Book

Hierarchical Linear Models: Applications and Data Analysis Methods

Authors:
... Variation in and covariation among repeatedly measured variables reflects both between-person and within-person variation. Thus, if these variables are not properly addressed, for example, if a time-varying outcome is naively regressed on a time-varying predictor, the resulting regression coefficient would be an"uninterpretable blend" of the between-person and within-person effects of the predictor (Cronbach, 1976;Raudenbush & Bryk, 2002). It is also called conflated effect (Preacher et al., 2010;Rights et al., 2020) and smushed effect (Hoffman, 2015(Hoffman, , 2019Hoffman & Walters, 2022) in the multilevel modeling literature, and is mathematically a weighted average of the between-person and within-person effects (Raudenbush & Bryk, 2002). ...
... Thus, if these variables are not properly addressed, for example, if a time-varying outcome is naively regressed on a time-varying predictor, the resulting regression coefficient would be an"uninterpretable blend" of the between-person and within-person effects of the predictor (Cronbach, 1976;Raudenbush & Bryk, 2002). It is also called conflated effect (Preacher et al., 2010;Rights et al., 2020) and smushed effect (Hoffman, 2015(Hoffman, , 2019Hoffman & Walters, 2022) in the multilevel modeling literature, and is mathematically a weighted average of the between-person and within-person effects (Raudenbush & Bryk, 2002). ...
... Centering is rescaling a variable, in particular, shifting the origin by subtracting a certain value from the variable. For example, subtracting a group-level mean is called group-mean centering, which is a popular centering option in multilevel analysis (Enders & Tofighi, 2007;Hamaker & Muthén, 2020;Hoffman, 2019;Kreft et al., 1995;Raudenbush, 1989;Raudenbush & Bryk, 2002;Snijders & Bosker, 2012). In a longitudinal setting, the mean of a time-varying predictor X ti for person i (i.e.,X ·i = T t=1 X ti /T ) 1 is subtracted from the raw predictor (i.e., person-mean centering), and the centered predictor has a form of X ti −X ·i . ...
Article
Full-text available
In longitudinal studies, researchers are often interested in investigating relations between variables over time. A well-known issue in such a situation is that naively regressing an outcome on a predictor results in a coefficient that is a weighted average of the between-person and within-person effect, which is difficult to interpret. This article focuses on the cross-level covariance approach to disaggregating the two effects. Unlike the traditional centering/detrending approach, the cross-level covariance approach estimates the within-person effect by correlating the within-level observed variables with the between-level latent factors; thereby, partialing out the between-person association from the within-level predictor. With this key device kept, we develop novel latent growth curve models, which can estimate the between-person effects of the predictor's change rate. The proposed models are compared with an existing cross-level covariance model and a centering/detrending model through a real data analysis and a small simulation. The real data analysis shows that the interpretation of the effect parameters and other between-level parameters depends on how a model deals with the time-varying predictors. The simulation reveals that our proposed models can unbiasedly estimate the between- and within-person effects but tend to be more unstable than the existing models. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
... In most multilevel applications, the errors in the level-1 model are assumed to have equal variance, σ 2 . According to [10], if the level-1 variance varies randomly over level-2 units, but these variances are assumed equal, consequences for inference about the level-2 coefficients will be mild, on the other hand if the =variances depend systematically as a function of level-1 or level-2 predictors, consequences may be more serious. Because causes of heterogeneity are quite different in their implications, it is strongly advocated to investigate possible sources of heterogeneity and model it if found. ...
... 00 is the intercept of (3), it can be understood as the average of the dependent variable Y when the independent variable is 0. is the value on the level-2 predictor 01 is the regression coefficients of the variables in (3), it can be understood as the impact of the variable to the initial value of the dependent variable Y. 10 is the intercept of (4), it can be understood as the changing rate of observed object when the variable is 0. ...
... In order to test the hypothesis, hierarchical linear modeling (HLM) is used to investigate variables at various levels of analysis (cross-level), namely unit level (Innovation-oriented culture, learning culture, adhocracy culture, hierarchical culture,clan culture, and market/rational culture) and individual level (Innovative behavior) (2-1). Raudenbush and Bryk (2002). are the creators of this model. ...
... When the predictor variables have multiple hierarchical levels, a move complicated version of ordinary least squares (OLS) regression named hierarchical linear modeling (HLM) is used to examine variation in the outcome variables. This model is developed by Raudenbush, and Bryk, (2002). HLM is the ideal analysis method for looking at cross-level models where individual and group outcomes vary. ...
... While it is useful to understand the population-level effects of input(s) on outcome(s), insights regarding how and why effects differ across individuals can be valuable. The random effects in a multilevel model indicate the presence of individual heterogeneity in input effects 50 . A simple approach to identify factors contributing to individual heterogeneity is to introduce each factor in an interaction term with the input variable and test its significance. ...
... A simple approach to identify factors contributing to individual heterogeneity is to introduce each factor in an interaction term with the input variable and test its significance. This is known as slopes-as-outcomes modeling 50 . However, this approach is sensitive to noise in longitudinal data and becomes cumbersome as the number of potential factors increases 50 such as in our case. ...
Article
Full-text available
We conducted a field study using multiple wearable devices on 231 federal office workers to assess the impact of the indoor environment on individual wellbeing. Past research has established that the workplace environment is closely tied to an individual’s wellbeing. Since sound is the most-reported environmental factor causing stress and discomfort, we focus on quantifying its association with physiological wellbeing. Physiological wellbeing is represented as a latent variable in an empirical Bayes model with heart rate variability measures—SDNN and normalized-HF as the observed outcomes and with exogenous factors including sound level as inputs. We find that an individual’s physiological wellbeing is optimal when sound level in the workplace is at 50 dBA. At lower (<50dBA) and higher (>50dBA) amplitude ranges, a 10 dBA increase in sound level is related to a 5.4% increase and 1.9% decrease in physiological wellbeing respectively. Age, body-mass-index, high blood pressure, anxiety, and computer use intensive work are person-level factors contributing to heterogeneity in the sound-wellbeing association.
... Specifically, we used packages psych (Revelle, 2021) and lme4 (Bates et al., 2015). Considering that we had repeated measures of HSCL-11 nested within patients, we decided to use Hierarchical Linear Models (HLMs; Raudenbush and Bryk, 2002;Gómez Penedo et al., 2019). These models account for the dependency of the data when there is a nested structure, providing more robust estimations of the parameters. ...
... handling missing data mimicking an intent-to-treat approach (Westra et al., 2016). First, we ran an unconditional model (i.e., without any predictor) to compute Intraclass Correlation Coefficients (ICC) representing the percentage of variability explained by the patient level (Raudenbush and Bryk, 2002). ...
