any modifications. The framework separates global variables used by identical, but independent, threads of legacy parallel programs without resorting to thread-based re-programming. At the same time, it also facilitates low-overhead collaboration among threads of a legacy parallel program through multi-granular selective sharing of global variables. Applications that follow the tenets of the Weaves framework can load multiple identical, but independent, copies of arbitrary object files within a single process. They can compose the runtime images of these object files in graph-like ways and run intra-process threads through them to realize various degrees of multi-granular selective sharing or separation of global variables among the threads. Using direct runtime control over the resolution of individual references to functions and variables, they can also manipulate program composition at fine granularities. Most importantly, the Weaves framework does not entail any modifications to either the source codes or the native codes of the object files. The framework is completely transparent. Results from experiments with a real-world process-based parallel application show that the framework can correctly execute a thousand parallel threads containing non-threadsafe global variables on a single machine - nearly twice as many as the traditional process-based approach can - without any code modifications. On increasing the number of machines, the application experiences super-linear speedup, which illustrates scalability. Results from another similar application, chosen from a different software area to emphasize the breadth of this research, show that the framework's facilities for low-overhead collaboration among parallel threads allows for significantly greater scales of achievable parallelism than technologies for inter-process collaboration allow. Ultimately, larger scales of parallelism enable more accurate software modeling of real-world parallel systems, such as computer networks and multi-physics natural phenomena. / Ph. D.

In evaluating clientsâ satisfaction with their lawyers, some research indicates that clients consider the interpersonal aspect of the lawyer-client relationship just as important as the legal competence of the lawyer. The purpose of this study is to assess factors hypothesized to be involved in clients* evaluations of lawyers and the legal system. These factors are ratings of procedural justice, perceptions of lawyers interactional styles and types of social power, and clientsf demographic and legal case characteristics. Data for the quantitative analysis in this study comes from a national telephone survey of legal clients. In addition, qualitative data from responses to an open-ended question in the telephone survey and focus groups of legal clients in divorce cases are analyzed. The results of the multiple regression and path analyses reveal that lawyers’ interactional orientation (combinations of social power and adoption of occupational role) nave the greatest effect on ratings of procedural justice. Ratings of procedural justice and lawyers’ interactional orientation have the greatest effect on satisfaction with lawyers. Smaller effects come from the outcome and the type of legal case. Satisfaction with the attorney has a greater effect on satisfaction with the courts for women than it does for men. Ratings of procedural justice and lawyer’s interactional styles have a large effect on satisfaction with the courts for both men and women. Comments by the survey respondents and the focus group participants support previous research that the major sources of dissatisfaction with lawyers are fees, discourtesy, and delays. Issues which coincide with elements of the rating of procedural justice emerged from the focus groups. The components of representation, quality and accuracy, and respect and concern for the client were brought up in various ways. A suggestion is made for the use of the components of procedural justice as a guideline or checklist for lawyers and clients as a way of improving the image of lawyers and increasing clients’ consumer power. / Ph. D.

The present study explored the direct influences of teacher classroom practices, including teacher support, conceptual teaching, and procedural teaching, on 9th grade students' mathematics achievement, and the indirect influences of these teacher variables on student mathematics achievement through students' mathematics self-efficacy and interest in mathematics courses. The base year data of High School Longitudinal Study of 2009 (HSLS: 09) was used for this study. Structural equation modeling method was used to estimate the relationships among variables. Results showed that teacher classroom practices influenced student mathematics achievement in different ways. Conceptual teaching positively, whereas procedural teaching negatively, influenced student mathematics achievement. Teacher support influenced student mathematics achievement indirectly through students' mathematics self-efficacy. It also had powerful influence on students' interest in mathematics courses. In addition, family socioeconomic status (SES) and student prior achievement were associated with teacher classroom practices. Students with higher levels of family SES and prior achievement were more likely to have teachers who use conceptual teaching strategies. Students with higher prior achievement were more likely to perceive higher levels of teacher support. / Ph. D.

Deep learning models can perform many tasks very capably, provided they are trained correctly. Usually, this requires a large amount of data. Pre-training refers to a process of creating a strong initial model by first training it on a large-scale dataset. Such a model can then be adapted to many different tasks, while only requiring a comparatively small amount of task-specific training data. Pre-training is the standard approach in most computer vision scenarios, but it's not without drawbacks. Aside from the cost and effort involved in collecting large pre-training datasets, such data may also contain unwanted biases, violations of privacy, inappropriate content, or copyright material used without permission. Such issues can lead to concerns about the ethical use of models trained using the data. This dissertation addresses a different approach to pre-training visual models by using abstract, procedurally generated data. Such data is free from the concerns around human bias, privacy, and intellectual property. It also has the potential to scale more easily, and provide precisely controllable sources of supervision that are difficult or impossible to extract from data collected in-the-wild from sources like the internet. The obvious disadvantage of such data is that it doesn't model real-world semantics, and thus introduces a large domain-gap. Surprisingly, however, such pre-training can lead to performance not far below models trained in the conventional way. This is shown for different visual recognition tasks, models, and procedural data-generation processes.

