Search

This study represents a large meta-analytic investigation of character education studies. We identified 214 studies up until 2017 (N[subscript Total] = 307,512) with at least one computable effect size comparing a character education program to a control condition. Initial analyses indicated a small, significant average positive effect for character education (g = 0.24, 95% CI [0.19, 0.28]). Single-session interventions and mentoring programs were associated with larger effects. However, all treatment lengths were associated with significant treatment effects, and the number of single-session interventions was small. Three programs--Cognitive Problem-Solving, Kohlberg's Moral Dilemma Discussion and Strong Kids--were associated with above-average effects in multiple assessments. Evidence emerged suggesting selection bias in published literature. Correcting this bias indicated lower estimates of mean effects while remaining positive and significant. We consider heterogeneity in reporting standards and discuss how to address biases.

This article summarizes issues in defining character education (CE) that are relevant to program development and evaluation. No authoritative definition exists for CE, but several features seem prominent. These include a focus on enhancing civic, moral, self-regulatory, and/or intellectual virtues; and (at least implicitly) prioritizing the shaping of participant identity over teaching skills. Within these broad parameters, programs vary widely in their targets and interventions. They can also vary in terms of emphasizing the same or individualized elements of character across students. The discussion finishes with a summary of elements of a CE program, and a call for well-planned evaluations.

Having an agreed-upon definition of character education would be useful for both researchers and practitioners in the field. However, even experts in character education disagree on how they would define it. We attempted to achieve greater conceptual clarity on this issue through a prototype analysis in which the features perceived as most central to character education were identified. In Study 1 (N = 77), we asked character education experts to enumerate features of character education. Based on these lists, we identified 30 features. In Study 2 (N = 101), experts assessed which features were central to character education through a categorization task. In Study 3 (N = 166), we assessed the extent of centrality using scalar items. We conclude by offering practical advice for the development of future character education studies and programs rooted in what is deemed central to such programs.

The "What Works in Character Education" project (WWCE; Berkowitz & Bier, 2007) was a descriptive summary of significant findings in the empirical literature on character education for the period 1945 to 2004. It employed a broad definition of character education to include related positive youth development programs such as socioemotional learning. Drawing on studies carried out in K-12 education settings, it encompassed research on published character development curricula as well as research on particular instructional practices such as moral discussion and cooperative learning. The current meta-analysis revisits WWCE by applying meta-analytic techniques to the same set of studies collected as part of the WWCE review. While the current analysis did not include any studies not covered by WWCE, it did modify its approach to examining 5 systematic reviews for which Berkowitz and Bier (2007) looked to the overall conclusions. In contrast, the current analysis expanded these reviews to individually examine each of their respective studies. It complements the already substantial referencing and discussion of the WWCE that exists in the character education literature. Random effects meta-analysis of the original WWCE studies examined 836 comparisons across 64 studies involving 96,930 participants and yielded a small but statistically significant average effect size, g = 0.33, 95% CI = [0.21, 0.45]. This finding is consistent with prior evidence on the effect size associated with various educational interventions (Lipsey & Wilson, 1993) and reaffirms that character education, as conceptualized and analyzed in the original WWCE, is effective in positively influencing character outcomes. The meta-analytic techniques also yielded new evidence of substantial heterogeneity of effects across studies, suggesting the potential value of focusing more narrowly, in future meta-analyses, on the character education programs that appear most promising. In the meta-analysis reported here, there were also indications of possible bias in some studies' reporting of results, but overall conclusions regarding effectiveness remained valid even after correcting for possible bias.

Character education specialists seem to know it when they see it, but what it means to call something a character education program remains unclear. One possible source for this uncertainty is the manner in which character education has been defined. By identifying certain features as necessary, existing definitions fail to encompass the universe of programs that can reasonably be considered instances of character education. Using recent literature on definition via prototype as a starting point, this article suggests evaluating programs as character education based on consistency with a model for such programs comprised of 7 features. Two strategies for evaluating whether and to what extent a program can be considered an example of character education are discussed, called polythetic classification and prototype matching. The goal is to help researchers and practitioners more effectively distinguish character education from other sorts of programs intended for personal growth.