Article
Full-text available
Introduction Routine Outcome Monitoring (ROM) has emerged as a strong candidate to improve psychotherapy processes and outcome. However, its use and implementation are greatly understudied in Latin-America. Therefore, the aim of the present pilot study conducted in Argentina was to implement a ROM and feedback system grounded on a psychometrically sound instrument to measure session by session outcome in psychotherapy. Methods The sample consisted of 40 patients and 13 therapists. At baseline, the patients completed the Patient Health Questionnaire-9 and the Generalized Anxiety Disorder-7, and they also completed the Hopkins Symptom Checklist-11 before each of the first five sessions. To estimate patient change during the first sessions, we conducted a quantitative analysis using Hierarchical Linear Models. Furthermore, we conducted a qualitative analysis using Consensual Qualitative Research to analyze therapist perception regarding the ROM and feedback system. Results Results showed a significant reduction in patients’ symptomatic severity during the first five sessions. Additionally, baseline depression significantly predicted the estimated severity at the end of the fifth session. Feedback was given to the therapists after the first four sessions based on these analyses. With regard to the perception of the feedback system, clinicians underlined its usefulness and user-friendly nature. They also mentioned that there was a match between the information provided and their clinical judgment. Furthermore, they provided suggestions to enhance the system that was incorporated in a new and improved version. Discussion Limitations and clinical implications are discussed.
... m.in. [Bryk, Raudenbush 1992;Hox 2010;Goldstein 1995]). Warunkiem zastosowania modeli wielopoziomowych jest zagnieżdżony charakter danych, w którym losowane są z populacji jednostki zarówno z pierwszego, jak i z drugiego poziomu losowania. ...
... An intraclass correlation coefficient (ICC) was calculated (ICC = 9.4%) to ensure HLM was appropriate for the nested data. ICC values ranging from 5% to 20% warrant social sciences research to employ HLM (Muthén, 1994;Raudenbush & Bryk, 2002;Peugh, 2010). After calculating the ICC from the unconditional model, the HLM technique had three steps. ...
... . 따라서 인지적 및 정의적 수준 양상이 상이한 한국과는 달리 우수한 교육 수준으로 여러 선행연구 (구자옥, 한정아, 김성숙, 2015;임해미, 이현숙, 김성숙, 2016;임해미, 한정아, 2016;Yoon & Järvinen, 2016 (구자옥 외, 2015; 김성숙, 한정아, 2016), 구조방정식 모형 (박현정, 2008; 임해미 외, 2016), 다층모형 (윤정일, 이범수, 2006; 임해미, 한정아, 2016), 다층 구조방정식 모형 (이금호, 정혜원, 2016 (최필선, 민인식, 2018), 기존 이론에서 간과되어 왔던 새로운 영향변인을 탐색할 필요성이 있다. 최근 이러한 필요성에 따라 PISA 자료에 머신러닝 기법인 랜덤포레스트 (Breiman, 2001)를 적용한 연구 (손윤희, 박현정, 박민호, 2020 (유진은, 2015), 랜덤포레스트 분석에서 도출된 주요 영향변인을 다층모형 (Raudenbush & Bryk, 2002) (강상진, 2016). ...
... 연구모형에서   는 학생수준 예측변수,   는 학교수준 예측변수,   는 학교별 평균 종속변수,   는 전체 종속변수 평균,   는 학교수준 주요 예측변수의 회귀계수를 의미한다.   는 학교별 학생 수준 예측변수의 회귀계수,   는 전체 학생수준 주요 예측변수 회귀계수를 의미하며, 본 연구에서는 모형의 간명성을 위해 학교 간 차이가 없다고 가정하는 무선절편(random intercept) 모형으로 분석 을 실시하였다(Bryk & Raudenbush, 1992). 혼합효과 랜덤포레스트 분석과 마찬가지로 다층모형도 10개의 유의측정값을 활용해 분석을 반복한 후 통합된 결과를 도출하였다.혼합효과 ...
... Then, before running any of the substantive models described below, we started with a null model to estimate the degree of "between-individual" variance in intraclass correlation coefficient. In order to explore if biological sex (coded 0 = men, 1 = women) moderated the relation of SUP-S and SUP-C with HRV, we estimated a two-level hierarchical linear model to account for the nested nature of the data (i.e., occasions nested within individuals; Bryk & Raudenbush, 1992), separately for SUP-S and SUP-C. All predictors were included as person-level (or Level 2) grand-mean centered covariates. ...
Article
The present study explored the association between perceived social support at work (from coworkers and supervisors), and heart rate variability (HRV), a measure of parasympathetic modulation of the heart which is considered an index of the ability to regulate emotional arousal in tune with environmental demands. To test these associations, we used data from 144 workers of various sectors (61.10% women; Mage = 40.45, SDage = 14.17 years). For these workers, HRV was assessed on three principal daily periods for which we computed three different HRV estimates: (a) during the working time, (b) during the pause(s) at work (averaging HRV across different pauses occurring randomly across individuals), and (c) during the evening (after work) till 10.00 p.m. Multilevel model results revealed that workers perceiving more social support by their supervisor showed increased HRV across different phases of the day (i.e., at work, during a work pause, and in the evening). Furthermore, the associations between social support perceived by supervisor and HRV were moderated by biological sex such that this association was stronger for men compared with women. Altogether, results provided support for the positive role of social interactions at work in fostering workers’ psychophysiological adjustment. Theoretical and practical implications of our results for individuals and organizations are discussed.
... The data-generating model was the heteroscedastic two-level random intercept only model, which is given by the following equations. Applying Raudenbush and Bryk's (2002) notation, the model reads: ...
Preprint
Full-text available
In random-effects models, hierarchical linear models, or multilevel models, it is typically assumed that the variances within higher-level units are homoscedastic, meaning that they are equal across these units. However, this assumption is often violated in research. Depending on the degree of violation, this can lead to biased standard errors of higher-level parameters and thus to incorrect inferences. In this article, we describe a resampling technique for obtaining standard errors—Zitzmann’s jackknife. We conducted a Monte Carlo simulation study to compare the technique with the commonly used delete-1 jackknife and the robust standard error in Mplus. Findings revealed that Zitzmann’s jackknife and the commonly used delete-1 jackknife clearly outperformed the robust standard error in rather small samples with high levels of heteroscedasticity. Moreover, Zitzmann’s jackknife tended to perform somewhat better than the commonly used delete-1 jackknife and was much faster.
... As the longitudinal data for children at ages 12-13, 14-15 and 16-17 were collected at different waves and nested within each individual, the data were analysed using a linear mixed-effects model. This method creates a two-level hierarchical model that nests time within individual (Bryk & Raudenbush, 1992;Miyazaki & Raudenbush, 2000), and simultaneously examines the relationship within and between hierarchical levels of nested data. Variables at Level-1 were used to estimate the average within-person trajectory and linear rate of change over time. ...
