Comparison of Some Improved Estimators for Linear Regression Model under Different Conditions

Multiple linear regression model plays a key role in statistical inference and it has extensive applications in business, environmental, physical and social sciences. Multicollinearity has been a considerable problem in multiple regression analysis. When the regressor variables are multicollinear, it becomes difficult to make precise statistical inferences about the regression coefficients. There are some statistical methods that can be used, which are discussed in this thesis are ridge regression, Liu, two parameter biased and LASSO estimators. Firstly, an analytical comparison on the basis of risk was made among ridge, Liu and LASSO estimators under orthonormal regression model. I found that LASSO dominates least squares, ridge and Liu estimators over a significant portion of the parameter space for large dimension. Secondly, a simulation study was conducted to compare performance of ridge, Liu and two parameter biased estimator by their mean squared error criterion. I found that two parameter biased estimator performs better than its corresponding ridge regression estimator. Overall, Liu estimator performs better than both ridge and two parameter biased estimator.

GIS-integrated mathematical modeling of social phenomena at macro- and micro- levels—a multivariate geographically-weighted regression model for identifying locations vulnerable to hosting terrorist safe-houses: France as case study

Adaptability and invisibility are hallmarks of modern terrorism, and keeping pace with its dynamic nature presents a serious challenge for societies throughout the world. Innovations in computer science have incorporated applied mathematics to develop a wide array of predictive models to support the variety of approaches to counterterrorism. Predictive models are usually designed to forecast the location of attacks. Although this may protect individual structures or locations, it does not reduce the threat—it merely changes the target. While predictive models dedicated to events or social relationships receive much attention where the mathematical and social science communities intersect, models dedicated to terrorist locations such as safe-houses (rather than their targets or training sites) are rare and possibly nonexistent. At the time of this research, there were no publically available models designed to predict locations where violent extremists are likely to reside. This research uses France as a case study to present a complex systems model that incorporates multiple quantitative, qualitative and geospatial variables that differ in terms of scale, weight, and type. Though many of these variables are recognized by specialists in security studies, there remains controversy with respect to their relative importance, degree of interaction, and interdependence. Additionally, some of the variables proposed in this research are not generally recognized as drivers, yet they warrant examination based on their potential role within a complex system. This research tested multiple regression models and determined that geographically-weighted regression analysis produced the most accurate result to accommodate non-stationary coefficient behavior, demonstrating that geographic variables are critical to understanding and predicting the phenomenon of terrorism. This dissertation presents a flexible prototypical model that can be refined and applied to other regions to inform stakeholders such as policy-makers and law enforcement in their efforts to improve national security and enhance quality-of-life.

Effectiveness of peer counseling on high school students who failed two or more classes in a nine week quarter

The purpose of this study was to examine the effectiveness of peer counseling on high school students who have previously failed two or more classes in a nine week quarter.^ This study was constructed by comparing students who previously failed and were subsequently given peer counseling with a matched group of students who failed and did not receive peer counseling.^ To test the proposed research question, 324 students from a large urban school system were randomly chosen from a computer generated list of students who failed courses, matched on variables of number of classes failed, grade level and gender. One student from each matched pair was randomly placed in either the experimental or control group. 162 students from Group 1 (experimental) were assigned a peer counselor with their pair assigned to Group 2 (control). Group 1 received peer counseling at least 4 times during the third nine week academic quarter (Quarter 3) while Group 2 did not.^ The Grade Point Averages (GPA) for all students were collected both at the end of Quarter 2 and Quarter 3, at which time peer counseling was terminated. GPA's were also collected nine weeks after counseling was terminated.^ Results were determined by multiple regression, analysis of covariance and t-test. A level of significance was set at.05. There was significant increase in the GPA's of those counseled students immediately after peer counseling and also nine weeks after counseling was terminated, while the group not receiving peer counseling showed no increase. It was noted that there were significantly more school drop-outs from the non-counseled group than the counseled group. The number of classes failed, high school attended, grade level and gender were not found to be significant.^ The conclusion from this study was that peer counseling does impact significantly on the GPA's of students experiencing academic failure. Recommendations from this study were to implement and expand peer counseling programs with more failing students, continue counseling for longer than one quarter, include students in drop-out programs and students from different socio-economic and racial backgrounds, and conduct subsequent evaluation. ^

