Book

Hierarchical Linear Modeling: Guide and Applications

Authors:
... At Level 2, representing between-person factors, we included proenvironmental self-identity. Level 1 predictors were centred around the group mean, while Level 2 variables were centred around the grand mean 108 . To check the moderation effect we chose high and low levels of proenvironmental self-identity corresponding to one standard deviation unit below and above its average, respectively. ...
... To assess the models implemented in our study, we employed a full unconditioned model for comparison, utilizing a chi-square statistic 108 . Subsequently, we used the deviance value as the foundation for evaluating model fit. ...
... Subsequently, we used the deviance value as the foundation for evaluating model fit. A substantial reduction in deviance (-2LL) indicates a significant enhancement in model fit, while a minor reduction suggests insignificant improvement 108 . In the area of proenvironmental behaviors, some scholars showed that accounting for multiple proenvironmental behaviors using within-subject analyses across multiple behaviors and a longitudinal design was effective in predicting proenvironmental intentions and behaviors and tested the moderating effect of proenvironmental self-identity 109 . ...
Article
Full-text available
Previous research investigated the impact of psychosocial predictors (e.g. attitude, social and moral norm, perceived behavioral control, intention) on sustainable clothing purchasing. To date, no studies considered whether proenvironmental self-identity moderates the effects of these predictors on behavior. In this study, we adopted an intrapersonal approach and a longitudinal design to assess the moderating role of proenvironmental self-identity in predicting intentions and behaviors, considering gender differences. 250 participants completed an initial questionnaire on the predictors of three sustainable clothing purchasing. A month later, they filled out a second questionnaire to self-assess these behaviors. The results showed that social and internalized norms (moral norms) were notably influential of participants’ intentions. Affective attitude influenced behavior positively, while cognitive attitude had a negative influence. When considering the moderating role of proenvironmental self-identity, significant gender differences emerged. Women with a weak proenvironmental self-identity expressed a higher intention to purchase sustainable clothing when they had high affective attitudes and descriptive norm but low cognitive attitudes. Women with a strong proenvironmental self-identity intended to purchase sustainable clothing when they had high moral norms and cognitive attitudes but low descriptive norm. Man with a weak proenvironmental self-identity and high positive affective attitude increased their future SCP.
... HLM (Raudenbush & Bryk, 2002) dyadic analysis was conducted for research questions 2 and 3. Unlike traditional regression, HLM allows for the separation of the two levels of effects (level-1-individual predictor variables of attachment avoidance, anxiety and relationship mindfulness; level-2-couple factors) (Huta, 2014). For each dependent variable (ESM, ESP), the intercept model (null model without predictors), unconditional models (intercept with before-DWC or gender), and conditional models (with predictors) were created (Garson, 2013;McRae et al., 2014). In order to address multicollinearity, level-1 predictor variables were included in separate conditional models. ...
... In order to address multicollinearity, level-1 predictor variables were included in separate conditional models. Intraclass Correlation Coefficient (ICC; Lorah, 2018;Woltman et al., 2012), group-mean centering (Enders & Tofighi, 2007), full maximum likelihood (Garson, 2020), deviance statistics, and homogeneity tests (Garson, 2013) were performed. ...
... We generated hierarchical linear models for dependent variables, namely ESM and ESP, to understand the predictors of change during-DWC (Table 3). According to Garson (2013), a homogeneity test significance of p > 0.500 showed that residual variances did not differ significantly for the data across couples. The unconditional model with gender indicated an ICC of 11% between-couple variance for ESM (Table 3). ...
Article
Full-text available
In Gottman Couple Therapy (GCT), the intervention of Dreams-within-Conflict (DWC) helps break down a gridlocked issue between couples through deeper emotional expression and experiencing (in-counseling exploration of emotions). The current study examined experiencing in a single session of DWC for N = 30 individuals (15 couples) using multiple methods such as self-assessment questionnaires, observation rating and coding of the video recording, and feedback interviews. The before and during DWC best experiencing video segments were selected and rated by two raters independently on the experiencing scale (ES) for partners. The changes in experiencing mode and peak scores (ESM and ESP) during DWC were investigated in the presence of individual characteristics of attachment (anxiety and avoidance) and relationship mindfulness traits. A paired-samples t-test showed a significant increase in experiencing for both partners. Hierarchical linear modeling analysis indicated that gender (women) significantly and positively predicted ESM. ESP was predicted positively by gender (women) and negatively by avoidance, though the results were not conclusive. Thematic analysis was used to look at the Indian couples' experiencing as shared by them in order to better grasp the therapeutic implications. The qualitative findings confirm the quantitative results that couples outside of intervention utilized experiencing levels 1–3 predominantly and moved to 3–4 levels during best experiencing segments of intervention. Couples reviewed positively to the emotional experiencing techniques used during the DWC intervention.
... The primary benefit to mixed effects models is that they can handle data where the observations are not independent of each other (Garson, 2013). For example, in the modeling approaches described in the previous section, many of the assumptions are often violated in time series data of human behavior-especially around independence of observations and residuals. ...
... For example, in the modeling approaches described in the previous section, many of the assumptions are often violated in time series data of human behavior-especially around independence of observations and residuals. This can lead to the predictor variables being misinterpreted in magnitude and sometimes even direction (Garson, 2013). For examples already within the behavior analytic corpora, multilevel modeling has been used to analyze indifference data in discounting (e.g., Young, 2017), to model cigarette purchase task data (e.g., Zhao et al., 2016), and to model reinforcer preference following task completion in children (e.g., DeHart & Kaplan, 2019 11 ). ...
... 11 NB: This article also offers a fantastic single article overview and tutorial for how to do this kind of work. For those who read this article and find themselves ready to bathe fully in these crystal blue waters, we recommend Garson (2013). ...
... The data collected in the present study have a hierarchical structure, so the linear mixed effect model is appropriate. Linear mixed effect models are designed for multilevel analyses addressing hierarchical data (Garson, 2013). Garson (2013) uses employee relations as an example to explain LMMs: employee data are at level 1, agency data are at level 2, and department data are at level 3. ...
... Linear mixed effect models are designed for multilevel analyses addressing hierarchical data (Garson, 2013). Garson (2013) uses employee relations as an example to explain LMMs: employee data are at level 1, agency data are at level 2, and department data are at level 3. In this example, employee data is nested in agency data, and agency data is nested in department data. ...