Article
Full-text available
To examine the role of neighbourhood, what Bronfenbrenner describes as an element in the exosystem, as a protective asset for adolescents’ social and emotional wellbeing. The study used a subset of national data reported by adolescents and their parents from the Longitudinal Study of Australian Children (LSAC). A linear mixed-effect model was used to estimate the association between neighbourhood indicators and trajectories in a measure of social and emotional wellbeing (the SDQ Total Difficulties score) at ages 12–13, 14–15 and 16–17, controlling for age and sex, peer and family relationships and household material resources. The analysis revealed that parents’ perceptions of belonging and their subjective assessments of the condition of housing in the neighbourhood, as well as externally sourced data on neighbourhood accessibility and socio-economic status, were significantly associated with adolescents’ total difficulties scores over time. The findings revealed the role of neighbourhood level protective assets as a potential influence on adolescents’ social and emotional wellbeing. Greater attention should be given to understanding the complex interactions between the resources mobilised by individuals and their families, and the influence of wider environments and social structures on young people’s social and emotional wellbeing.
... Because adult children are grouped within the category of parents, the intergenerational relationship quality of adult children within the same family will be correlated. This correlation violated a core statistical assumption of standard regression methods that all observations should be independent of one another (Raudenbush & Bryk, 2002). If this correlation is ignored, incorrect inferences can result in the false estimation of the effects of both parent-and child-level factors. ...
Article
Full-text available
Intergenerational relationships have become increasingly crucial for maintaining well-being in aging families. Under a changing sociocultural background, families in Hong Kong increasingly exhibit diverse intergenerational relationships and functioning. Focusing on families with mutiple children, this study investigated how the characteristics of parents and their adult children jointly affect different domains of intergenerational relationship quality. A two-stage stratified random sampling design was adopted. Face-to-face questionnaire interviews were conducted between November 2016 and March 2017 with 1,001 Hong Kong residents aged ≥ 50 years. Data of 612 parents and 1,745 adult children were included for analysis. Hierarchical linear modeling was performed to examine child- and parent-level correlates of intergenerational relationship quality. Parents who were female, were married, had higher self-perceived economic status, owned a house, and had fewer depressive symptoms, exhibited higher intergenerational relationship quality. Parents’ age was positively related to affectual closeness, whereas their educational level was negatively related to both affectual closeness and conflict. More favorable intergenerational relationships were reported by aging parents whose adult children were younger, female, and married. Children with higher educational levels exhibited higher levels of both affectual closeness and conflict with their parents. Moreover, affectual closeness was found to be transmitted between generations. The findings can help improve awareness of the factors affecting the different domains of intergenerational relationships, thus informing the development of targeted services and interventions to promote family relationships and well-being.
... The within-and between-person correlations reported above handle the multilevel structure of the daily data separately. Multilevel modeling (MLM) statistical techniques handle both levels simultaneously and permit analysis of interdependent data without violating the assumptions of ordinary least squares (OLS) regression (Bryk & Raudenbush, 1992). Thus, we used MLM path analysis as our primary analytic strategy for examining associations among focal predictors and MIL. ...
Article
Do supernatural attributions to God and the devil relate to a sense of meaning in life? We investigated this question by conducting a three-week experience sampling study (N = 75), in which we measured daily causal attributions for both positive and negative events and assessed daily meaning (n = 1,425 total daily reports). Correlational and multilevel path analyses revealed several reliable between-person associations. People who made more attributions to God (particularly for positive events) had higher meaning, and people with stronger religious worldviews who made more attributions to the devil/demons (particularly for negative events) had higher meaning. These findings were largely robust when controlling for relevant meaning and religious/spiritual covariates. We did not find evidence for within-person associations; daily changes in attributions and meaning were not related. Consistent and enduring attributions to God and the devil/demons may be particularly meaningful ways of understanding the world.
... The hypothesis was examined using multilevel modeling (MLM) with repeated-observation (within-person) data at level-1 and person-level data at level-2 (see Raudenbush & Bryk, 2002) with Hierarchical Linear Modeling (HLM) 7 software (Raudenbush et al., 2011). Consistent with Schmidt and DeShon (2007), group centering was used for all level-1 predictors. ...
... The province-level data is incorporated into the individual-level through a random intercept hierarchical model. Specifically, the study follows the hierarchical or multilevel modeling used by Raudenbush and Bryk (2002). Please see Appendix 1 for the full description of the model. ...
Article
Full-text available
Obesity has recently developed into a global epidemic. This study follows an integrated approach of examining the determinants and prevalence of obesity in Canada. A multimethod analysis revealed that almost two-thirds of Canadians are overweight or obese. It was found that males, the aged, the married, the less educated, physically inactive people, and people with poor self-rated health have increased probability of being overweight/ obese. Several regulatory policy options which could reduce the incidence of obesity and, in turn, increase social welfare and individual well-being are discussed.
... Long-term analysis of record scores in selected exercises is essential, both from the sporting perspective as well as the adaptive changes occurring in young athletes under the influence of strength training [26,27]. In strength training itself, the correct selection and distribution of loads to increase the athletes' exercise capacity is of primary importance [28]. Front and back squats are among the leading exercises used in this context [22]. ...
Article
Full-text available
Introduction. Front and back squats are multi-joint exercises used in sports to develop strength, power and muscle mass. Due to their movement structure and involvement of the main muscle groups, they are training measures that are used in strength training in various sports. A long-term training process leading to sports mastery requires the correct selection of loads as well as objective control of the training effects to date, both short-term and long-term. The aim of this study was to create a statistical model and to determine with it the character of the changes in the record scores in the front and back squat of young weight-lifters over a two-year training cycle. Material and Methods. The study included 17 young athletes training in competitive weightlifting. Recorded scores in the front squat and back squat were measured seven times (every three months over a two-year training cycle). The progression method was used to determine the maximum load, and the repetition with the highest load (1RM) was used in the calculations. Results. The subjects performed better in the back squat, but the difference between the results for both squats was approximately constant over the two-year training cycle. The time courses of both squats were similar over the analysed period. Systematic increases in maximum results were observed. The first statistically non-significant differences between the mean record values in successive measurements were found between the sixth and seventh measurements for both squats. Conclusions. Determining the relationship between front and back squat record scores can contribute to the optimisation of training loads in sports using strength training. Obtaining an analytical form of the course of record scores over time for both squats helps to assess the skill level of athletes and predict their performance in the next training macro-cycle.
... The time-varying predictors (within person-variables: state negative affect, co-rumination, and job search intentions) were centered on individual means (i.e., group-mean centered), and the between-person variables (multidimensional scale of perceived social support) were grand-mean centered. Centering the predictor variables ensure accurate interpretation of statistical estimates in multilevel modelling (Raudenbush & Bryk, 2002). Additionally, Bauer and exceeds the threshold of 0.7 suggested by Hair et al. (2019). ...