Prior literature on the use of character strengths suggests that both deficiencies and excesses in the use of strengths can be problematic. While most school-based training in character strengths tends to focus on the former issue, an example is provided of a school-based program offered by the Mayerson Academy in partnership with the VIA Institute on Character that does not focus on deficiency but rather celebrates enhancing one's own personal key strengths. This program, entitled Thriving Learning Communities, will be implemented in 42 Cincinnati public schools this year. It aims to motivate students, as well as educators, to perform at their highest levels through encouraging increased use of character strengths. We provide an overview of the program and discuss how it may be modified by the teacher to offer a more nuanced perspective on character strengths within a gifted student population. Specifically, within a gifted classroom setting, we discuss ways in which educators can encourage gifted students to understand the importance of balancing their strengths and effectively matching their use of strengths to the situation. In addition to focusing on the use of signature or top strengths, we suggest a program that would aid gifted students in deciphering when it is most and least appropriate to use top strengths, through a dial-up/dial-down framework for targeted strength use.

Larry Nucci, in his article "Character: A Multifaceted Developmental System," has outlined a model of character that focuses on 4 domains of the moral system. This response focuses on 2 issues. The first is Nucci's equation of character with moral functioning. My research has led me to propose that 3 elements are essential to any adequate theory of character, called caring, inquisitiveness, and self-control. These emerge across a number of virtue theories, as well as models of essential skills and even literary references. In particular, I suggest that inquisitiveness and self-control are indicators of character even when not used to support moral ends, so long as they are not used to support immoral ones. Second, I think the model Nucci proposes for moral character is incomplete, as it fails to outline the competing considerations that enter into moral decision-making. Character education programs should acknowledge and even emphasize how these competing considerations can lead to nonmoral decision-making when environmental factors encourage poor choices.

The VIA Classification of Strengths and Virtues attempts to provide a comprehensive model of character based on 24 character strengths. The present study is the largest study to date exploring the structure of the 24 strengths in youth. One sample (N = 23,850) completed the VIA-Youth, a teen measure of the VIA Classification. Based on a random subsample, it was determined the data were best modeled using four factors. The remainder of the sample was used to demonstrate measurement invariance for the four-factor model across ages 10-17 and country. Comparison with 471 English academy school students who completed two alternate measures of the VIA Classification also demonstrated measurement invariance. The results suggest a four-factor model that includes two primarily interpersonal factors, one reflecting general engagement, the second other-directedness. Other factors involved intellectual and self-control strengths. Implications for the understanding of character strengths in youth versus adults are discussed.

Statistical analyses investigating latent structure can be divided into those that estimate structural model parameters and those that detect the structural model type. The most basic distinction among structure types is between categorical (discrete) and dimensional (continuous) models. It is a common, and potentially misleading, practice to apply some method for estimating a latent structural model such as factor analysis without first verifying that the latent structure type assumed by that method applies to the data. The taxometric method was developed specifically to distinguish between dimensional and 2-class models. This study evaluated the taxometric method as a means of identifying categorical structures in general. We assessed the ability of the taxometric method to distinguish between dimensional (1-class) and categorical (2-5 classes) latent structures and to estimate the number of classes in categorical datasets. Based on 50,000 Monte Carlo datasets (10,000 per structure type), and using the comparison curve fit index averaged across 3 taxometric procedures (Mean Above Minus Below A Cut, Maximum Covariance, and Latent Mode Factor Analysis) as the criterion for latent structure, the taxometric method was found superior to finite mixture modeling for distinguishing between dimensional and categorical models. A multistep iterative process of applying taxometric procedures to the data often failed to identify the number of classes in the categorical datasets accurately, however. It is concluded that the taxometric method may be an effective approach to distinguishing between dimensional and categorical structure but that other latent modeling procedures may be more effective for specifying the model. (Contains 4 tables and 8 footnotes.)