The influence of perception of forgetting and gender, ethnicity, and closeness on likelihood of caregiving

As life expectancy increases, the population of older adults is increasing rapidly. The caregiving of older parents by adult children has become a normative experience. Much of the gerontological literature has examined the caregiving experience, particularly in terms of the stresses involved. However, research is only beginning to examine the factors which motivate adult children to begin caregiving. The research described here examined how an elderly parent's memory behavior might influence caregiving decisions. In addition, gender, ethnicity, and parent-adult child closeness were examined to explore how these individual difference variables might influence those caregiving decisions.^ Participants read one of two vignettes describing a social visit with an elderly widow (target). In the vignette, the elderly target experiences several instances of forgetting. The vignettes depicted forgetting behavior established in pilot work as normal or serious. The normal forgetting vignette did not arouse concern and the serious forgetting vignette did arouse concern when the middle-aged participants imagined their mothers in the role of the vignette target. Participants rated their likelihood of engaging in eight caregiving behaviors if their mothers behaved like the vignette target. They also rated their closeness with their own mothers.^ Multivariate analyses of variance indicated main effects for vignette type, gender, ethnicity, and attachment. The likelihood of caregiving was higher when forgetting was more serious and when participants were female, Hispanic, and were highly attached to their mothers. Interaction effects showed that gender differences decreased with increased seriousness of forgetting, and ethnic differences were only significant for the normal forgetting condition.^ Multiple regression analyses indicated that attachment was the most significant predictor of likelihood of caregiving. Gender and ethnicity predicted specific caregiving behaviors. Females were more likely to engage in phoning and cooking, and Hispanics were more likely to engage in visiting and suggesting mother move in. ^

Consequences of interorganizational information systems: An empirical investigation of the effects of EDI use on interorganizational relationships in the retail industry

This dissertation examines the consequences of Electronic Data Interchange (EDI) use on interorganizational relations (IR) in the retail industry. EDI is a type of interorganizational information system that facilitates the exchange of business documents in structured, machine processable form. The research model links EDI use and three IR dimensions--structural, behavioral, and outcome. Based on relevant literature from organizational theory and marketing channels, fourteen hypotheses were proposed for the relationships among EDI use and the three IR dimensions.^ Data were collected through self-administered questionnaires from key informants in 97 retail companies (19% response rate). The hypotheses were tested using multiple regression analysis. The analysis supports the following hypothesis: (a) EDI use is positively related to information intensity and formalization, (b) formalization is positively related to cooperation, (c) information intensity is positively related to cooperation, (d) conflict is negatively related to performance and satisfaction, (e) cooperation is positively related to performance, and (f) performance is positively related to satisfaction. The results support the general premise of the model that the relationship between EDI use and satisfaction among channel members has to be viewed within an interorganizational context.^ Research on EDI is still in a nascent stage. By identifying and testing relevant interorganizational variables, this study offers insights for practitioners managing boundary-spanning activities in organizations using or planning to use EDI. Further, the thesis provides avenues for future research aimed at understanding the consequences of this interorganizational information technology. ^

Factors associated with social work student satisfaction with field placements in child welfare

