Search results

In the evaluation of most interventions in criminal justice settings, evaluators have no control over assignment to treatment and control/comparison conditions, which means that the treated and comparison groups may have differences that lead to biased conclusions regarding treatment effectiveness. Propensity score analysis can be used to balance the differences in the groups, which can be used in a number of ways to reduce biased conclusions regarding effectiveness. A review of propensity scoring studies was conducted for this chapter, where the limited number of evaluations of criminal justice interventions using these methods was identified. Due to the small number of these studies, research was also reviewed if propensity scoring had been employed to evaluate interventions that are similar to those in criminal justice systems. These studies are used as examples to demonstrate how the methods can be used to evaluate criminal justice interventions, the different ways propensity scores can be used to analyse treatment and comparison group differences, and the strengths and limitations of this approach. It is concluded that, while not appropriate for all interventions/settings, propensity score analysis can be useful in criminal justice arenas, at least to investigate the comparability of treatment and comparison groups, with suspected non-comparability being a common weakness of traditional quasi-experimental studies and frequently cited limitation in terms of drawing efficacy conclusions from such evaluations.

Without controlling for selection bias and the potential endogeneity of the treatment by using proper methods, the estimation of treatment effect could lead to biased or incorrect conclusions. However, these issues are not addressed adequately and properly in higher education research. This study reviews the essence of self-selection bias, treatment assignment endogeneity, and treatment effect estimation. We introduce three treatment effect estimators – propensity score matching analysis, doubly robust estimation (augmented inverse probability weighted approach), and endogenous treatment estimator (control-function approach) – and examine literature that applies these methods to research in higher education. We then use the three methods in a case study that estimates the effects of transfer student pre-enrollment debt on persistence and first year grades. The final discussion provides guidelines and recommendations for causal inference research studies that use such quasi-experimental methods.

This paper aims to introduce matching in propensity score analysis (PSA) as an alternative statistical approach for researchers looking to make causal inferences using intact groups.

An illustrative example demonstrated the varying results of analysis of variance, analysis of covariance and PSA on a heuristic data set. The three approaches were compared by results and violations of statistical assumptions.

Through the illustrative example, it is demonstrated how different statistical approaches can produce varied results. Only PSA mitigated pre-existing group differences without violating the assumption of independence.

This paper attempts to answer calls in the literature for more robust statistical methodologies to better inform human resource development practice and theory.

The purpose of this paper is to illustrate how to examine the effectiveness of a pilot summer bridge program for elementary algebra using propensity scores. Typically, selection into treatment programs, such as summer bridge programs, is based on self-selection. Self-selection makes it very difficult to estimate the true treatment effect because the selection process itself often introduces a source of bias.

By using propensity scores, the authors can match students who participated in the summer bridge program with equivalent students who did not participate in the summer bridge program. By matching students in the treatment group to equivalent students who do not participate in the treatment, the authors can obtain an unbiased estimate of the treatment effect. The authors also describe a method to conduct a sensitivity analysis to estimate the amount of hidden bias generated from unobserved factors that would be needed to alter the inferences made from a propensity score matching analysis.

Findings suggest there is no significant difference in the pass rates of the subsequent intermediate algebra course for students who participated in the summer bridge program when compared to matched students who did not participate in the summer bridge program. Thus, students who participate in the summer bridge program fared no better or worse when compared to similar students who do not participate in the program. These findings also appear to be robust to hidden bias.

This study describes a unique way to estimate the causal effect of participating in a treatment program when there is self-selection into the treatment program.

Among the research studies related to the relevance between religious belief and mental health, most of them highlight people with religious belief who tend to obtain mental comforting more easily. However, the research studies mentioned above were cross-sectional studies, and they only verified that religious beliefs and mental health are relevant, but they did not prove their cause-and-effect relationship. That is, they do not identify “due to people's religious beliefs, they have healthier mind” or “due to people's healthier minds, they have religious beliefs.” Therefore, the study aims to explore the benefit evaluation of religious belief affecting mental health.

The study uses propensity score matching (PSM) and treatment effect (ATT) to carry out the causal inference between religious beliefs and mental health. First, the propensity score (PS) is calculated from the impact factors that affect people's religious belief before establishing counterfactual analysis based on the PS to analyze the effect of religious beliefs to further understand the difference of mental health index between people with religious belief and without it, and confirm the cause-and-effect relationship between them.

Religious beliefs and participation are ubiquitous within and across populations. The associates between religious participation and health are considerably in great magnitude. Most of the research in the past related to religious beliefs and mental health only verified that religious beliefs and mental health are relevant but not proved its cause-and-effect relationship. This paper aimed to explore the causal relationship between religious belief and mental health. The experimental results showed religious belief has treatment effect toward “daily functioning,” “feeling affect,” “spirituality” and “mental health.” On a whole, religious belief can promote mental health.

In academic and practical circles, there are a lot of research studies exploring the relationship between religious belief and mental health. However, there is no research investigating the cause-and-effect relationship between religious belief and mental health. It also causes some questioning toward the relevant research studies. Therefore, the outcome of this study not only can clarify the legitimacy, importance, and practicality on the researches in the past but also provide the practical support for psychology and counseling.