The taxometric method effectively distinguishes between dimensional (1-class) and taxonic (2-class) latent structure, but there is virtually no information on how it responds to polytomous (3-class) latent structure. A Monte Carlo analysis showed that the mean comparison curve fit index (CCFI; Ruscio, Haslam, & Ruscio, 2006) obtained with 3 taxometric procedures--mean above minus below a cut (MAMBAC), maximum covariance (MAXCOV), and latent mode factor analysis (L-Mode)--accurately identified 1-class (dimensional) and 2-class (taxonic) samples and produced taxonic results when applied to 3-class (polytomous) samples. From these results it is concluded that using the simulated data curve approach and averaging across procedures is an effective way of distinguishing between dimensional (1-class) and categorical (2 or more classes) latent structure. (Contains 3 tables, 1 figure and 2 footnotes.)

Professional psychologists are often confronted with the task of making binary decisions about individuals, such as predictions about future behavior or employee selection. Test users familiar with linear models and Bayes's theorem are likely to assume that the accuracy of decisions is consistently improved by combination of outcomes across valid predictors. However, neither statistical method accurately estimates the increment in accuracy that results from use of additional predictors in the typical applied setting. It was demonstrated that the best single predictor often can perform better than do multiple predictors when the predictors are combined using methods common in applied settings. This conclusion is consistent with previous findings concerning G. Gigerenzer and D. Goldstein's (1996) "take the best" heuristic. Furthermore, the information needed to ensure an increment in fit over the best single predictor is rarely available. (Contains 5 tables, 2 figures, and 5 footnotes.)

The increased use of effect sizes in single studies and meta-analyses raises new questions about statistical inference. Choice of an effect-size index can have a substantial impact on the interpretation of findings. The authors demonstrate the issue by focusing on two popular effect-size measures, the correlation coefficient and the standardized mean difference (e.g., Cohen's "d" or Hedges's "g"), both of which can be used when one variable is dichotomous and the other is quantitative. Although the indices are often practically interchangeable, differences in sensitivity to the base rate or variance of the dichotomous variable can alter conclusions about the magnitude of an effect depending on which statistic is used. Because neither statistic is universally superior, researchers should explicitly consider the importance of base rates to formulate correct inferences and justify the selection of a primary effect-size statistic.

The extent to which the Comprehensive System for the Rorschach is reliably scored has been a topic of some controversy. Although several studies have concluded it can be scored reliably in research settings, little is known about its reliability in field settings. This study evaluated the reliability of both response-level codes and protocol-level scores among 84 adolescent psychiatric inpatients in a clinical setting. Rorschachs were originally administered and scored for clinical purposes. Among response codes, 87% demonstrated acceptable reliability(> .60), and most coefficients exceeded .80. Results were similar for protocol-level scores, with only one score demonstrating less than adequate reliability. The findings are consistent with previous evidence, indicating reliable scoring is possible even in field settings.

A study of 74 undergraduates examined whether creativity training and interpersonal problem-solving training are equivalent or complementary. Results showed the training programs were complementary. Each training program specifically affected performance only on related measures of performance. A combination of the training programs affected both abilities. (CR)

In their comment, M. L. Rohling et al. (2011) accused us of offering a "misleading" review of response bias. In fact, the additional findings they provided on this topic are relevant only to bias assessment in 1 of the domains we discussed, neuropsychological assessment. Furthermore, we contend that, even in that 1 domain, the additional findings they described do not merit revision of our conclusion that the data are insufficient for evaluating the status of bias indicators. We remain hopeful that our review will spur researchers to publish additional tests of the validity of bias indicators in real-world settings and reduce the reliance on analogue studies as an evidence base for their use. (Contains 4 footnotes.)

The author agrees with the editors of this informative series of articles (Haig & Borsboom, 2008) that the number of articles appearing in recent years questioning basic assumptions about psychosocial measurement instruments, such as their mathematical qualities and their potential as valid indicators, augurs well for serious reconsideration of measurement practice. In this commentary the author focuses on two issues. The first is his own pet hypothesis about how psychosocial measurement evolved in the way it did. The second is what he thinks would need to happen before such reconsideration is likely to lead to praxis. (Contains 1 footnote.)

Using samples of psychiatric, medical, and chronic pain patients, the effectiveness of the Hysteria scale and of various combinations of Minnesota Multiphasic Personality Inventory scales as predictors of the simultaneous occurrence of two characteristics was evaluated: denial of psychological problems and admission of physical problems. The value of the Hysteria scale is discussed. (Author/KS)