This study was conducted during the 1994-1995 academic year. Seven social work education programs in the state of Florida, all accredited by the Council on Social Work Education, participated in this study. Graduate and undergraduate social work students in child welfare field placements, and their field instructors, were surveyed during the Spring 1995 semester to assess their satisfaction with field placements in this area and the relationship of this satisfaction to employment interests and field placement recommendations.^ The majority of social work students responding to this survey were generally satisfied with several aspects of their field placements--the learning, field work program, field instructor, child welfare agency, and overall field experience. The field instructors were generally more satisfied than the students, but only statistically different from the students in the areas of satisfaction with the field work program and the child welfare agency. Multiple regression analysis revealed that learning assignment opportunities, field instructor relationship characteristics, placement preference, and pre-placement interview contributed to the prediction of student satisfaction.^ Student satisfaction in field placement was significantly related to the acceptance of employment, if offered, and the recommendation of the field placement to other students. Logistic regression analysis revealed that satisfaction with the child welfare agency was the greatest contributor to the prediction of acceptance of employment, and satisfaction with the field work program was the greatest contributor to the prediction of field placement recommendation. ^

A study of the relationship between parental self-efficacy and indicators of school involvement exhibited by Hispanic parents of elementary school children

The purpose of this study was to examine the relationship between Hispanic parents' sense of self-efficacy at various degrees of acculturation to the United States and specific indicators of school involvement in their elementary school children's education. It assessed the effects of acculturation on the level of parental self-efficacy and their degree of school involvement. The theoretical framework guiding this investigation was Bandura's theory of self-efficacy which advocates that the amount of effort a person devotes to the accomplishment of a specific outcome is related to a person's beliefs in their capabilities regardless of actual competencies.^ The research method involved a correlational design measuring levels of parental self-efficacy, acculturation, degree of school involvement and related demographic characteristics. Multiple regression analysis was performed to determine the degree of relationships existing between the predictor variables of self-efficacy and level of acculturation, and level of school involvement. The data was subjected to a path analysis to test the validity of the causal model advanced in this study specifying a positive relationship between the constructs of acculturation, parental self-efficacy and level of school involvement.^ A total of 109 Hispanic parents of students enrolled in five elementary public schools in Dade County, Florida, were selected for participation in the study. Results revealed a significant positive correlation r =.23, p $<$.05 between level of parental self-efficacy and number of hours parents spent helping their children with homework. A statistically significant positive correlation was found between acculturation and level of self-efficacy r =.21, p $<$.05. Statistically significant positive correlations were also found between acculturation and such indicators of parental school involvement as participation in parent-teacher conferences r =.20, p $<$.05, volunteering at school, r =.22, p $<$.05, attendance at school sponsored sports activities r =.26, p $<$.01, and volunteering in field-trips r =.28, p $<$.01.^ The multiple regression analysis equation predicting level of homework assistance provided by parents and self-efficacy was statistically significant, F(2,106) = 3.59, p $<$.03. The beta weights revealed that self-efficacy contributed the most to the prediction of homework assistance by parents, B =.258, p $<$.009. In turn, the variable of acculturation was the most significant predictor of number of school based parent involvement activities, B =.281, p $<$.05 level. The path analysis confirmed the results obtained in the multiple regression analyses, establishing self-efficacy as having a direct effect on the level of homework assistance provided by parents. Conversely, the variable of acculturation had a direct effect on the number of school based parent involvement activities. ^

A study of the relationship between U.S. hotel companies and travel agency companies: Determinants of the success of the relationship

The primary objectives of this study were to describe the relationship that U.S. hotel companies have with travel agency companies and to identify the determinants of a successful relationship.^ The unit of analysis was hotel companies operating in U.S. One hundred and three hotel companies contributed information for this research. The data were collected through a questionnaire developed from previous empirical studies and from interviews with hospitality management professors and hotel sales and marketing directors.^ Simple and multiple regression analyses indicated that in order to have successful relationships with travel agency companies, hotel companies should (a) show more commitment or dedication to working with the travel agents, (b) have more trust in the travel agents themselves, and (c) be less dependent on any one travel agent for their business.^ Additionally, the results suggest that hotel companies who coordinate activities with travel agents companies and who communicate with them in a timely, accurate, adequate, complete and credible manner have more successful interorganizational relationships than those who do not.^ Furthermore, hotel companies who share proprietary sales and any other information with travel agents reported better relationships. (Abstract shortened by UMI.) ^