Thesis
This study looks at the L2 and L3 perception of Quebec French (QF) tense and lax vowels [y, ʏ] and /e, ɛ/ and rounded vowels /y-u/ and /oe-ɔ/. Inspired by the Linguistic proximity model (LPM) (Westergaard, 2021), I predict that the trilingual participants will outperform the bilingual participants as the trilingual can transfer phonological features from both the L1 and the L2 to acquire L3 contrasts. The contrastive hierarchy theory, a representational and learning model proposed by Dresher (2009) is adopted to explain the sources of potential transfer in phonological acquisition. According to Dresher, phoneme inventories are best understood in relation to contrastive feature specifications, assigned in language-specific hierarchies. In a language-specific hierarchy, features are assigned to divide the inventory into smaller binary subsets until each phoneme is uniquely specified. The selection of the features is determined by examining the phonological processes in a given language (Dresher, 2009). The present study provides a comparison of the perceptual performance of four groups: (1) L1 Mandarin; L2 English; L3 QF (n=22), (2) L1 English; L2 QF (n=20) and (3) QF natives (NS) (n=20), (4) naïve learners (n=20). Two learner groups are at the upper-intermediate level of QF proficiency that was measured by self-rated background questionnaire (based on instructional hours and course level). The Mandarin speakers’ L2 English proficiency level was measured by IELTS (average 7.0). An ABX discrimination task (with 1500msISI) was conducted by embedding [y, ʏ] and /e, ɛ/ and /y-u/ and /oe-ɔ/ in CVC syllables ([bVb], [dVt], [sVz]) in a total of 120 trials. The primary findings of the study demonstrate that 1) The L3 QF learners are able to transfer [±front] > [±round] from L1 Mandarin and [±tense] from L2 English to successfully parse L3 QF tense and lax vowels [y, ʏ] and /e, ɛ/. 2) The L3 QF learners, transferring [±front] > [±round] from L1 Mandarin, are able to successfully parse /y-u/ and /oe-ɔ/. 3) Lack of [±round] in the English hierarchy and the transfer of L1 phonetic roundedness cue make the L2 QF learners misparse the rounded vowels and tense and lax vowels [y, ʏ]. 4) The contrastive hierarchy theory is able to predict the ease and difficulty of acquiring these contrasts based on different types of restructuring actions. 5) The present findings are in support of the LPM (Westergaard, 2021) and the Scalpel model (Slabakova, 2017) and reinforce the importance of developing a model in L3 phonology that takes contrastive hierarchy theory and restructuring principles into consideration.
... Another measure is RMSEA (Root Mean Square Error of Approximation), which is the root of the mean square error of approximation, which is a measure of the divergence of the model adjusted for its level of complexity, i.e., the number of parameters. The RMSEA value for the obtained model was 0.062, the recommended range (≤0.08); as a further diagnostic activity, no problems were found with the reliability of the measuring scales (high composite reliability-CR) [156]. The model fit with the empirical data is described by the following set of standard diagnostic measures; the values in parentheses show the recommended thresholds for models with an acceptable fit based on: [156]. ...
... The RMSEA value for the obtained model was 0.062, the recommended range (≤0.08); as a further diagnostic activity, no problems were found with the reliability of the measuring scales (high composite reliability-CR) [156]. The model fit with the empirical data is described by the following set of standard diagnostic measures; the values in parentheses show the recommended thresholds for models with an acceptable fit based on: [156]. The built model based on empirical research conducted (sample size n = 442; male = 183; female = 259) is the recursive model ( Figure 2). ...
Article
Full-text available
The sharing economy substitutes owning with accessing, promoting sustainable development by reducing excessive consumption and resource overuse, which harm the environment. Sharing reduces resource and energy use, lowering emissions and waste disposal costs, thus reducing environmental damage. This study identifies key factors that encourage Generation Z to embrace the sharing economy for goods and services, emphasizing its role in sustainable development. Conducted in May 2023, the study surveyed 442 Polish Generation Z individuals to examine their attitudes and behaviours regarding climate change. The research focused on this demographic due to their crucial role in addressing global issues. Data was collected using the CAWI method and analyzed with IBM SPSS and AMOS software through structural equation modelling (SEM). The analysis revealed three factors: Willingness to Share for Savings (WSS), Digital Customer Engagement (DCE), and Environmental Concern (EC). The results show that ecological concerns and digital engagement significantly influence people’s willingness to share, boosting environmental awareness and cost-saving behaviours. Generation Z’s sharing propensity and environmental consciousness are significantly shaped by digital engagement.
... It is suitable not only because the dependent variable is count data with overdispersion, a common feature of count data and a violation of the principal assumption of the Poisson regression model (Cameron & Trivedi 2009;Hilbe 2014), but also because it can test both fixed-and random-effects (Heck & Thomas 2020;Kwayu et al. 2020;StataCorp 2021). It addresses the issues of intra-cluster correlation of clustered count data (Garson 2013;Kwayu et al. 2020;Tseloni 1999) and omitted-variable bias (Sessions & Stevans 2006). In this case, when ridership counts from individual stops are nested in one census block group, these observations are not independent but rather are more similar due to the shared neighborhood characteristics that are relevant to transit demand. ...
... Moreover, neighborhood and transit service characteristics are often measured at different geographic levels. Such contextual and individual variables, if not treated properly, could lead to over-or under-estimates of model outcomes (Garson 2013). ...
Article
Full-text available
Although crime is well recognized as a factor detrimental to ridership, fewer empirical studies have tested the effect of crime on ridership and results are inconclusive. Moreover, existing studies seldom perform multilevel analysis despite using data with a hierarchical structure. This research addresses these gaps using the 2018 data of the five largest cities in the Texas Triangle and multilevel negative binomial regression. The results reveal a non-linear relationship between crime and ridership after controlling for other effects. Transit service characteristics and crime are the top predictors of ridership. The effect of transit trip rate on ridership varies contingent upon crime. The findings have significant implications for research on transit ridership and the improvement of transit services.
... The intercept model (null model without predictors) and unconditional model (intercept with before-DWC and gender variables) were generated for each dependent variable (EW, IW, EI, II) in HLM, with the couple as the grouping variable. Gender and before-DWC variables were treated as a simple level-1 fixed effects to arrive at unconditional models (Garson, 2013;McRae et al., 2014). We entered the level-1 variables (cognitive reappraisal, suppression, and emotion regulation beliefs) in separate conditional models as predictors of emotion regulation strategies, to address multicollinearity. ...
... When we tested two-level conditional models using predictor variables with during-DWC EI, the gender variable showed a coefficient of b = 0.238 (t (25) = 2.15, p = 0.041), cognitivereappraisal showed a coefficient of b = 0.203 (t (25) = 2.23, p = 0.035), and suppression showed a coefficient of b = 0.132 (t (25) = 2.06, p = 0.050). A homogeneity test significance of p > 0.500 indicated that residual variances did not differ significantly for the data across couples (Garson, 2013). ...