Article
Full-text available
Based on social cognitive, reasoned action, and basic needs theories, this study examines whether co-rumination with others about the job search mediates the positive relation between state negative affect and job search intentions. In addition, we looked at how this positive indirect effect interacts with the social support received from different sources (such as friends, family, and significant others) at the beginning and six months after the initial job search process. Using a sample of 87 graduates (job seekers) from a Portuguese masters program, we used multilevel modeling to test this moderated mediation. Ages ranged from 22 to 53 years old (M = 29.45; SD = 7.60). Data were collected using measures to assess negative affect, co-rumination, perceived social support, and job search intentions. We found an indirect effect of state negative affect on job search intentions; when job seekers perceive higher levels of social support from significant others at the beginning of the job search, the positive, mediating role of co-rumination in this relationship is increased. Further, for a sub-sample of six-month job seekers, this positive indirect effect increased when there was also an increased perception of social support from family. Discussion focuses on implications for theory and practice and the role of co-rumination for unemployed people during job search.
... Researchers in educational psychology also have more statistical software firepower available to them than ever before -and these are proving extremely exciting for analysing within-person data. For example, structured appropriately, real-time data can be analysed using multilevel modeling (Goldstein, 2003;Raudenbush & Bryk, 2002). Again, taking motivation as a case in point, if we collect motivation responses from many students numerous times a day, fivedays a week, over the course of a month at school, we have a multilevel design where time is nested within day, day is nested within week, week is nested within month, and month is nested within student. ...
... In other words, the data were nested. Therefore, the model was designed as a hierarchical linear one (Bryk and Raudenbush, 1992); this method has also been adopted in other studies (for a similar approach, see Table 1). ...
Article
Full-text available
Employee innovative behavior is significant in maintaining an organization's sustainable development. This study explored the impact of team psychological safety and workplace anxiety on the association between self-serving leadership and employee innovation behavior by synthesizing social information processing theory, conservation of resources theory, and ego depletion theory. We conducted a hierarchical linear model analysis using three-wave paired data collected from 86 leaders and 392 employees. The research results showed that self-serving leadership is negatively correlated with employee innovation behavior. Meanwhile, team psychological safety and workplace anxiety mediated this relationship. In addition, team psychological safety mitigates the impact of workplace anxiety on employee innovation behavior and the indirect impact of self-serving leadership on employee innovation behavior via workplace anxiety. These findings have a number of theoretical and practical implications in the domains of self-serving leadership and employee innovation behavior.
... Note. R&B R 2 indicates the R 2 statistics for each variance component in a model proposed by Raudenbush and Bryk (2002). The change in R&B R 2 when a predictor is added to the model is indicative of the amount of additional individuallevel variance that is explained by the addition of the predictor. ...
Article
Full-text available
Place attachment, as an important emotional and functional bond between people and place, has received extensive attention in the early-21st century. This trend is especially evident in China, where rapid market transitions have fundamentally changed urban neighborhoods. Using a survey of residents in Guangzhou, China, remote sensing imagery, geographic information system (GIS) techniques, and hierarchical linear modeling (HLM), this study investigates the role of physical and social environments in place attachment. Results indicate that commodity housing residents generally do not necessarily have strong place attachment. However, higher quality physical and social environments do contribute to stronger place attachment. Specifically, perceived housing conditions, social ties, sense of security, and lower floor area ratio (FAR) are positively associated with place identity, while perceived housing conditions, perceived neighborhood environment, social ties, lower FAR, and built environment quality are positively associated with place dependence. Social ties are the strongest predictor of place identity while housing conditions are the strongest predictor of place dependence. We conclude that social environments still matter as social and physical environments play different roles in place identity and place dependence. Based on the findings, we advocate that urban (re)development and housing policies should shift the focus from large-scale demolitions to micro-scale dwelling improvements; and in addition to objective measures, subjective evaluation by residents should also be used to evaluate urban (re)development projects.
... The r ðf1vÞ w statistic is defined as the square root of the proportion of variance explained by within-person predictors via fixed slopes and random slope variation/covariation. This is similar to a measure of the square root of the proportion reduction in variance, akin to a correlation (Hox, 2002;Kreft & de Leeuw, 1998;Raudenbush & Bryk, 2002). 8 NEWMAN, NEZLEK, AND THRASH Coefficients from Models 2a and 2b are presented in Table 3 and shown in Figure 3. Supplication had a negative effect on well-being the following day. ...
Article
Full-text available
Prayer is an important aspect of many people's daily lives, yet little is known about the relationships between prayer and daily experiences and well-being in ecologically valid settings. In three studies, participants (N = 350) completed questionnaires once a day for 2 weeks (4,437 daily reports) regarding the events they experienced each day, their emotions, well-being, and the prominence of the four types of prayer constituting the ACTS (adoration, confession, thanksgiving, and supplication) taxonomy. Thanksgiving and adoration were more prominent in prayers on days when positive events were reported and well-being was high (relative to individuals' own average reports of positive events and well-being). In contrast, supplication was more prominent on days when negative events were reported and well-being was low. Relationships between daily events, states of well-being, and prayers of confession were mixed. Lagged analyses indicated that present-day supplication, thanksgiving, and adoration negatively predicted well-being the following day. These lagged effects were weaker among people who prayed more frequently. Finally, each prayer type was predicted by distinct, nonreligious emotional states-supplication by envy, thanksgiving by gratitude, confession by guilt, and adoration by awe. By moving beyond cross-sectional and experimental paradigms, we have provided insights about the dynamic nature of prayer through repeated measurement in naturalistic contexts. The content of individuals' prayers reflects their daily experiences and has consequences for their well-being. (PsycInfo Database Record (c) 2023 APA, all rights reserved).
... The intraclass correlation for a model that only includes an intercept, based on a within-study variance of 0.201 and a between-study variance of 0.138, amounts to 0.138/ (0.138 + 0.201) = 0.407. This value shows that a high percentage of the overall variance in elasticities is due to differences between studies, which justifies the use of HLM (Raudenbush & Bryk, 2002). The Q-test (Q = 26,887.07; ...
Article
Full-text available
How does advertising affect supply and demand in the entertainment industry? Different advertising and distribution mechanisms and unique product characteristics limit the transferability of findings from other industries to the entertainment industry. This meta-analysis focuses on 290 documented elasticities, drawn from 59 studies of movies and video games, and establishes new findings and empirical generalizations. First, the average advertising elasticity in the entertainment industry is .33 (method bias-corrected .20), approximately three times higher than the average identified for other industries. Second, average advertising elasticities are higher for demand (e.g., revenue) than for supply (e.g., screens). Third, elasticities of pre-launch advertising are higher than those of overall advertising budgets, but with respect to the success period, elasticities are higher for later periods, and in total, compared to the launch period. Fourth, elasticities tend to be rather recession-proof and consistent across geographic regions but decreased after the rise of social media platforms.
... For Research Question 2, three-level hierarchical linear models (HLM; Raudenbush & Bryk, 2002) were used. For all models, time (i) was a Level 1 predictor, with students (j) nested within Level 2, and group (k) association as a Level 3 predictor to account for variance in implementation at the group level and controlling for differences in student outcomes related to group membership. ...