The influence of weak parental and peer attachment on academic achievement among late adolescent college students

The current study examined the influence of weak parental and peer attachment on academic achievement among late adolescent college students. In previous research, weak attachment to parents and/or peers had been found to have an adverse influence on the academic success of college students. This study also examined the potential moderating influence of several cognitive and non-cognitive psychosocial variables that might act as protective factors for weakly attached students and, therefore, enhance their academic competence. Data regarding attachment, cognitive variables, and non-cognitive variables were collected using several self-report measures. The multi-ethnic sample of students in this study (n = 357) attended an urban university. Students were classified into one of nine parental-peer attachment groups (e.g., Low-Low, Medium-Medium, High-High). Attachment groups were compared in terms of cognitive and non-cognitive variables. Contrary to the hypothesis, no statistically significant academic achievement differences were revealed for the group of college students who perceived themselves to be weakly attached to both parents and peers. Analysis of variance (ANOVA) identified the High-High group to be significantly different in terms of academic outcome variables from the other eight groups while the Low-Low group had significantly lower levels of non-cognitive variables than several of the other attachment groups. Hierarchical multiple regression analyses revealed that cognitive variables accounted for significant amounts of variance in academic outcomes and that several non-cognitive variables were significant predictors of scholastic competence. Correlational analyses revealed that parental and peer attachment were positively correlated with several cognitive and non-cognitive variables but neither was significantly correlated with self-reported college GPA. In general, the findings do not provide support for a main effect of weak attachment to parents and peers upon academic adversity among college students. Results suggest that both cognitive variables and non-cognitive variables may moderate academic risk due to weak attachment to parents and peers. Descriptive within group analyses of the Low-Low group revealed a heterogeneous group of students with regards to academic outcomes and scores on non-cognitive measures. Gender and ethnic differences were found for attachment status but not for cognitive or non-cognitive variables. Implications for interventions and suggestions for future research are presented. ^

The relationship between the teacher's experience, the teacher's college major, and the teacher's level of education in predicting classroom attitudes in high school science students

A review of the literature reveals few research has attempted to demonstrate if a relationship exists between the type of teacher training a science teacher has received and the perceived attitudes of his/her students. Some of the teacher preparation factors examined in this study include the college major chosen by the science teacher, the highest degree earned, the number of years of teaching experience, the type of science course taught, and the grade level taught by the teacher. This study examined how the various factors mentioned, could influence the behaviors which are characteristic of the teacher, and how these behaviors could be reflective in the classroom environment experienced by the students.^ The instrument used in the study was the Classroom Environment Scale (CES), Real Form. The measured classroom environment was broken down into three separate dimensions, with three components within each dimension in the CES. Multiple Regression statistical analyses examined how components of the teachers' education influenced the perceived dimensions of the classroom environment from the students.^ The study occurred in Miami-Dade County Florida, with a predominantly urban high school student population. There were 40 secondary science teachers involved, each with an average of 30 students. The total number of students sampled in the study was 1200. The teachers who participated in the study taught the entire range of secondary science courses offered at this large school district. All teachers were selected by the researcher so that a balance would occur in the sample between teachers who were education major versus science major. Additionally, the researcher selected teachers so that a balance occurred in regards to the different levels of college degrees earned among those involved in the study.^ Several research questions sought to determine if there was significant difference between the type of the educational background obtained by secondary science teachers and the students' perception of the classroom environment. Other research questions sought to determine if there were significant differences in the students' perceptions of the classroom environment for secondary science teachers who taught biological content, or non-biological content sciences. An additional research question sought to evaluate if the grade level taught would affect the students' perception of the classroom environment. (Abstract shortened by UMI.) ^