Article
Full-text available
Gottman Couple Therapy (GCT) is based on 40 + years of empirical findings and advocates process research, enabling an understanding of how an intervention works. Dreams-within-Conflict (DWC) is a GCT technique that softens the stand on unresolvable issues by facilitating positive emotion regulation strategies such as expressing vulnerabilities, understanding, and soothing in place of destructive strategies such as criticism and defensiveness. The aim of the study is to understand the emotion regulation process during a one-session DWC intervention using a convergent parallel mixed-methods design examining N = 30 individuals (15 couples) during the DWC intervention. The changes in emotion regulation strategies (Extrinsic/Intrinsic affect Worsening/Improving strategies–EW, IW, EI, II) in partners were examined in the presence of individual characteristics of emotion regulation traits (cognitive-reappraisal and suppression) and beliefs using self-assessment questionnaires, feedback reports, thematic coding of video recordings, and a semi-structured interview. Paired-samples t-test results showed that DWC fosters emotion regulation strategies by significantly decreasing partners’ EW and increasing EI and II strategies. Though IW strategies declined during-DWC, the changes were not significant. Hierarchical linear modeling findings showed that before-DWC emotion regulation strategies, gender, and individual emotion regulation traits of cognitive-reappraisal and suppression predicted EI, and before-DWC strategies predicted II, but none of the variables predicted EW and IW during-DWC. To further understand the interventional implications, the emotional regulation strategies and preferences for expression (over suppression) shared by the Indian couples were examined using thematic analysis. The results show that avoidance, conflict behaviors, and prioritizing parents’ emotions over partners’ (in men) were the most often employed regulatory strategies. Simultaneously, Indian couples unanimously agreed that expression of emotions was a crucial factor for marital satisfaction.
... Multilevel modeling adjusted the estimated standard errors, allowing for the clustering of observations within communities. This means respondents were nested within households and households nested within communities to account for cluster-level effects [34]. Model 1 was a null model with no covariates. ...
... The log-likelihood ratio (LLR) and Akaike's information criterion (AIC) tests were used to compare models with the highest log-likelihood and the lowest AIC indicating the best-fit model (see Table 3). Random effects were expressed in terms of community level variance, while the intra-class correlation coefficient (ICC) was used to examine clustering and the extent to which community/contextual factors explain the unexplained variance of the empty model [34]. All models were fitted at a 95% confidence level. ...
Article
Full-text available
Background Despite efforts from the government and developmental partners to eliminate gender-based violence, intimate partner violence (IPV) remains a pervasive global health and human rights problem, affecting up to 753 million women and girls globally. Few studies on IPV have focused on pregnant and parenting adolescent (PPA) girls in Africa, although the region has the highest rates of adolescent childbearing. This limited attention results in the neglect of pregnant and parenting adolescents in policies and interventions addressing IPV in the region. Our study examined IPV prevalence and its individual, household, and community-level correlates among pregnant and parenting adolescent girls (10–19 years) in Blantyre District, Malawi. Methods We collected data from a cross-section of pregnant and parenting adolescent girls (n = 669) between March and May 2021. The girls responded to questions on socio-demographic and household characteristics, lifetime experience of IPV (i.e., sexual, physical, and emotional violence), and community-level safety nets. We used multilevel mixed-effect logistic regression models to examine the individual, household, and community-level factors associated with IPV. Results The lifetime prevalence of IPV was 39.7% (n = 266), with more girls reporting emotional (28.8%) than physical (22.2%) and sexual (17.4%) violence. At the individual level, girls with secondary education (AOR: 1.72; 95% CI: 1.16–2.54), who engaged in transactional sex (AOR: 2.29; 95% CI: 1.35–3.89), and accepted wife-beating (AOR: 1.97; 95% CI: 1.27–3.08) were significantly more likely to experience IPV compared to those with no education/primary education, who never engaged in transactional sex and rejected wife beating. Girls aged 19 (AOR: 0.49; 95% CI: 0.27–0.87) were less likely to report IPV than those aged 13–16. At the household level, girls with fair and poor partner support had higher odds of experiencing IPV, but the effect size did not reach a significant level in the parsimonious model. A high perception of neighborhood safety was associated with a lower likelihood of experiencing IPV (AOR: 0.81; 95% CI: 0.69–0.95). Conclusion Intimate partner violence is rife among pregnant and parenting adolescent girls in Malawi, underscoring the need for appropriate interventions to curb the scourge. Interventions addressing IPV need to target younger adolescents, those engaging in transactional sex, and those having weaker community-level safety nets. Interventions to change social norms that drive the acceptance of gender-based violence are also warranted.
... Notably, although the outcome variable was not normally distributed, HLM 6.02 can fit hierarchical generalized linear models, which allows for the estimation of nonnormal response variables (McCoach, 2010). To make the results less biased, fixed effects with robust standard errors were reported to address the misspecification of the distribution of the dependent variable (Garson, 2013). 3 ...
... Intraclass correlation was calculated for each outcome variable, ranging from 0.01 to 0.13. Although the intraclass correlation was relatively small (Garson, 2013), a multilevel approach was used to properly account for the common variance shared at the group level (Nezlek, 2011;Pornprasertmanit et al., 2014). ...
Article
Full-text available
The purpose of this study was to examine which early childhood (EC) teachers’ qualifications (i.e., degree, major, and teaching experience) are linked to teachers’ Metacognitive Awareness (MA) and science teaching efficacy, and to investigate the relation among EC teachers’ MA components and science teaching efficacy. A total of 153 Head Start teachers from eight U.S. states completed validated surveys that measured their science teaching efficacy and MA. Results from multilevel ANOVA and regression analysis showed that teachers with an early childhood education background were more positive about their ability to teach science, more mindful of their teaching strategies, and more likely to self-evaluate their teaching as compared to teachers without an EC education background. Also, teachers who were more aware of their teaching strategies and instructional goals, and monitored their teaching practices reported higher confidence in their ability to teach science. Our results revealed the role of MA in early science teaching efficacy and highlighted the importance of supporting EC teachers’ professional development, particularly for those whose backgrounds are not in EC.
... Another critical concern in our research models is the endogeneity problem presented by our independent variables, which may be correlated with unobserved physician characteristics. However, the HLM estimation results help mitigate our concern over the endogeneity of physicians' integrity, benevolence, and outstanding ability within the team, as we controlled these endogeneity sources to a certain extent by our three-level model specifications, which include a series of fixed effects and random effects (Garson, 2013). ...