Article
Assessing implementation allows for a better understanding of an intervention's effects and the mechanisms that influence its impact. Two main areas of implementation are (a) the quality with which an intervention is delivered and (b) instructors’ adherence to the programmed intervention. The current study used data from a kindergarten mathematics intervention program to (a) examine if and how treatment adherence was associated with implementation quality and (b) explore implementation measures’ relation to student mathematics outcomes. Results indicated high implementation scores across time for both adherence and quality. Neither treatment adherence nor implementation quality was found to relate to a general outcome measure of student mathematics achievement; however, both were similarly related to the curricular‐aligned measure.
... Random intercept ANOVAs (one for each outcome; [39]) were utilized to address study aims and hypotheses. Along with a random intercept, time, condition, and the time by condition interaction were all modeled as independent variables. ...
Article
Forced migrants suffer from significant psychological distress. However, they often prioritize urgent practical resettlement needs over mental health needs. The present study used a quasi-experimental design to compare pathways of treatment for survivors of torture (N = 369) from 42 different counties receiving care from a refugee health clinic. Random intercept ANOVAs were used to compare combined case management services and psychological treatment (CM-PT) to case management services only (CM) on changes in cultural adaptation and global functioning over time. Results showed that both groups improved on each outcome. Importantly, the CM-PT group endorsed greater improvements in cultural adaptation (b = 0.28, 95% CI 0.14, 0.41, p ≤ 0.001) and global functioning (b = 3.29, 95% CI 1.33, 5.25, p = 0.001) compared to the CM group. These findings suggest that treatment for survivors of torture should be multifaceted and include case management and psychological treatment. Case management services alone may be beneficial when socio-cultural and resource barriers exist for mental health treatment.
... In the second step, two intra-class coefficients [ICC(1) and ICC(2)] were calculated for each item to justify the requirement of running multilevel analyzes. ICC(1) was used to detect whether there were significant and substantial variations located at the school level for each item (Raudenbush and Bryk, 2002). Large values of ICC(1) indicate substantial variations occurring at the school level, suggesting that multilevel analysis might be required to simultaneously incorporate both the teacher-level and school-level varieties. ...
Article
Full-text available
Introduction Understanding the sources and the effects of collective teacher efficacy has been one of the central interests to many educational researchers and practitioners, because it is critical to understand how teachers can shape, and are shaped by, the educational processes in schools. Following the social cognitive perspective on the sources and consequences of efficacy beliefs, this study examined how school support influences collective teacher efficacy which in turn affects teachers’ organizational commitment. Method The participants included 969 teachers sampled from 28 primary and secondary schools in Hong Kong. To appropriately address the nature of collective teacher efficacy and school support as school-level variables, the doubly latent multilevel structural equation modeling approach was used to analyze the data. Results The results revealed the mediation mechanism played by collective teacher efficacy in explaining the effect of school support on teachers’ organizational commitment. Discussion Schools are suggested to consider fostering a supportive school environment as a strategy to improve teachers’ collective efficacy beliefs if it is wished to enhance teachers’ commitment to schools.
... As recommended in the literature, no participants were excluded due to missingness in the diary part (Hox, 2002;Singer & Willett, 2003). Instead, we retained participants with missings and used maximum likelihood estimation as a missing data technique (Raudenbush & Bryk, 2002;Singer & Willett, 2003;Wang et al., 2017). Taken together, the 410 participants delivered between 7,777 and 6,083 observations for the daily variables (e.g., work engagement was only assessed on working days). ...
Article
Full-text available
Burnout negatively affects employees’ health, life satisfaction, and performance. However, little is known about how burnout shapes employees’ resilience process in daily life to produce these adverse effects. Therefore, we present a 30-day diary study among an international sample of 410 employees, studying burnout-related differences in response to an acute stressor (i.e., learning about the COVID-19 diagnosis of a close friend or family member). Specifically, we investigate how this event affects COVID-19-related worrying, positive and negative affect, and work engagement, both on the day itself and across several post-event days. Multilevel analyses with cross-level interactions between individual-level burnout and day-level stressor occurrence reveal that employees high in burnout score significantly higher on negative affect and lower on positive affect and work engagement on the day the stressor occurred. Additionally, discontinuous random coefficient growth modelling with burnout-time interactions shows that employees high in burnout sustain higher levels of COVID-19 worrying, but their negative and positive affect return to pre-event levels in the post-event days. These findings shed important new light on how burnout affects employees’ resilience process in response to acute stressors, thereby potentially identifying a key proximal mechanism by which burnout’s negative distal effects on health, well-being, and performance emerge.
... Using Mplus 4.0 (Muthén & Muthén, 2006), multilevel modeling was run to test hypotheses (see , table 3) since each member in the sample was nested under the corresponding country culture from which he or she came, avoiding underestimate or overestimate standard errors for parameter estimates. (Bryk & Raudenbush, 1992;Snijders & Bosker, 1999). To test the main effects of FBS sources on cultural variables, the author estimated an intercept-only regression model for FBS sources at Level 1 and predictive effects of cultural practice variables on the Level-1 random intercept were Level-1 random regression slopes of collectivism and individualism in predicting cultural value both by group means when they were entered into corresponding multilevel models to avoid interpretation difficulties and spurious findings according to Hofmann and Gavin's (1998) suggestion. ...
... Given that they sometimes related to several children per mother, the hypotheses were examined from a nested perspective. Thus, the analyses employed hierarchical linear models, using mixed models accounting for the interdependency of the data on children within each family (Raudenbush & Bryk, 2002). ...
Article
The quality of life (QoL) evaluations of children of divorced parents have been little studied. The literature focuses on the consequences of divorce for their wellbeing as reported by adults or examined using tools designed from the adult perspective. This study examines how risk (parental conflict) and resilience factors (perceived overall social support, closeness to grandparents, and open discourse with them about divorce) are related to the self-reported QoL of children of divorce. Cross-sectional online questionnaires for a child and one parent were completed by 122 children aged 7-17, from 86 Israeli families. Hierarchical linear models were employed, with mixed models accounting for the interdependency of children's data within each family. All risk and resilience factors were examined from the children's perspective, except parental conflict that was examined also from the parent's perspective. Parental conflict was found to be negatively associated and overall social support and close relationship with grandparents were positively associated with children's QoL evaluations. The more the children were involved in parental conflict, the more QoL measures were associated with it. The study also supported three models in which perceived overall social support and perceived grandparental support moderated the association between parental conflict and QoL evaluations. The importance of examining the wellbeing of children of divorce from their own perspective is discussed. In terms of practical implications, the findings suggest the need to reduce the involvement of children of divorce in parental conflict and to strengthen their social support systems, particularly their relations with grandparents.