The current study seeks to assess the predictive utility of personality, family violence, associations with criminal friends, peer rejection, parental attachment, and parental supervision as predictors of homicidal behaviour among a sample of 144 male recidivistic offenders.

This research project utilized a quasi‐experimental design with propensity score matching in order to minimize the effect of selection bias. Post‐matching binary logistic regression analysis was subsequently conducted in order to determine what factors predict homicidal behaviour.

Post‐matching regression results indicated that experience of family violence, psychoticism, and parental attachments were significant predictors of being a homicidal murderer.

The findings provide strong empirical support for the important role of early childhood experiences in the prediction of homicidal acts, along with the crucial role of personality (psychoticism). These findings provide additional support for Eysenck's theoretical indications regarding the role of psychoticism in the prediction of violent criminal behaviours.

This study examines the relationship between small-firm managers' propensity to participate in a growth-oriented training program and their subsequent program outcome in terms of strategic reorientation. From a policy perspective, this relates to the important question of what benefit would come from recruiting managers who are normally not easily recruitable for training programs.

A control group design including pre- and post-training surveys is used to assess the effects of a large-scale management training program. Accounting for selection bias, the difference-in-difference method, together with propensity score matching, was applied to assess average program effects. The matching-smoothing method was used to assess heterogeneity in program effects associated with participation propensity.

Overall, program participation associated positively with change in strategic orientations. This effect was especially pronounced for managers with either low or medium to high inclinations for program enrollment, while diminishing in the modest to medium range.

The findings have important practical implications for selection of target groups and recruitment strategies in relation to small-firm management training programs. From the results, recruitment strategies may effectively include managers with either high or low participation propensities, rather than aiming to “fill up” with managers with moderately low participation propensity.

Several extant studies have examined average treatment effects from small-firm training programs. Yet there has been a lack of examination of the extent to which participation propensity modifies the effect of training on outcomes. This study brings new knowledge of the direction and magnitude of such heterogeneous training effects.

In this paper, the authors explore the relationship that slack resources and technology-mediated human capital investments can have on individuals’ entrepreneurial intentions. Focusing on human capital investments that individuals make through education and work, the authors analyze the relationship among formal online learning opportunities, informal skill development in the gig economy and entrepreneurial intentions.

Leveraging a novel dataset that merges administrative tax data with a survey of over 8,528 low- and moderate income (LMI) households, this study uses machine learning and propensity score weighting to examine the likelihood that individuals who make these technology-mediated human capital investments will have increased odds of entrepreneurial intentions when compared to similar individuals who do not make these investments.

The authors find that both partaking in online learning and working in the gig economy are significantly associated with increased odds of entrepreneurial intentions. Furthermore, through a variety of robustness and mechanism checks, the authors find that technology-mediation is an important factor in these relationships and that informal skill development and career preparation is one way in which gig employment influences entrepreneurial intentions.

As the study’s data come from a cross-sectional survey, the authors cannot make causal inferences about the relationship between online learning, gig employment and entrepreneurial intentions. Thus, future research should explore sources of longitudinal data.

This study has practical implication for individuals and policymakers that seek to increase entrepreneurship among LMI households.

Despite a wealth of research on the relationships among slack resources, technology and innovation at the firm level, there is little of this research at the individual level – especially among LMI individuals. The authors begin to fill this important gap.

This paper aims to examine whether Andersen’s audit quality in the five years preceding its collapse lagged that of other Big-Five auditors.

This paper compares Andersen’s audit quality and the other Big-Five auditors using five methodologies, namely, earnings response coefficients, magnitudes of abnormal accruals, propensities to issue going-concern opinions, usefulness of going-concern opinions in predicting bankruptcy and the frequency of Accounting and Auditing Enforcement Releases. The comparisons are based on both pooled samples of all observations and propensity-score-based matched-pairs.

The preponderance of evidence shows that Andersen’s audit quality did not differ materially in audit quality from other Big-Five auditors prior to its failure. However, it was found that Andersen’s independence was compromised in the year leading to its collapse (2000), as indicated by the lower likelihood to issue going-concern opinions.

This paper complements and improves on Cahan et al. (2011) by using more measures of audit quality, as no one measure is perfect, showing that their results using discretionary accruals are sensitive to the model used and showing that there is a more powerful direct measure of audit quality, namely, going-concern opinions.

In the existing literature on crowdfunding project performance, previous studies have given little attention to the impact of investors’ hedonic value and utilitarian value on project results. In a crowdfunding setting, utilitarian value is somehow hard to satisfy due to information asymmetry and adverse selection problem. Therefore, the projects with more hedonic value can be more attractive for potential investors. Lucky draw is a method to increase consumer hedonic value, and it can influence investors’ behavior as a result. The authors hypothesize that projects with hedonic treatment (lucky draw) may have higher probability to win their campaign than others. The paper aims to discuss these issues.

A unique self-extracted two-year Chinese crowdfunding platform real data set has been applied as the analysis sample. The authors first employ propensity score matching methods to control for the endogeneity of hedonic treatment adoption (lucky draw). The authors then run OLS regression and probit regression in order to test the hypotheses.

The analysis suggests a significant positive relationship not only between project lottery adoption and project results but also between project lottery adoption and project popularity.

The results suggest that an often ignored factor – hedonic treatment (lucky draw) – can play an important role in crowdfunding project performance.