... Separate multi-level linear regression models (adjusted and unadjusted) were completed to estimate the association between household poverty and neighborhood poverty at age 8 with child life satisfaction and health at age 12. Clustering of observations within groups may lead to correlated error terms and biased estimates of regression coefficients (Garson, 2013). To account for clustering effects of children attending the same school, multi-level models were completed with children grouped at the school level. ...
Article
Full-text available
Objectives This study examined whether poverty (neighborhood and household) was associated with future health or life satisfaction outcomes and whether the association operated through social support (adult support at home, adult support at school, peer belonging), or differed by the immigration background (nonimmigrant family or immigrant family) of the family. Methods This study utilized a retrospective, longitudinal, population‐based cohort that included self‐reported survey data from the Middle Years Development Instrument (MDI) completed by children at age 9 and age 12, linked to administrative records. Participants included 5906 children in British Columbia, Canada. Neighborhood and household poverty were observed at age 8. Social support from adults and peers was self‐reported at age 9. Outcomes (overall health; life satisfaction) were self‐reported at age 12. Adjusted multi‐level multiple linear regression analyses and parallel mediation analyses were utilized. The interaction between poverty exposure and immigration background was also examined. Results Exposure to either poverty type was associated with lower levels of life satisfaction and overall health at age 12, though household poverty appeared to be associated with lower outcomes in comparison to neighborhood poverty. The indirect effects of poverty on outcomes appeared to operate primarily through adult support at home and peer belonging. Children in immigrant families had a larger negative association between neighborhood poverty and life satisfaction. Conclusions Household poverty had a larger negative association to outcomes in comparison to neighborhood poverty. The association of poverty to outcomes differed by immigration background and operated partially through adult support at home and peer belonging.
... Linear regression models are unsuitable for ESM data, due to its clustered nature (Carter & Emsley, 2019). Multilevel regressions extend linear models by allowing for variation within and between participants, thereby accounting for ESM clustering (Garson, 2013). This study employed two-level (observations nested within participants) as opposed to three-level models (observations nested within days nested within participants), as the latter has been shown to be suboptimal in the presence of autocorrelation (de Haan-Rietdijk, Kuppens, & Hamaker, 2016), which temporal networks suggest is typical of psychotic experiences (Contreras, Valiente, Heeren, & Bentall, 2020;Jongeneel et al., 2020). ...
Article
Full-text available
Background Post-traumatic stress disorder (PTSD) has been shown to predict psychotic symptomology. However, few studies have examined the relative contribution of PTSD compared to broader post-traumatic sequelae in maintaining psychosis. Complex PTSD (cPTSD), operationalized using ICD-11 criteria, includes core PTSD (intrusions, avoidance, hyperarousal) as well as additional “disturbances of self-organisation” (DSO; emotional dysregulation, interpersonal difficulties, negative self-concept) symptoms, more likely to be associated with complex trauma histories. It was hypothesized that DSOs would be associated with positive psychotic symptoms (paranoia, voices, and visions) in daily life, over and above core PTSD symptoms. Methods This study (N = 153) employed a baseline subsample of the Study of Trauma And Recovery (STAR), a clinical sample of participants with comorbid post-traumatic stress and psychosis symptoms. Core PTSD, DSO and psychosis symptoms were assessed up to 10 times per day at quasi-random intervals over six consecutive days using Experience Sampling Methodology. ResultsDSOs within the preceding 90 min predicted paranoia, voices, and visions at subsequent moments. These relationships persisted when controlling for core PTSD symptoms within this timeframe, which were themselves significant. The associations between DSOs and paranoia but not voices or visions, were significantly stronger than those between psychosis and core PTSD symptoms. Conclusions Consistent with an affective pathway to psychosis, the findings suggest that DSOs may be more important than core PTSD symptoms in maintaining psychotic experiences in daily life among people with comorbid psychosis and cPTSD, and indicate the potential importance of addressing broad post-traumatic sequelae in trauma-focused psychosis interventions.
... Specifically, multilevel modeling was used to explore how contextual, classroom-level factors, such as the various aspects of iSTEM, contribute to change in student attitudes toward STEM when student-level factors such as gender and race are considered. This research design and approach were used in order to determine predictive relationships between the dependent and independent variables without making strong causal inferences and to account for the nested structure of the data (students within classrooms) (Ferron et al., 2008;Garson, 2013;Snijders & Bosker, 2012). SPSS v. 29 was the primary statistical software used in the analyses. ...
Article
Full-text available
Integrated STEM education (iSTEM) is recognized for its potential to improve students’ scientific and mathematical knowledge, as well as to nurture positive attitudes toward STEM, which are essential for motivating students to consider STEM-related careers. While prior studies have examined the relationship between specific iSTEM activities or curricula and changes in student attitudes, research is lacking on how the aspects of iSTEM are operationalized and their influence on shifts in student attitudes towards STEM, especially when considering the role of demographic factors. Addressing this gap, our study applied multilevel modeling to analyze how different iSTEM aspects and demographic variables predict changes in student attitudes. Drawing on data from two five-year NSF-funded projects, we evaluated pre- and post-attitude survey responses from 948 students. Our analysis identified two key iSTEM aspects—relating content to students’ lives and engagement in engineering design—that significantly influence positive attitude change. The results highlight the importance of curriculum relevance and hands-on, problem-solving activities in shaping student attitudes. However, the impact of these instructional strategies varies across demographic groups. The study’s insights into the differential impact of iSTEM aspects on diverse student groups provide actionable guidance for educators, curriculum developers, and policymakers aiming to enhance STEM learning experiences and outcomes.
... The multilevel modeling technique considered three levels: individual characteristics (level-1), household characteristics (level-2), and built environment characteristics (level-3). Consequently, this study used a three-level nested random-intercept modeling approach in which intercepts were allowed to vary; therefore, the dependent variable was predicted by an intercept that varied across the groups (Cohen et al., 2013;Garson, 2013). The estimation of MMETRMs was conducted, using "xtmixed" command of the STATA statistical analysis software (StataCorp, 2011). ...
... ). Yet, the estimation of fixed effects and standardized errors remained reliable when the homogeneity assumption is not met (Garson, 2012;Wei & Low, 2017). Additionally, examinations of the histograms and normal Q-Q plot indicated normality in the distribution of Level 1 residuals. ...
