Full Sample Model Research Articles

Living outside the Financial Mainstream: Alternative Financial Service Use among People with Disabilities

ABSTRACT This study explored factors related to the use of alternative financial services (i.e., payday loans, prepaid debit cards, pawn shops, auto title lenders, and rent-to-own stores) among a nationally representative sample of U.S. adults (N1 = 27,564), and a sub-sample of people with disabilities (PWDs) (N2 = 1,232). Using secondary data from the Financial Industry Regulation Authority’s 2015 National Financial Capabilities Study (NFCS), two separate logistic regression models revealed several significant individual and institutional level factors that predicted alternative financial service use among both samples. The full-sample model was found to be significant, χ2(10) = 3,917.22, p < .001, as was the sub-sample model, χ2(10) = 193.43, p < .001. Additionally, it was found that PWDs were twice as likely to use alterative financial services than people without disabilities. Further research is recommended to empirically explore the consequences of alternative financial service use among PWDs. Implications of this study are targeted toward social work researchers, policy makers, and human service practitioners.

Sexual Minority Youth Report High-Intensity Binge Drinking: The Critical Role of School Victimization

PurposeTo examine standard binge drinking (≥4 drinks for females, ≥5 drinks for males) and high-intensity binge drinking (≥8 drinks for females, ≥10 drinks for males) among heterosexual and sexual minority youth in the US and whether reports of school-based victimization mediate this association. MethodsSurvey data are from the 2015 Youth Risk Behavior Survey (YRBS; n = 10,839, Mage = 16.07). Logistic regression adjusted for race/ethnicity and age compared rates of standard and high-intensity binge drinking among heterosexual and sexual minority youth and whether experiences of school-based victimization mediated this association. Effects were tested in full sample and sex-stratified models. ResultsLesbian and bisexual girls and girls with male and female partners were more likely than heterosexual girls to report standard rates of binge drinking. Lesbian girls and girls reporting male and female sexual partners were more likely than heterosexual girls to report high-intensity binge drinking in the past 30 days. Compared with heterosexual boys, gay boys were significantly less likely to participate in high-intensity binge drinking. School-based victimization mediated all significant associations between sexual minority status and standard and high-intensity binge drinking, with the exception of lesbian girls. ConclusionLesbian and behaviorally bisexual girls have elevated risk for high-intensity binge drinking relative to heterosexual girls. Findings point to the importance of policies that reduce school-based victimization as these experiences are associated with higher rates of standard and high-intensity binge drinking among sexual minority girls.

General Strain Theory, Depression, and Substance Use: Results From a Nationally Representative, Longitudinal Sample of White, African-American, and Hispanic Adolescents and Young Adults

ABSTRACTBased on the interpretations of Agnew's general strain theory (GST), White, African-American, and Hispanic adolescents and young adults were examined longitudinally to identify the intersection between strain and depression that could produce a long-term effect on substance use. Results from full sample and group-based path models indicate that some support was found for the general tenets of GST. African-Americans were significantly different from Whites and Hispanics in regard to experiencing certain types of strain; however, the results were not always in the anticipated direction. Possible theoretical reasons for the findings and directions for future research are discussed.

The Impact of Accounting Conservatism and Voluntary Disclosure on the Cost of Capital of Industrial Companies in Jordan

This study aims at examining the impact of accounting conservatism and voluntary disclosure on the cost of capital of industrial companies in Jordan during the period (2009-2013). Panel OLS regression analysis was employed to test the hypotheses of the study. The results of the full sample model revealed that accounting conservatism and voluntary disclosure have significant negative impacts on the firms’ cost of capital.Furthermore, the results of the sub-samples which distinguish between large and small, as well as between high and low leverage firms showed that the sub-sample of large and small firms conforms to the full sample results. Across the sub-sample of high leverage firms, the results showed that only voluntary disclosure has a significant negative impact on the firm’s cost of capital. On the other hand, only accounting conservatism has a significant negative impact on the firm’s cost of capital across the sub-sample of low leverage firms.

Comparing Predictive Accuracy, Twenty Years Later: A Personal Perspective on the Use and Abuse of Diebold–Mariano Tests

The Diebold–Mariano () test was intended for comparing forecasts; it has been, and remains, useful in that regard. The test was not intended for comparing models. Much of the large ensuing literature, however, uses -type tests for comparing models, in pseudo-out-of-sample environments. In that case, simpler yet more compelling full-sample model comparison procedures exist; they have been, and should continue to be, widely used. The hunch that pseudo-out-of-sample analysis is somehow the “only,” or “best,” or even necessarily a “good” way to provide insurance against in-sample overfitting in model comparisons proves largely false. On the other hand, pseudo-out-of-sample analysis remains useful for certain tasks, perhaps most notably for providing information about comparative predictive performance during particular historical episodes.