Academic performance of college students in financial management and related prerequisite courses: A comparison of senior high school and commercial high school graduates attending Tamsui Oxford University College in Taiwan

In Taiwan, the college freshmen are recruited graduates of both senior high school and senior vocational school. The Ministry of Education (MOE) of the Republic of China prescribes the standards of curriculum and equipment for schools at all levels and categories. There exists a considerably different curriculum arrangement for senior high schools and vocational high schools in Taiwan at the present time. The present study used a causal-comparative research design to identify the influences of different post-secondary educational background on specialized course performance of college business majors. ^ The students involved in this study were limited to the students of four business-related departments at Tamsui Oxford University College in Taiwan. Students were assigned to comparison groups based on their post-secondary educational background as senior high school graduates and commercial high school graduates. The analysis of this study included a comparison of students' performance on lower level courses and a comparison of students' performance in financial management. The analysis also considered the relationship between the students' performance in financial management and its related prerequisite courses. The Kolb Learning Style Inventory (LSI) survey was administered to categorize subjects' learning styles and to compare the learning styles between the two groups in this study. The applied statistical methods included t-test, correlation, multiple regression, and Chi-square. ^ The findings of this study indicated that there were significant differences between the commercial high school graduates and the senior high school graduates on academic performances in specialized courses but not in general courses. There were no significant differences in learning styles between the two groups. These findings lead to the conclusion that business majors' academic performance in specialized courses were influenced by their post-secondary educational background. ^

Risk factors that predict maladaptive outcomes among sexually abused girls presenting for services at a community agency

The purpose of this study was to determine which factors predicted maladaptive outcomes in sexually abused children. Key factors were aggregated into four categories: abuse characteristics risk factors, individual-level risk factors, family disruption risk factors, and social disruption risk factors. It was hypothesized that (a) individual-level risk factors (e.g., school performance, child alcohol/substance abuse) and (b) abuse characteristics risk factors (e.g., longer duration/frequency of abuse, use of force/threats of force, intrafamilial abuse) would predict higher levels of trauma symptoms. Furthermore, it was hypothesized that (a) family disruption risk factors (e.g., family alcohol/substance use, family psychopathology) and (b) social disruption risk factors (e.g., parental divorce, homelessness, witnessing homicide or violence) would moderate the impact of prior sexual abuse and predict higher levels of trauma symptoms. ^ The participants were 110 female children (5 to 18 years old) presenting for treatment for sexual abuse at a community agency (The Journey Institute) in Miami, Florida. This study conducted a retrospective analysis of an archival data set collected over a three-year period (1998–2001). The measures completed upon intake included The Journey Psychosocial Assessment and The Trauma Symptom Checklist for Children (TSCC; Briere, 1996). Using Pearson correlations and hierarchical multiple regression analysis, this study found that abuse characteristics risk factors and individual-level risk factors were predictive of maladaptive outcomes in this sample of sexually abused girls. However, no moderating effects were found for family disruption risk factors or social disruption risk factors. Therefore, the results of these analyses provided support for the contention that abuse characteristics and individual-level risk factors were appropriate targets for treatment for sexually abused girls. Moreover, limitations of this study, implications for treatment, and directions for future research were discussed. ^

Teacher efficacy of special education teachers of English language learners with disabilities