Article
Full-text available
The current investigation examined multilevel models aimed at predicting two dimensions of homework emotion regulation: emotion management and cognitive reappraisal. Our models integrated key components from self-determination, expectancy-value, achievement goal, learning approaches, and self-regulatory or volitional perspectives. The investigation involved 1,282 Chinese middle school students. Results revealed that emotion regulation (i.e., emotion management or cognitive reappraisal) was associated with at least one variable for each of the five theoretical perspectives at the student level. Results at the class level revealed while cognitive reappraisal was positively linked to teacher autonomy support, it exhibited a negative association with homework quality. These findings underscore the significance of our models in unraveling the multifaceted nature of influences on emotion regulation during middle school homework. We discuss implications for both research and practice in homework in light of the results.
... All hypotheses were tested using multilevel modeling, because the data on the situational level are nested within persons. All multilevel models (also known as hierarchical linear models; Garson, 2013) were conducted using MPlus Version 8.6 (L. K. Muthén & Muthén, 1998. ...
Article
Full-text available
At universities, as in other educational settings, multitasking among students is a widespread phenomenon, leading educational staff to worry about its negative consequences. Using the experience sampling method and adopting a self-determination theory perspective, we examine the situative relationships of autonomous and controlled motivation for studying with multitasking, as well as students’ concentration and mood. In total, 138 undergraduate students answered up to six daily questionnaires for 7 days, yielding 1,770 responses during study situations. Results suggest that more autonomous study motivation is associated with less multitasking, while more controlled motivation is associated with more multitasking during studying. Additionally, we found that autonomous motivation is linked to higher levels of concentration and better mood, while controlled motivation is unrelated to concentration or mood. Furthermore, results show that in situations with multitasking in comparison to situations with monotasking concentration was lower, whereas affect did not differ. We used multilevel mediation analyses to test if the situational relationship of motivation and concentration/affect is mediated by multitasking. Mediation analyses did not reveal significant results, although the indirect effect of autonomous motivation on concentration via multitasking was marginally significant. Implications for research on multitasking and self-determination theory are discussed. Also, potential functionalities as well as costs and benefits of multitasking are considered.
... df = 27, p < .001). However, the estimation of fixed effects and standardized errors are robust to the violation of homogeneity assumption (Garson, 2012). Furthermore, normal Q-Q plot of level 1 residuals and histograms were reasonably normal. ...
Article
This study aimed to examine multilevel models posited to predict student perceptions of teacher feedback quality. A cross‐sectional survey design was used, involving 1072 middle school students. We incorporated two clusters of variables: (a) student characteristics (gender, prior knowledge, parent education, homework expectancy, homework value, homework cost, and help seeking) and (b) the characteristics of the classroom context (perceived homework quality, autonomy support, and teacher monitoring). Perceived feedback quality was positively related to perceived autonomy support and homework quality at the individual and class levels. Meanwhile, perceived feedback quality was positively related to homework expectancy, homework value, and help seeking at the individual level.
... df = 27, p < 0.001). However, the fixed effects and standardized errors could be estimated reliably even when homogeneity assumption is violated (Garson, 2012). In addition, histograms and normal Q-Q plot of Level 1 residuals were normal. ...
Article
The present study investigated multilevel models posited to predict student approaches to homework. Participants were 1,072 middle school students in China. Results revealed that deep and surface approaches were positively associated with performance-approach. Furthermore, deep approach to homework was associated negatively with homework cost, yet positively with mastery-approach, homework expectancy, and prior knowledge. Surface approach to homework was associated positively with homework cost and parent education, yet negatively with homework expectancy and mastery-approach. Females were less likely to use surface approach to homework than males. At the class level, surface approach to homework was negatively associated with parent education. Implications for homework practice and future investigation are discussed in terms of these results.
... The choice of lag structures has been explored for the identified controls to address this issue. By studying the autoregressive process of involuntary turnover, it is found that it follows a first-order autoregressive structure with homogenous variances (i.e., AR(1) structure) (Garson, 2013). ...
Article
Full-text available
In this study, we seek to understand firm behavior during times of crisis, with a particular focus on family firms in different contexts. We theorize that family control mitigates (i.e., negatively moderates) the relationship between economic crisis and the layoff of employees, resulting in a higher propensity of family firms to retain their employees during a crisis compared to their nonfamily counterparts. Furthermore, taking a closer look at family firms, based on their location, we argue that family firms in rural regions are more likely to adopt measures leading to involuntary job turnover than family firms in urban areas due to a higher sensitivity to the loss of socioemotional wealth following a business closure. Relying on a panel dataset of Swedish private firms active in the period 2004–2012, our study contributes to a better understanding of family firms as employers in different contexts.
... This strategy takes into account of clustering of households by geographic region also known as the primary sampling units (PSUs). For details, please refer to Garson (2013). Table 3. Multilevel models estimating the association between capital assets and food insecurity in Nepal, 2011. ...
Chapter
Full-text available
Food security is a multidimensional concept. According to the Global Food Security Index (GFSI) score (an indicator of food security), Nepal’s GFSI score in 2017 was 44.5 out of 100, ranking 81 out of 113 countries. This scenario shows that food security is Nepal’s challenge as well. Thus, this chapter examines the various socioeconomic, cultural, and structural factors that determine food security at the household level using the Sustainable Livelihood (SL) framework. The SL framework, an integrated approach to studying various livelihood outcomes, is widely used to examine the factors influencing food security in developing countries. To fulfill the aim of this chapter, we introduce the subject matter in the context of Nepal. We define the various concepts of food security and discuss its four major dimensions, and ways of measuring food security. We report on the food security situation of the country from a geographic perspective. Then, we illustrate the sustainable livelihood framework and associated components. We also provide theoretical and empirical links between various livelihood assets and food security along with a review and empirical evidence of the influence of the access to various assets on food security in the Nepali context. Finally, we conclude that food security is a geographically driven issue. We reveal that food security is a geographic problem in Nepal, which varies by north-south as well as east-west geographic regions. We also find that access to various household capitals is important in addressing food security challenges thereby suggesting the significance of household assets in solving the food security problem in the country.
... The explanatory variables were standardized (ztransformation, using the rescale function) to improve the linearity and comparability of coefficient estimates. We analyzed the suitable error distribution for the count data and the appropriate link functions (Garson, 2013). Frequent issues to be handled in count data are zeroinflation (e.g., Hartig, 2019) and overdispersion (causing incorrect standard errors, e.g., Bell & Grunwald, 2011, Meyer, 2021. ...