Individual and jurisdictional factors associated with voluntary HIV testing in Canada: Results of a national survey, 2011.

HIV testing remains a central strategy for HIV prevention for its ability to link those who test positive to treatment and support. In Canada, national guidelines have recently changed as part of standard primary care to recommend voluntary HIV testing for those aged 16-64 years. Using results from a nationally representative survey, we examined individual and jurisdictional factors associated with voluntary testing. A total of 2,139 participants were sampled using a regionally stratified, two-stage recruitment process. English or French interviews (by phone or online) were conducted during May 2011. Voluntary testing was defined as testing at least once for reasons other than blood donation, insurance purposes, immigration screening or research participation. Weighted logistic regression analysis (including socio-demographic, sexual activity, HIV/AIDS knowledge and jurisdictional factors of HIV prevalence and anonymous testing availability) were conducted for the overall sample, and stratified by sex. Twenty-nine percent (29%) of survey participants reported at least one lifetime voluntary HIV test. For the full-sample model, the following were associated with increased odds of testing: age &lt;60 years, female sex, sexual minority status, perceived HIV knowledge, casual sex partner in previous year, and living in a higher-prevalence jurisdiction. For men, the strongest factor related to testing was sexual minority status (OR = 5.15, p &lt; 0.001); for women, it was having a casual sex partner in the previous year (OR = 2.57, p = 0.001). For both men and women, residing in a jurisdiction with lower HIV prevalence decreased odds of testing. Sex differences should be considered when designing interventions to increase testing uptake. Jurisdictional factors, including HIV prevalence and testing modality, should be investigated further.

Identifying Theoretical Predictors of Risky Alcohol Use Among Noncollege Emerging Adults

Studies show that emerging adults who do not obtain postsecondary education are at greater risk for developing alcohol use disorders later in life relative to their college-attending peers. Research examining constructs amenable to change within this population is necessary to inform intervention efforts. Thus, the current study aimed to identify psychosocial correlates of risky alcohol use for noncollege emerging adults. A secondary goal was to examine whether gender moderated the relationships between the psychosocial constructs and alcohol use. Participants were a nationally representative sample of noncollege emerging adults (18-22 years old) who reported using alcohol in the past year, recruited through an established Internet panel (N = 209; 125 women). A path model was used to examine the relationship between theoretically derived constructs (expectancies, attitudes, normative beliefs) and risky (peak) drinking. A second model examined a multigroup solution to assess moderating effects of gender. The full-sample model revealed significant associations between attitudes toward drinking and risky drinking. The model assessing gender differences revealed association between normative beliefs and drinking for women but not men, whereas attitudes were significantly associated with risky drinking for both men and women. Findings highlight the importance of attitudes and, for women, descriptive norms in the etiology of risky drinking among noncollege emerging adults, which emphasizes their potential utility in the development and adaptation of interventions for this at-risk population.

Can Ownership Diversification Enhance Banks’ Efficiency? An Analysis Based on Super Efficiency and Tobit Regression on the Chinese Listed Commercial Banks

Ownership diversification can improve the monitoring level in commercial banks and hence increase the efficiency. A good case is the Chinese commercial banks which have achieved outstanding performance in the recent years. In the past decade, many Chinese banks have been listed. Going listed has provided opportunities for the banks’ ownership to be diversified. Since those banks were controlled more or less by the state before going listing, it is meaningful to study the impacts of ownership diversification on the Chinese listed commercial banks’ efficiency. Our sample in this paper includes sixteen Chinese listed commercial banks with data from 2007 to 2011. Herfindahl index is the most widely used proxy to measure the degree of concentration. Hence it can reflect the degree of diversification too. In this paper, we have used Herfindhal index of the largest, the largest three, the largest five and the largest ten shareholders as the proxies for ownership structure. We include these four proxies because in the Chinese listed banks, some are owned by several largest shareholders, while others have much fewer large shareholders. There are several methods to measure the banks’ efficiency, to get more precise measurement, we adopt the super efficiency method. Super efficiency has the advantage of higher comparability for the sampled banks. on the other hand, we have used Tobit regression model rather than the ordinary least square (OLS) method because the dependent variable (the super efficiency) is positive. In this paper, we have measured the super efficiency with Data Enveloping Analysis (DEA) and then studied the impacts of the ownership diversification on the efficiency with Tobit regression model. The Chinese listed banks can be categorized into two groups. One is the four formerly state-owned largest commercial banks (hereinafter “Big Four”), the other is the group of smaller banks with less control from the government (hereafter non-Big Four). Because these two groups have different ownership structure, the ownership diversification can influence their efficiency differently. To discover the differences of the impacts from the ownership diversification, we try to compare the impacts of ownership diversification on the efficiency of the formerly state-owned larger banks (Big Four) and smaller banks (Non-Big Four). To overcome the problem of inadequate observations in Big Four sample, we first run Tobit regression with the full sample, and then run Tobit regression with non-Big Four sample. The difference in the coefficients can reveal the different impacts of ownership diversification on the efficiency. Our results indicate that ownership diversification can enhance the listed commercial banks’ efficiency. But the coefficients in the non-Big Four sample is smaller than those in the full sample models, which indicates ownership diversification has greater positive impacts on the efficiency of the formerly state-owned larger banks than that of the smaller banks. Our contribution is that our paper is the first one to study the impacts of the ownership diversification on the listed commercial banks which were previously held by the state. DOI: http://dx.doi.org/10.5755/j01.ee.24.1.2471