The number of Ombudsman offices in U.S. federal agencies rose dramatically in the 1990s. This study investigates why, despite the efforts of policymakers to force staff reductions across the federal government, Ombudsman offices continued to be established to the point that almost every agency has an Ombudsman. This study uses neo-institutionalist theory to pinpoint indicators that explain what has triggered the proliferation of external facing Ombudsmen in the federal government. The results of this historical retrospective investigation, which uses a mixed methods approach, indicate that the offices were created to ensure procedural justice and as a response to both: stakeholder pressures (since the population became more vocal and active, demanding access to the government, transparency, and accountability) and congressional mandates (such as the Administrative Dispute Resolution Act (ADRA) of 1990 and 1996 the Alternative Dispute Resolution Acts from 1998). This is consistent with neo-institutionalist expectations that organizations change as the result of pressures from forces in the environment combined with the drive for survival. As change accelerates, isomorphism occurs as organizations and agencies adopt strategies that have worked for other similar organizations in their environment. / Ph. D. / This study investigates why, despite the efforts of policymakers to force staff reductions across the federal government, external facing Ombudsman offices continued to be established in the U.S. federal government, to the point that almost every agency has an Ombudsman. This study uses neo-institutionalist theory to pinpoint indicators that explain what has triggered the proliferation of these entities. The results of this historical retrospective investigation indicate that the offices were created to ensure procedural justice and as a response to both: stakeholder pressures (since the population became more vocal and active, demanding access to the government, transparency, and accountability) and congressional mandates (such as the Administrative Dispute Resolution Act (ADRA) of 1990 and 1996 the Alternative Dispute Resolution Acts from 1998).

Herein I document my exploration into the intersection of convolutional neural networks and raw non-lexical audio signals by detailing the development and results of four projects, each representing a unique problem in this domain: mutation detection, upscaling, classification, and generation. Convolutional neural networks, within the class of computational models which approximate a functional relationship between spaces of data expressed through a bio-inspired structure of modular interconnected neural nodes, are a subcategory suited to data with features that are spatially correlated while variable in absolute position. Dilated convolutional neural networks are of particular interest for operating on audio signals, as the exponential dilation stack both greatly expands the receptive field and extracts features at a progression which reflects the logarithmic properties of human hearing. More generally, I seek to study at a granular level the application of convolutional neural networks to any discrete temporal signals with dense periodic features, though the primary focus is on music and components of audio composition for music and video game production.

Measuring procedural fidelity helps behavior analysts assess staff performance and make decisions about a treatment's effectiveness. However, there is limited research comparing different methods of measuring procedural fidelity. Therefore, we systematically replicated a study that examined how procedural fidelity of discrete-trial-instruction sessions varies depending on the measurement system and calculation. We scored videos depicting behavior technicians conducting discrete-trial instruction with a child with autism and did not permit data collectors to pause, rewind, or fast-forward to simulate a live session. Specifically, we compared individual-component and global fidelity from the occurrence/nonoccurrence method and all-or-nothing by component method. Like previous research has demonstrated, our findings indicated that global values were likely to mask component errors, whereas the all-or-nothing by component method was less likely to do so. When data collected in the current study was compared to those in previous research, interobserver agreement was high for both occurrence/nonoccurrence and all-or-nothing by component methods. Overall, the procedural-fidelity values obtained from each study differed minimally. These results suggest that both measurement methods yield consistent results whether derived from recorded sessions with the ability to pause, rewind, and fast-forward or simulated live sessions. Differences that were obtained are discussed in detail with particular emphasis on connecting our findings to the utility of procedural-fidelity assessment in practice.

No / This study aims to investigate whether, and how, the use of performance measures, procedural justice, and interpersonal trust interact to affect organisational commitment of police officers. Drawing on a survey based on a sample of 57 senior officers within a single police force, we find that the use of performance measures, procedural justice, and interpersonal trust are positively associated with organisational commitment. Further analysis reveals that trust between officer and superior mediates the relationship between the use of non-financial measures and organisational commitment, but procedural justice does not have a mediating effect on commitment. These findings are further explored through selected interviews with respondents.