... This situation may result in a biased estimation of standard errors (Hofmann, 1997). A multilevel regression model is an appropriate alternative and allows one to test cross-level hypotheses-in particular, cross-level interaction hypotheses-obtaining the estimates of robust standard errors (Raudenbush & Bryk, 2002). Scholars generally treat cross-sectional designs with skepticism, given their inherent limitations including deficient causal inferences. ...
Article
Full-text available
Organizational image, defined as internal members' evaluations of their organization, is a critical managerial issue since it influences employee work attitudes and subsequent performance. While the public administration discipline has observed an increasing scholarly interest in the influence of organizational image on employee engagement, it leaves the potential interrelationships between two distinctive organizational image dimensions-construed external image (CEI) and perceived organizational identity (POI)-unexplored. We examine the influence of these two organizational images on employee engagement , with an emphasis on both the mediating and moderating role of POI between CEI and employee engagement. Empirical results show that both image dimensions positively influence employee engagement. More importantly, we find that POI acts as a moderator in the CEI-employee engagement relationship, that is, the positive influence of CEI on the outcome depends on the level of POI. Empirical findings imply that public employees treat two dimensions of organizational image differently and weigh POI more than CEI. In particular, the findings suggest that attractive or positive POI may serve as a catalyst to reinforce employee engagement even in the presence of a long-lasting negative external image for public organizations.
... We used hierarchical linear modeling (HLM) as our primary analytic technique (Raudenbush and Bryk, 2002) to control for the intra-class correlation (ICC) that was evident and attributable to the nesting of firms within countries (ICC = 0.126, p ≤ .001 in model 1c; ICC = 0.086, p ≤ .01 in model 2c; and ICC=0.110, p ≤ .001 in model 3c, see Table 2). ...
Preprint
Full-text available
Despite recent evidence linking gender diversity in the firm with firm innovativeness, we know little about the underlying mechanisms. Building on and extending the Upper Echelon and entrepreneurship literature, we address two lingering questions: why and how does gender diversity in firm ownership affect firm innovativeness? We use survey data collected from 7,848 owner-managers of SMEs across 29 emerging markets to test our hypotheses. Our findings demonstrate that firms with higher gender diversity in ownership are more likely to invest in R&D and rely upon a breadth of external capital, with such differentials explaining sizeable proportions of the higher likelihood of overall firm innovativeness, product and process, as well as organizational and marketing innovations exhibited by their firms. Our findings are robust to corrections for alternative measurement of focal variables, sensitivity to outliers and subsamples, and endogenous self-selection concerns.
Article
Full-text available
Este trabajo explora las relaciones y efectos que existen entre los antecedentes socioeconómicos del alumno, la composición social de las escuelas y la repetición de grado en el rendimiento de los alumnos. Utilizando los resultados de la prueba en Competencia Lectora de la evaluación PISA 2009 aplicada en el Perú, el interés se focaliza en los probables beneficios que puede aportar la repetición como práctica escolar para mejorar los aprendizajes. Aunque los datos no permiten inferencias causales, los análisis efectuados sugieren que los alumnos que han conocido la repetición en su carrera escolar obtienen sistemáticamente rendimiento escolares más bajos que sus compañeros no-repetidores. Se ha constatado, también, que las escuelas favorecidas socialmente tienen altos rendimientos asociados a una baja repetición, mientras que las escuelas desfavorecidas tienen más alta repetición y más bajos rendimientos. Además, se ha identificado tres ventajas producidas por efectos de los antecedentes socioeconómicos de los alumnos. La primera es para aquellos que provienen de hogares favorecidos socialmente, quienes obtienen en todos los casos mejores puntajes conforme su capital social se acrecienta. Una segunda ventaja viene de la textura social de la escuela frecuentada por un público escolar favorecido, que también se ve bonificado con mayores puntajes. En fin, estudiar en una escuela favorecida socialmente protege mejor de la repetición que el propio antecedente socioeconómico del alumno. Esta exploración de variables que se asocian con el rendimiento del alumno, así como los efectos de la repetición, brindan información que puede contribuir al debate sobre la pertinencia de esta práctica escolar.
Article
Full-text available
Evaluaciones internacionales, regionales y nacionales han mostrado que los estudiantes peruanos tienen serias deficiencias en el aprendizaje de las matemáticas. Así, el identificar variables de procesos escolares que podrían ayudar a mejorar los resultados educativos, en términos de calidad y equidad, es fundamental para la mejora de la educación en nuestro país. El presente estudio se sirve de las bases de datos de Pisa 2012 para estimar un modelo multinivel de coeficientes aleatorios que permita no solo modelar el rendimiento de los estudiantes, sino las brechas educativas asociadas a diferencias socioeconómicas. Los principales resultados del estudio muestran que, entre las variables de procesos escolares usadas en el presente estudio, el clima disciplinario en clase tiene un efecto positivo y significativo sobre el rendimiento promedio de los estudiantes, mientras que el clima escolar y el sentido de pertenencia promedio de los estudiantes en la escuela tienen un efecto sobre las brechas de rendimiento debido a diferencias socioeconómicas.
Article
Background: The weight of a newborn is measured for the first time shortly after birth. The World Health Organization divides newborns’ birth weight into three categories: low birth weight (2.5 kg), normal birth weight (2.5 kg-4 kg), and high birth weight (> 4 kg). Both the mother and the infant are at risk of mortality and morbidity as a result of their birth weight. Using hierarchical data, there is scant evidence in Ethiopia of factors linked with birth weight. The goal of this study was to use a multilevel ordinal logistic regression model to investigate geographical variance and factors related to baby birth weight. Methods: Using missing factors in datasets, data for this study was collected from the Ethiopia Demographic Health Survey 2016. To address missing data and increase the inference’s reliability, hot deck multiple imputations were utilized. A multilevel ordinal logistic regression model was used to examine factors associated with birth weight. R software was used for analysis. Results: The study took into account a total of 8,328 newborns. According to a descriptive study, 1292 (15.5%) of the 8,328 babies were born with low birth weight, 6143 (73.8%) were born with normal birth weight, and 893 (10.7%) were born with high birth weight. Mother’s age, residence, mother’s age at first birth, wealth index, BMI, anemia level, gestational age, total children, mother delivery, multiple pregnancies, and baby’s sex were all found to be significant factors associated with a birth weight of Ethiopian babies in a multilevel ordinal logistic regression analysis. Conclusions: The multilevel ordinal logistic regression analysis revealed that there was significant variance in baby birth weight between and within Ethiopian regions. Among the three multilevel models, the random coefficient model fits the data the best.