Comparing Predictive Accuracy, Twenty Years Later: A Personal Perspective on the Use and Abuse of Diebold-Mariano Tests

The Diebold-Mariano (DM) test was intended for comparing forecasts; it has been, and remains, useful in that regard. The DM test was not intended for comparing models. Unfortunately, however, much of the large subsequent literature uses DM-type tests for comparing models, in (pseudo-) out-of-sample environments. In that case, much simpler yet more compelling full-sample model comparison procedures exist; they have been, and should continue to be, widely used. The hunch that (pseudo-) out-of-sample analysis is somehow the only, or best, or even a good way to provide insurance against in-sample over-fitting in model comparisons proves largely false. On the other hand, (pseudo-) out-of-sample analysis may be useful for learning about comparative historical predictive performance.

Location Decision of Korean Manufacturing FDI: A Comparison between Korean Chaebols and Non-Chaebols

This study analysed the location selection determinants of Korean foreign direct investment (KFDI) from the perspectives of OLI paradigm and resource- and knowledge-based views. The study involved a sample of 1911 Korean multinational enterprises (MNEs) and their 3095 foreign subsidiaries during the period of 1999–2004. Specifically, the study compared the differences of foreign direct investment (FDI) location decisions (1) between less-developed and developed countries (LDCs and DCs) and (2) between chaebols and non-chaebols. The results showed that the partitioned sample model of chaebols versus non-chaebols was found to have more detailed effects on the location decision between the two regions than the combined full sample model. In addition, the results showed that there are clear, effectual differences for subsidiary scale and prior host country experience between chaebols and non-chaebols.

One Size Does Not Fit All: An Examination of Low Birthweight Disparities Among a Diverse Set of Racial/Ethnic Groups

To examine disparities in low birthweight using a diverse set of racial/ethnic categories and a nationally representative sample. This research explored the degree to which sociodemographic characteristics, health care access, maternal health status, and health behaviors influence birthweight disparities among seven racial/ethnic groups. Binary logistic regression models were estimated using a nationally representative sample of singleton, normal for gestational age births from 2001 using the ECLS-B, which has an approximate sample size of 7,800 infants. The multiple variable models examine disparities in low birthweight (LBW) for seven racial/ethnic groups, including non-Hispanic white, non-Hispanic black, U.S.-born Mexican-origin Hispanic, foreign-born Mexican-origin Hispanic, other Hispanic, Native American, and Asian mothers. Race-stratified logistic regression models were also examined. In the full sample models, only non-Hispanic black mothers have a LBW disadvantage compared to non-Hispanic white mothers. Maternal WIC usage was protective against LBW in the full models. No prenatal care and adequate plus prenatal care increase the odds of LBW. In the race-stratified models, prenatal care adequacy and high maternal health risks are the only variables that influence LBW for all racial/ethnic groups. The race-stratified models highlight the different mechanism important across the racial/ethnic groups in determining LBW. Differences in the distribution of maternal sociodemographic, health care access, health status, and behavior characteristics by race/ethnicity demonstrate that a single empirical framework may distort associations with LBW for certain racial and ethnic groups. More attention must be given to the specific mechanisms linking maternal risk factors to poor birth outcomes for specific racial/ethnic groups.