... The explanatory variables were standardized (ztransformation, using the rescale function) to improve the linearity and comparability of coefficient estimates. We analyzed the suitable error distribution for the count data and the appropriate link functions (Garson, 2013). Frequent issues to be handled in count data are zeroinflation (e.g., Hartig, 2019) and overdispersion (causing incorrect standard errors, e.g., Bell & Grunwald, 2011, Meyer, 2021. ...
Article
Full-text available
Synonyms are part of the scientific progression in taxonomy and nomenclature and reflect the evolving knowledge about species based on revisionary systematics. However, synonyms frequently cause problems in biodiversity repositories, so understanding the causes of the variation of botanical synonyms is essential. Recent studies attribute variation in synonyms to intrinsic and extrinsic drivers, such as nomenclature, taxonomic group membership (e.g., of orchids), and the age of the accepted name. Here, we examine the drivers of the synonyms for a large global subset of all angiosperms. Across 137,378 accepted names of 193 angiosperm families and 5,019 genera present in 355 botanical countries and regions worldwide, range size, the age of the accepted name, and insularity (insular or mainland occurrence, or occurrence on both) emerged as drivers with a positive effect on angiosperm synonyms. After accounting for these three factors, the residual differences in the number of botanical continents and the interaction between insularity and the range size became less significant. The combined multi-predictor model explained about 41% of the global variation in angiosperm synonymy (96%, including the random effects of the families, genera, and the presence patterns of accepted species on one or more botanical continents). We suggest that geographic distance between taxonomists enables wide-ranging species and species with insular distributions to accumulate more synonyms. Also, the age of an accepted name plays a vital role in synonym accumulation. Our results can help to set priorities in revising floras and checklists and to resolve synonymy problems in biodiversity databases, likely leading to more realistic global species numbers. As the drivers may also impact other plant taxa, the study likely has implications for a wider range of families and genera.
... Since multicollinearity is possible to occur in logistic regression, such occurrences should not change the estimates of the parameters, only their reliability [66]. According to the correlation matrix in Table 2, the coefficients showed no serious problem for multicollinearity among the explanatory variables used in the model. ...
Article
Full-text available
Based on traditional market theory, this study aims to investigate whether conventional market investment slopes affect the unconventional Bitcoin market, considering both normal conditions and crises. This study examines three main characteristics of the economy-intensive blockchain system, namely reliability, investment slopes, financial and accounting aspects that ultimately determine the confidence in the choice to invest in cryptocurrency. The analysis focuses on the study of the Bitcoin (BTC) investment slopes during January 2014-April 2023, considering the specifics of blockchain technology and the inferences of ethics, reliability and real-world data on investment Tas-sets in the context of conventional regulated markets. Using an econometric model that incorporates reliability analysis techniques, factorial comparisons and multinomial regression using economic crisis periods as a dummy variable, this study reveals important findings for practical and academic purposes. The results of this study show that the investment slopes of Bitcoin (BTC) are mostly predictable for downward trends, when statistically significant correlations with the investment slopes of conventional stock markets are observable. The moderate or high increase in performance slopes pose several challenges for predictive analysis, as they are influenced by other factors than conventional regulated market performance inferences. The results of this study are of intense interest to researchers and investors alike, as they demonstrate that investment slopes analysis sheds light on the intricacies of investment decisions, allowing a comprehensive assessment of both conventional markets and Bitcoin transactions.
... The results of the HLM analysis are shown in Figure 3. We confirmed that HLM analysis was appropriate for the model because the intraclass correlation suggests 54% of the variance in the governance disclosure score is explained at the country level and the remainder at the firm level (Model 1; Garson, 2013;Woltman et al., 2012). parency. ...
Article
Full-text available
This study empirically investigates the direct and interactive effects of firm‐level ethics policies and country‐level investor protection on firm corporate governance transparency. Using data on 9298 firms collected from the Bloomberg Terminal, we find that there is a positive relationship between country‐level investor protection and firm corporate governance transparency. The results also support the argument that firms with existing ethics policies exhibit greater corporate governance transparency. We also find that in countries with weaker investor protection, the impact of firm‐level ethics policies on corporate governance transparency is stronger. This study advances our understanding of the corporate governance transparency determinants, and the empirical evidence supports the notion that firm‐level factors such as ethics policies may compensate for the lack of formal national investor protection regulations.
... .γ 100 and the γ 00 of the variables at Level 1 in the intercept of Y ij are treated as fixed effects. This model explores whether the effect of city variables discovered in the null model may be attributed, in part, to the fact that different individuals are grouped under different city variables [73]. In general, constructing the Level 2 model aims to explore the predictors of the intercept and slope parameters, with the Level 1 coefficients (i.e., intercepts and slopes) serving as the dependent variables [66]. ...
Article
Full-text available
A large number of studies have provided evidence regarding the factors that influence commuting time. However, few studies have explored such effects in the context of considering spatial heterogeneity across cities, which limits the generalizability of the findings. This study addresses this gap by utilizing a dataset of 113 cities in China across the years 2014, 2016, 2018, and 2020. A two-level hierarchical linear model (HLM) was developed to explore the combined effects of city-level and individual-level factors on commuting time by constructing a nested “city-individual” relationship. The results show that (1) built environments at the city level significantly impact commuting time; (2) a non-linear association between population density and commuting time (U-shaped relationship) was identified, as well as between the number of buses and commuting time (inverted U-shaped relationship); (3) the urban construction land area and road area per capita exert negative effects on commuting time; (4) the impacts of individuals’ jobs–housing balance, travel allowances, and education on commuting time vary across cities. These findings might contribute to optimizing the design of a built environment, addressing the challenge posed by longer commuting times, and providing a better understanding of the effects of individuals’ characteristics on commuting time while considering the inherent differences across cities.
... Multi-level models (MLMs) were used to examine direct associations between neighborhood environment attributes (eight scales; independent variables) and each of the QoL outcomes (physical health, psychological health, social relationships, and environmental health; dependent variables), as well as the moderating effects of leisure PA (some leisure PA vs no leisure PA) and overall (total) PA (sufficient vs insufficient). MLMs can handle data where observations are not independent, correctly modeling correlated error, and account for dependency in error terms due to clustering (participants recruited from selected neighborhood units) (Garson, 2013). For main-effects of each QoL outcome, we estimated separate MLMs for each environment attribute (single environment-attribute models) adjusting for covariates and neighborhood types. ...