Chapter
This chapter is focused on the connection between flow and deep learning, and the potential for learning environments fostering flow and deep learning to prepare emerging adults for times of crises. Topics addressed include how crisis affects learning, as well as the importance of deep learning in times of crisis, and how it is enabled by motivation, engagement, and flow. In this context, I present a 3 year, quasi-experimental study comparing students’ motivation, engagement, and deep learning of course materials between emerging adults who took an undergraduate engineering course that compared a video game approach to a control group. The video game, EduTorcs, provided challenges in which the student participants created control algorithms to drive virtual cars through a simulated game environment. Students taking the game-based course reported greater intrinsic motivation, work-play integration, and engagement in deep learning characteristic of flow than students taking the course in the traditional way; and they performed significantly better on tests of complex course concepts designed to measure deep learning. Results showed positive effects for concentration and engagement, which can be specific barriers to learning in times of crisis. Findings also suggest the potential role of the video game approach in facilitating intrinsic motivation, flow, and deep learning. Implications for educational and learning strategies in future crises are discussed.KeywordsFlowDeep learningCrisisGame-based learningEmerging adulthood
Article
Drawing on survey data of Christian students (N = 4261) enrolled at 11 colleges that are part of the Council for Christian Colleges and Universities, we examine color‐blindness, different types of color‐cognizance, and views on race in Christian organizations. We find significant differences between students of color and White students in color‐blindness and color‐cognizance but find racial convergence in the importance of racial reconciliation. Net of controls, greater evangelical orthodoxy is associated with color‐cognizant views that perceive prejudicial treatment against White people, undercut the experiences of people of color, and is associated with more positive views of Christian organizations’ roles in race relationships. Concurrently, upper year students differ significantly from first years in their racial attitudes and views on race in Christian organizations, suggesting that Christian education may shape racial views positively. Our findings have implications for theorizing how evangelical Christianity and evangelical colleges maintain and challenge color‐blindness and White supremacy.
Article
Using a sample of up to 2,503 initial public offerings (IPOs) in 32 countries from 2011–2017, we predict and find that higher levels of country‐level accounting enforcement are associated with lower levels of IPO underpricing. IPOs in countries with a relatively low accounting enforcement score (second quintile) exhibit a mean underpricing of 19%, whereas the mean underpricing amounts to just 9% in countries with a relatively high score (fourth quintile). The results remain qualitatively the same when we employ a multi‐level model or a difference‐in‐difference design. In countries that substantially strengthened their accounting enforcement in the 2003–2009 period, the level of IPO underpricing decreased significantly. We show that accounting enforcement matters for the cost of going public.
Article
Experiences during emerging adulthood hold particular importance for future health and economic well-being. This study uses Panel Study of Income Dynamics Transition into Adulthood Supplement data to explore growth trajectories of perceived discrimination across the transition into adulthood and to examine how gendered racialized group status might shape perceived discrimination among U.S. emerging adult men and women racialized as white, non-Hispanic, Black, non-Hispanic, or Hispanic ( n = 2532) using multilevel models. Results showed that as emerging adults age there is a decreasing pattern of perceived everyday discrimination, though there were individual differences in the rate and direction of change. Women racialized as Black, non-Hispanic had significantly lower perceived discrimination than all other groups except Hispanic women. Each of the six gendered racialized subgroups had similar patterns of perceived discrimination over time. More research is needed to better understand differences in the rate and direction of change in perceived discrimination over time.
Article
We investigate whether the characteristics of audit committee (AC) chairs are associated with decisions about auditor choice, audit fees and audit quality. Using hand‐collected Australian data, firms with AC chairs who have longer tenure and multiple AC memberships across several boards are found to be more likely to choose Big 4 and/or industry specialist auditors, pay higher audit fees and have lower discretionary accruals. Those AC chairs with higher business qualifications are more likely to hire a Big 4 auditor, pay higher audit fees and have lower discretionary accruals, while AC chairs with professional qualifications are more likely to hire a Big 4 and/or industry specialist auditor. In contrast, firms with AC chairs who are executive directors are less likely to hire a Big 4 auditor and have higher discretionary accruals. Our findings contribute to the literature by documenting that various characteristics of AC chairs are important for enhancement of auditor selection and audit quality.
Thesis
Full-text available
Interesse ist eine bedeutsame motivationale Variable, die sich positiv auf die Persistenz bei der Bearbeitung von Lernaufgaben, die Schulleistung und das leistungsbezogene Wahlverhalten von SchülerInnen auswirkt. Im Verlauf der Schulzeit ist ein deutlicher Rückgang des mathematisch-naturwissenschaftlichen Interesses von SchülerInnen zu beobachten. Es ist eine zentrale Aufgabe naturwissenschaftsdidaktischer Forschung, Strategien für die Interessenförderung zu erproben und zu deren praktischer Implementierung beizutragen. Diese Studie evaluiert die Wirksamkeit von zwei mathematisch-naturwissenschaftlichen Profilierungsangeboten, die im Zeitraum von 2011 bis 2018 im Rahmen des Schulversuchs zur „Förderung mathematisch-naturwissenschaftlich hochbegabter, begabter und interessierter Schülerinnen und Schüler“ an einem Thüringer Gymnasium erprobt wurden. In Form eines Mehrkohorten-Längsschnitts wurde die Ausprägung und Entwicklung des Fachinteresses von insgesamt 643 SchülerInnen mehrerer Jahrgänge in den Fächern Mathematik, Mensch-Natur-Technik, Biologie und Physik von Klassenstufe 5 bis 8 mithilfe eines standardisierten Fragebogens erfasst. Zusätzlich wurde dokumentiert, für welche Profilierungsangebote des Schulversuchs sich die SchülerInnen im Verlauf der Sekundarstufe I entschieden. Als geeignete Maßnahme zur Förderung des Fachinteresses wurden die „Klassen besonderer Profilierung“ identifiziert, in die sich die SchülerInnen jeweils beim Übergang in Klassenstufe 7 einwählen konnten. Jede dieser Klassen erhielt zusätzlichen Fachunterricht in Mathematik, Physik und Biologie im Umfang von mehreren Wochenstunden. Die Teilnahme an den Klassen besonderer Profilierung wirkte sich insbesondere auf das Mathematik- und Physikinteresse der Schülerinnen positiv aus. Die Wirksamkeit war dabei abhängig von der Anzahl der zusätzlichen Unterrichtsstunden. Je mehr Stunden zusätzlich erteilt wurden, umso deutlicher zeigte sich der interessenfördernde Effekt. Demgegenüber zeigte der mathematisch-naturwissenschaftliche Profilunterricht langfristig keine messbare interessenfördernde Wirkung. Der Profilunterricht war thematisch eher bereit abgelegt, stand nicht in Bezug zu den Fachlehrplänen und fand im Umfang von einer Wochenstunde statt. Auf Basis der Ergebnisse werden Schlussfolgerungen für die Verstetigung der Maßnahmen an der Versuchsschule, die Übertragbarkeit der Maßnahmen auf andere allgemeinbildende Schulen sowie für die naturwissenschaftsdidaktische Interessenforschung gezogen.