This research examined the perceived teacher efficacy of special education teachers of English Language Learners (ELLs) with educational disabilities by surveying 202 elementary special education teachers using the EXCEL Teacher Inventory. EXCEL consists of 20 teacher efficacy items, three open-ended questions, and participant demographic items. ^ Overall participant teacher efficacy scores were high. Of the variables assessed, a statistically significant difference in perceived efficacy was found with self-reported proficiency in the language of the target students. No statistically significant differences in teacher efficacy scores were found for (a) levels of teacher preparation; (b) number of years of teaching experience; or (c) socioeconomic status of the students. A multiple regression analysis indicated that, of the variables listed above, proficiency in the language of the target students accounted for significant variance in predicting the level of teachers' perceived efficacy. ^ Responses to the open-ended questions about what was most helpful when working with ELLs with disabilities yielded two major themes: organizational issues and teacher issues. Participants wrote numerous comments about the value of support from educational professionals and parents. Many participants recommended individuals in preservice programs take ESOL content specific courses. ^ The results demonstrate the positive correlation between proficiency in the language of the target students and teacher efficacy. This suggests that teachers of ELLs with disabilities have proficiency in the ELLs' native language (or have support from others who are language proficient) that allows them to distinguish between language difference and language disability and provide instruction in the native language when needed. Further, results from open-ended questions suggest that special education teacher preparation programs should include courses related specifically to strategies for teaching ELLs with disabilities, not just ELLs. ^

Predictors of developmental gains in children enrolled in an early intervention setting

The purpose of this research was to determine whether initial developmental delay, site of intervention, frequency of intervention, age of the child, socio-economic status (SES), gender and ethnicity significantly predict developmental gains in a group of children enrolled in an early intervention setting. The records of 134 children enrolled in an inner-city program in Miami, Florida were reviewed for inclusion in this study. ^ Demographic variables, site placement and treatment frequencies were collected during a retrospective chart review. Level of delay was expressed using the developmental quotient and developmental gain was calculated using the mean gain on age equivalent scores or developmental tests. A multiple regression analysis was performed to determine which of the above variables significantly predicted developmental gains. Multivariate analysis compared developmental gains for all the developmental domains based on intervention site (center versus home-based) while controlling for developmental delay. ^ Children made greater developmental gains if they had higher developmental quotients and if they were younger at the time services were initiated. Frequency of intervention significantly improved developmental outcomes in children attending center-based programs. Children attending center-based programs also made significantly greater gains in gross motor skills compared to children attending home-based programs. ^ These findings emphasize the importance of early screening and referral of children with developmental delay and adjusting intervention for the child's developmental quotient. Children should receive intense treatment to maximize results. Decisions regarding program placement should be individualized according to the child's unique developmental pattern. Policy and program decisions affecting the curriculum of a child in early intervention need to reflect these multivariate considerations. ^

Selected psychosocial and demographic factors affecting levels of pre-retirement planning

In September 2002, the State of Florida implemented a new retirement structure for those employed in the Florida Public School System. Teachers were given the option to maintain their existing defined benefit plan or choose the newly offered defined contribution plan. The variables that affect planning for retirement are innumerable. Identifying the most significant variables is essential to understanding how one plans for retirement. ^ This study examined the relationship between hypothesized psychosocial and demographic factors and an individual's level of pre-retirement planning. The criterion variable, the level of pre-retirement planning, comprised two concepts. First, the time spent thinking about retirement was determined by the score an individual received on a pre-retirement planning scale. This scale included the concepts of information gathering, goals, anticipated resources, and long-range planning. Second, implementation of retirement plan procedures was determined by the percentage an individual annually deferred to retirement. ^ The survey used for data collection contained 50 close-ended items. It was distributed to all full-time teachers in nine randomly selected elementary, middle, and senior high schools throughout Miami-Dade County Public Schools. Multiple regression and crosstabulation indicated that math anxiety, general risk, years of service, and total family income were significant predictors of the level of pre-retirement planning, as measured by the pre-retirement planning scale. In addition, the statistical analysis indicated that math anxiety, internal locus of control, years of service, and total family income were significant predictors of the level pre-retirement planning, as measured by the amount deferred to retirement. An individual's level of math anxiety and family income were the two factors that were the most significant predictors for both concepts on the level of pre-retirement planning. ^ Based on the findings of the study, recommendations focused on assessing an individual's level of math anxiety and educating teachers, particularly pre-service candidates, about the factors that affect pre-retirement planning. Further research should investigate the benefit of such educational programs. ^