Article
Full-text available
Understanding how neighborhood environments are related to older adults’ quality of life (QoL) and physical activity (PA) is important for public health actions on healthy ageing in sub-Saharan Africa. We examined associations of perceived neighborhood environment attributes with QoL among older adults in Nigeria and investigated the moderating effects of PA on these associations. We conducted a cross-sectional study of 353 older adults (mean age = 68.9 ± 9.1 years) selected from 5 high- and low-income communities in Maiduguri, Nigeria. QoL, attributes of the neighborhood environments and PA were self-reported using validated questionnaires. Multi-level models were used to examine the direct associations between neighborhood environment attributes and each of the four domains of QoL (physical health, psychological health, social relationships, and environmental health), as well as the moderating effects of leisure-time and total PA. Seven of nine neighborhood environment features were positively associated with multiple domains of QoL. Residential density, land-use diversity, land-use mix-access, walking infrastructure, traffic safety and ‘overall walkability’ were positively related to both or either physical health and environmental health QoL among those who are physically active. In contrast, walking infrastructure, traffic safety, and ‘overall walkability’ were negatively related to psychological health QoL among those not physically active. Our findings suggest being physically active moderates the association of neighborhood environments with QoL among Nigerian older adults. We suggest that designing age-friendly communities and simultaneously promoting PA may be needed to improve QoL and help prepare the Nigerian society for the predicted increase in the older adult population.
... In order to refute the statistical hypothesis that the error relations had a permanent adjustment, the test for heteroscedasticity intended that previous error terms predisposed other error relations. However, the financial performance of mutual funds, which is the dependent variable, has an equivalent level of inconsistency for any value of the independent variables, according to the homoscedasticity hypothesis (Garson, 2012). A homoscedasticity test was used to look for consistency in the residuals from the regression equation. ...
Article
Full-text available
Several Kenyan mutual funds' financial performance has recently been on the downturn. Overall, variations in critical criteria reveal variable economic performance over time among Kenya's mutual funds. It was crucial to assess whether the portfolio performance pattern may be attributed to behavioral investing traps. The objective of this study was to determine how the disposition effect affected the financial performance, and how closely the fund size affected the relationship between behavioral investment traps and mutual funds' performance. The research method used was a causal research design. Data from a study panel were gathered. Over an eleven-year period, from 2011 to 2021, secondary data was obtained from mutual funds' previously published financial statements. Secondary data was acquired using the data collection tool. Data analysis was done with Stata software, version 15. The unit root test, a stationarity test, was carried out. Panel data regression was applied. The use of regression analysis with fixed and random effects was also carried out. The Levin-Lin-Chu test, the Augmented Dickey-Fuller test, the Im-Pesaran and Shin tests, the Philips-Peron test, and the Hadri 2000 test were all used to evaluate the validity and reliability of the data. Jarque-Bera test was employed to evaluate normality. In the panel analysis, the random effects model and the fixed effects model were separated using the Hausman test. The variables were distributed properly, as shown by the skewedness and kurtosis tests. There was no approaching multicollinearity among the variables, according to the pairwise correlation study. The results of this study revealed that disposition effect had a negative but substantial influence on financial performance of mutual funds in Kenya with a regression coefficient of -0.5455628. The study additionally found that fund size with a probability value of 0.1560 which was not significant. This therefore shows that fund size does not have a significant effect on the relationship between disposition effect and mutual funds financial performance. The results of this study demonstrated that the disposition effect had a negative but considerable impact on the financial performance of the Kenyan mutual fund. This suggested that the financial performance of mutual funds in Kenya is subject to a sizeable yet adverse disposition effect. According to the results of the multiple regression study, the financial performance of mutual funds in Kenya was negatively but significantly impacted by the disposition effect. But when institutional investors and fund managers are dissecting financial investing decisions, they should not use behavioral investment traps in isolation. Based on the findings, stakeholders should be aware of the information that fund managers in institutional investors are not insusceptible from behavioral biases arising from behavioral finance in the financial investment decision making processes.
... Son olarak, regresyon ve ANOVA'dan farklı olarak HLM analizinde bireylerden alınan ölçüm sayısının eşit olmasına gerek yoktur ve kayıp verilerin varlığında da HLM analizi yürütülebilmektedir. HLM analizleri; Sosyal Bilimler için İstatistik Programı (SPSS), LISREL 8'de çok seviyeli bir modül olan MLwiN, R ve HLM programları ile yapılabilmektedir (Garson, 2014;Nezlek, 2003). ...
... To analyze group-or team-level relationships, the research mostly aggregates team members' reports of leaders' transformational or charismatic behavior (Wang et al. 2011). Aggregating individuals' scores for a group score is usually reported on a methodological (Garson 2013), theoretical (consistent with the theoretical argument of other studies that leaders direct many of their transformational behaviors towards the entire group rather than to each individual) (Shamir et al. 1998), or empirical basis, when cases belong to known groups and there is intragroup congruency (Lopez-Zafra et al. 2017). Therefore, to incorporate the TFL perception scores, a direct consensus model was followed, which is a composition model in which individual ratings of higher-level phenomena are used to represent higher-level constructs (Chan 1998). ...
Article
Full-text available
In the business context, models are needed to facilitate our understanding on the emergence of processes that transcend the individual level. In the case of affective organizational commitment, such models are even more necessary, due to the benefits associated with affective organizational commitment at the organizational level. From a time-lagged multilevel perspective, a model to explain the emergence of affective organizational commitment was tested by integrating the contribution of group processes. In this study, at two time points, 63 work teams from different organizations and sectors in Spain (n = 233 employees) were evaluated for transformational leadership, workgroup emotional intelligence and affective organizational commitment. The data were analyzed by a multilevel structural equation modelling (MSEM). The results showed that supervisors’ transformational leadership style to both directly and indirectly (through workgroup emotional intelligence levels) mediates the development of affective organizational commitment at the individual level. The results are replicated at the team level but a direct relationship between transformational leadership and affective organizational commitment was not found. In conclusion, the results of this multilevel analysis of the relationships between transformational leadership, workgroup emotional intelligence, and affective organizational commitment contribute to the development of so-called “hybrid theories of homology” in the search for the generalization of relationships between variables across levels.
... Model 3 (χ 2 = 59.093, df = 27, p = .001). However, estimating standardized errors and fixed effects are robust to violations of the homogeneity assumption (Garson, 2012). Additionally, histograms and normal Q-Q plots of level 1 residuals were normal. ...
Article
The current investigation examined models of homework interest according to the data from middle school students in China. Homework interest was positively related to homework favorability, feedback quality, deep approach, monitoring motivation, peer interest, self‐concept, teacher control, and family homework help. Additionally, homework interest was negatively associated with the surface approach. At the class level, homework interest was related positively to feedback quality, yet negatively to parent education. Thus, the current study extended prior research on homework, by indicating that student interest in homework was further associated with peer interest and homework approaches. Implications for further investigation and practice are discussed based on these findings.