Article
Investigaram, com base em um modelo teórico, os fatores socioeconômicos que determinam a alocação entre estudo e mercado de trabalho da juventude no Brasil. Mais especificamente, analisaram o que leva um jovem a se tornar um nem-nem, isto é, não estar empregado e nem em processo de educação ou treinamento. Para tanto, além de se considerar as características dos indivíduos já testados pela literatura, como as relacionadas aos atributos pessoais (gênero, idade e raça/cor) e atributos do lar (tamanho da família e escolaridade do chefe), procurou-se verificar o efeito do contexto social na decisão de estar na condição nem-nem. Em tal análise, considerou-se os dados do Censo Demográfico 2010, do Datasus e do Ipeadata.
Article
Previous research on linguistic relativity and economic decisions hypothesized that speakers of languages with obligatory tense marking of future time reference (FTR) should value future rewards less than speakers of languages which permit present tense FTR. This was hypothesized on the basis of obligatory linguistic marking (e.g., will) causing speakers to construe future events as more temporally distal and thereby to exhibit increased “temporal discounting”: the subjective devaluation of outcomes as the delay until they will occur increases. However, several aspects of this hypothesis are incomplete. First, it overlooks the role of “modal” FTR structures which encode notions about the likelihood of future outcomes (e.g., might). This may influence “probability discounting”: the subjective devaluation of outcomes as the probability of their occurrence decreases. Second, the extent to which linguistic structures are subjectively related to temporal or probability discounting differences is currently unknown. To address these, we elicited FTR language and subjective ratings of temporal distance and probability from speakers of English, which exhibits strongly grammaticized FTR, and Dutch, which does not. Several findings went against the predictions of the previous hypothesis: Framing an FTR statement in the present (“Ellie arrives later on”) versus the future tense (“…will arrive…”) did not affect ratings of temporal distance; English speakers rated future statements as relatively more temporally proximal than Dutch speakers; and English and Dutch speakers rated future tenses as encoding high certainty, which suggests that obligatory future tense marking might result in less discounting. Additionally, compared with Dutch speakers, English speakers used more low‐certainty terms in general (e.g., may) and as a function of various experimental factors. We conclude that the prior cross‐linguistic observations of the link between FTR and psychological discounting may be caused by the connection between low‐certainty modal structures and probability discounting, rather than future tense and temporality.
Article
This study investigated the effectiveness of ABRACADABRA (ABRA), a web-based literacy program, developed by the Center for the Study of Learning and Performance at Concordia University, in facilitating the literacy attainment of third-grade primary school students in an area of rural China. Researchers shared the experiences of localizing ABRA to make it more readily accepted by the regional teachers. Ten schools (five experimental, five control) with 339 students participated in a one-year-long matched control study; participants in both groups took pre- and post-intervention tests. The treatment group outperformed the controls on all six key reading outcomes: Phonological Awareness (ES = +0.64), Phonemic Awareness (ES = +0. 57), Early Literacy Skills (ES = +0.70), Initial Letter Sound Fluency (ES = +0.38), Segmenting (ES = +0.81), and Non-word Reading (ES = +0.92). These outcomes suggest ABRA is effective in a context where English is used as a foreign language, and the learning resources are limited. ABRA is useful in facilitating the integration of information and communication technologies (ICTs) into the course syllabus.
Article
Background and Objectives: A large portion of veterans referred to substance use disorder (SUD) treatment do not attend their first appointment or leave treatment prematurely. Telehealth options may increase access to care and treatment attendance. This study compared treatment initiation and retention between inperson treatment pre-COVID-19 and telehealth treatment during COVID-19 in a VA hospital outpatient alcohol and drug treatment clinic. Methods: Electronic health record data were collected on 373 veterans out of 481 referrals. Mixed logistic and negative binomial regression models were used for estimating the effect of treatment modality on treatment initiation and retention, respectively. Results: The odds of initiating treatment were 2.6 times greater when referred to a telehealth appointment during COVID-19 compared to an in-person appointment pre-COVID-19. Months retained in treatment postinitiation was similar between treatment modalities. Each month in treatment corresponded to a subsequently lower rate of treatment session attendance and by the fourth month, the average number of sessions attended was significantly lower for those receiving in-person treatment compared to those receiving telehealth treatment. Conclusions and Scientific Significance: The odds of initiating SUD treatment and rate of treatment attendance over time was greater for services delivered via telehealth during COVID-19 compared to in-person pre-COVID-19. Extending telehealth services as an option for individual and group SUD treatment may promote treatment initiation and potentially higher rate of attendance over time. This study provides additional evidence for the feasibility and utility of rapidly expanding telehealth for veterans seeking outpatient substance use treatment.
Article
Full-text available
Unlabelled: This article describes racial and ethnic differences in mortgage payment difficulties during the COVID-19 pandemic and examines whether disparities exist in the benefits of the unemployment insurance (UI) program. The sample consisted of 80,797 jobless mortgage borrowers who received or waited for UI benefits between August 2020 and May 2022. Considering individual- and state-level variables in multilevel logistic regressions, we examined rates of mortgage delay in the last month and payment concerns about the next month by racial and ethnic group. Minority borrowers were more likely to have a difficulty in paying mortgage than White borrowers. UI recipients-regardless of race and ethnicity-were less likely to experience mortgage difficulties, but the positive unemployment benefit was reduced disproportionately among Blacks. Blacks were also at a higher risk of mortgage difficulties compounded by other pandemic-induced hardships-loss of household, lack of food, and mental illness-even after the receipt of UI. Findings on the intersection between race and ethnicity and UI suggest that pandemic policy interventions should be race conscious and consider the longstanding and systematic barriers experienced by minority mortgage borrowers. Supplementary information: The online version contains supplementary material available at 10.1007/s10901-022-10006-w.
Article
Full-text available
Introduction Existing studies have focused on the impact of economic development and urban expansion on public healthcare environment but has ignored the importance of regional integration. Regional integration reflects the spatial distribution of the labor force, which significantly affects healthcare workforce and healthcare infrastructure development. Methods Based on panel nested data for 137 cities in 16 major city clusters in China from 2001 to 2019, this paper assesses the impact of regional integration on the public healthcare environment through a hierarchical linear model (HLM). Results Our findings indicate that a 1% increase in regional integration leads to a 6.6 and 1.9% improvement in healthcare workforce and healthcare infrastructure. The results of the mechanism analysis indicate that regional integration affects the public healthcare environment through improving transportation infrastructure and industrial upgrading. In addition, regional integration has a stronger effect on cities with lower levels of economic development and healthcare environments. Finally, the endogeneity test based on the difference-in-difference (DID) model and the robustness test based on high-dimensional fixed effects model conduct the consistent conclusions. Discussion Policies to improve the public healthcare environment through promoting regional integration are proposed. Government should develop a more comprehensive regional cooperation plan to improve the public healthcare environment. Also, financial spending on improving the healthcare environment in peripheral cities should be increased. In addition, regional integration policy development needs to consider differences across regions.
ResearchGate has not been able to resolve any references for this publication.