Characteristics of acquired firms: the case of the banking industry

Purpose – The purpose of the paper is twofold: to determine firm characteristics that explain mergers in the banking industry and to predict the likelihood of a merger.Design/methodology/approach – A logit model is used to estimate coefficients. The paper also tests the effect of the Financial Services Modernization Act on market structure and uses a national sample from 1997 to 2003.Findings – Results indicate that profitability (PROF), size (SIZE), asset management (ASSETMGT), and solvency (SOLV) are factors in explaining mergers in the full sample model. Poor ASSETMGT, large target banks, low PROF, and solvent banks represent merged firms in the full sample model. The results also indicate that loan activity is not a factor in explaining mergers. It appears that banks merged before the Financial Services Modernization Act (Pre‐1999 Merger Model) are characterized by poor ASSETMGT, and are solvent. In contrast, the only factor explaining mergers after the Financial Services Modernization Act (Post‐1999 ...

Improving the Accuracy of Fetal Foot Length to Confirm Gestational Duration

To establish normative fetal foot length ranges using last menstrual period (LMP) and ultrasound dating by biparietal diameter and to examine variations in these ranges by ethnicity. A consecutive series of 1,099 eligible subjects receiving abortions had fetal foot lengths measured directly. Models of fetal foot length were developed by using assessment of gestational duration by LMP alone, ultrasonography alone, and "best estimate" (LMP confirmed by ultrasonography). The full sample model using ultrasound dating (n = 1,099) yielded the following equation: foot length = -30.3 + days of gestation x 0.458 (R(2) of 0.92). Regression by LMP-determined gestational duration by using the "best estimate" sample (n = 491) provided an almost identical equation (foot length = -29.8 + days of gestation x 0.45) and a similar R(2) value of 0.87, although the standard errors were larger. Gestational duration by ultrasonography alone produced a better model fit than duration by LMP alone. Regressions by ethnicity were not significantly different compared with the simple regression, regardless of method used to determine gestational duration. A reconsideration of fetal foot length measurements to confirm gestational duration is important. More accurate tables of these measurements allow for greater precision in correlating gestational duration and foot length. Fetal foot length tables using ultrasonographically confirmed gestational duration and current statistical standards should replace tables currently used. Biparietal diameter as a single measurement provides adequate estimation of gestational duration in the second trimester for pregnancy termination, proving more reliable than LMP dating. II-2.

On the relationships between gender, early onset, and the seriousness of offending

Building on previous research implicating the age of onset as a predictor of later patterns of delinquent activity, data from subjects participating in the Philadelphia Collaborative Perinatal Project (CPP) are combined with criminal history records from the Philadelphia Police Department to explore the relationships between gender, early onset, and the seriousness of offending. Findings suggest that an early onset before age fourteen is predictive of serious offending by age eighteen, with the earliest onset ages relating to the most severe delinquency patterns. This relationship holds in the full sample after controls for relevant variables; however after controlling for variables considered in the full sample model, the early onset→seriousness of offending relationship holds for males, but vanishes for females. The theoretical and policy implications of these findings and potential directions for future research are discussed.

Land Leasing and Debt on Farms: Substitutes or Complements?

Abstract Land Leasing and Debt on Farms: Substitutes or Complements? Theoretically, leasing and debt are thought to be substitutes. This assumes that a leasepayment, which is a fixed obligation like a loan, displaces debt and reduces debt capacity, i.e., iffirms have optimal debt to equity ratios, then, to the extent that it represents “off-balance sheet”financing, leasing reduces debt capacity. Ang and Peterson—the seminal work in theliterature—fit Tobit models with 1976 to 1981 data from 600 firms in which a leasing to bookvalue of equity ratio is the dependent variable and a debt to book value of equity ratio and othervariables are the explanatory variables. Contrary to expectations, their model results indicate thatleasing and debt are complementary activities. This study follows the Ang and Petersonmethodology, but utilizes a set of firms which are distinct from those of earlier studies—non-corporate U.S. commercial farms—to test a land leasing-debt substitution hypothesis. Anadvantage of the land lease example is that by focusing on a single industry—productionagriculture—the problem of potential industry affects is substantially reduced. In a departurefrom earlier studies, the issue of whether the leasing-debt relation is sensitive to heterogenous firmcharacteristics and shifting business conditions is examined.OLS leasing models corrected for heteroskedasticity are fit with 1977 through 1992Kansas farm-level data in which a leasing ratio is the dependent variable and a debt ratio and otherexplanatory variables serve as independent variables. The models account for fixed time and farmtype effects. Results strongly indicate that land leasing and debt are substitutes, albeit not dollarfor dollar. A coefficient estimate of about -0.43 on the total asset full sample model indicates thaton average, leasing decreases by $0.43 for each dollar of debt incurred. The rate at which farms