Article
Full-text available
Ecological environmental protection and poverty alleviation are of great significance for the study of human–land relationship coordination and sustainable development, and they have also been a focus of attention in China in the past few decades. In this study, we chose 13 contiguous poverty-stricken areas in China as the study area. Using MODIS Leaf Area Index (LAI) data from 2000 to 2020, the spatial–temporal changes in greenness were obtained using the Bayesian spatial–temporal model (BYM). Spatial autocorrelation was used to identify the spatial distribution of poverty using socio-economic statistical data. Driving factors, including natural factors, poverty factors, and the Grain for Green Policy (GTGP), and their influence on greenness were analyzed by using the Geodetector model for detecting spatial differentiation and factors’ interactions. The results showed the following: (1) In 13 contiguous poverty-stricken areas (CPSAs) in China, 59% of the area presented an increasing trend of greenness. (2) In 2000, the high poverty levels with larger MPI values were widely distributed. After 20 years, the overall MPI value was lower, except in some northwest regions with increased MPI values. The spatial autocorrelation of poverty, which relates to the mutual influence of poverty in adjacent areas, also decreased. (3) In the study area, 65.24% of the regions showed strong synergistic effect between greening progress and poverty reduction in the interaction between poverty status and green development. With the improvement of greenness level, the positive correlation between poverty alleviation and ecological environment improvement has become increasingly close. (4) The impacts of interaction factors with the highest q values changed from temperature interacting with precision to regional division interacting with the Grain for Green Policy. The conclusions are that from 2000 to 2020, the impact of natural factors, geographical division, and poverty status on greenness has shown a decreasing trend; The effect of the Grain for Green Policy is gradually increasing; At the same time, the interaction and overlapping effects between the Grain for Green Policy and poverty were increasing. Taking into account the needs of ecological environment, poverty alleviation, and rural revitalization, this research provides valuable reference for formulating and implementing relevant policies based on the actual situation in different regions to promote harmonious coexistence between human-land relationship.
Article
Full-text available
This article examines the relationship between school infrastructure and academic performance in Ecuador. The objective of this research is to identify which types of infrastructure are associated with better student outcomes in elementary schools. The study employs data from the 2019 UNESCO standardized test, ERCE-2019, for Ecuadorian primary schools. The findings indicate a significant positive correlation between various types of school infrastructure and student achievement. Such infrastructure includes art and music installations, on-site nursing facilities, and basic service connections, especially in rural areas. Significant correlations between educational outcomes and large-scale infrastructure investments are either not observed or are inconsistently evidenced. These results challenge the heavy focus on prominent educational infrastructure projects in Ecuador, suggesting an opportunity to reorient educational spending to enhance outcomes cost-effectively. These research findings may apply not only to Ecuador but potentially to the broader Latin American context.
Article
While many countries have made significant strides toward meeting the United Nations Sustainable Development Goals (SDGs), others continue to struggle to make progress. We connect progress toward the SDGs with capital market outcomes for firms going public. We find that the degree of underpricing in initial public offerings (IPOs) is lower in countries that are closer to achieving the SDGs, particularly those focused on economic and social outcomes. The relation between SDG progress and underpricing is sensitive to IPO characteristics associated with uncertainty, and country‐level indicators of democracy and cultural attributes. SDG progress is also associated with important IPO characteristics correlated with uncertainty, including offer prices, underwriter reputation, VC backing, offer size, and underwriter spreads. Together, our results support the hypothesis that SDG progress reduces uncertainty for investors, which lowers the cost of capital for IPO firms.
Article
The significance of green human resource management (GHRM) practices in promoting employees' environmentally friendly behaviors and enhancing environmental performance is well‐documented in current research. However, the impact of GHRM practices at the organizational level on the eco‐conscious behaviors of individual employees remains unexplored. Moreover, scholarly attention to employees' green innovative work behavior (GIWB) has been limited. This study explores the novel concept of GIWB from a strategic human capital (SHC) perspective. Drawing on signaling theory, this study proposes that GHRM practices act as signals emphasizing the importance of environmental sustainability, which may influence GIWB at the individual level. This research aims to assess the effect of various GHRM practices on GIWB and to explore how unit‐level GHRM practices moderate the relationship between an employee's creative personality (CP) and their GIWB, using social exchange theory (SET) as the guiding framework. Data were collected from 163 HR managers and plant supervisors and 815 employees within Taiwan's environmental protection sector and then analyzed using hierarchical linear modeling (HLM) techniques. The findings reveal that specific GHRM practices significantly enhance employees' GIWB. Furthermore, it demonstrates that aggregated GHRM practices moderate the relationship between CP and GIWB. These findings contribute to a deeper understanding of how signals from organizational GHRM practices can foster GIWB among individual employees.
Article
The term Proteus effect refers to the changes in attitudes and behavior induced by the characteristics of an embodied virtual agent. Whether the effect can extend to the moral sphere is currently unknown. To deal with this issue, we investigated if embodying virtual agents (i.e., avatars) with different characteristics modulate people's moral standards differentially. Participants were requested to embody an avatar resembling the Christian God in His anthropomorphic appearance or a control human avatar and to perform a text-based version of incidental and instrumental dilemmas in a virtual environment. For each participant, we recorded (1) chosen options (deontological vs. utilitarian), (2) decision times, (3) postdecision feelings, and (4) physiological reactions (skin conductance response and heart rate). We found that embodying God vs. a control avatar did not change the performance in the moral dilemma task, indicating that no strong Proteus effect was at play in our experimental conditions. We interpreted this result by examining the constraints and limitations of our task, reasoning about the necessary conditions for eliciting the Proteus effect, and discussing future developments and advances in the field. Moreover, we presented compelling effects concerning dilemma type, chosen option, personality traits, and religion affiliation, thus supporting and extending literature on decision making in moral dilemmas.
Article
Informed by multiple theoretical perspectives pertaining to academic procrastination, this investigation examined multilevel models aimed to predict homework procrastination using the data from 1,072 middle schoolers in China. Our model incorporated student gender, homework motivation, homework approach, homework behavior, and time allocated to extracurriculars outside of school. At the individual and class levels, homework procrastination was associated positively with homework cost and negatively with homework expectancy. Furthermore, at the individual level, homework procrastination was associated positively with surface approach and time on videogames, but negatively with homework effort. Implications for practice and further investigation(e.g. homework expectancy and surface approach) are discussed in relation to our findings
Preprint
Full-text available
This study examines the impact of per capita carbon consumption on life satisfaction and assesses how country-level policies to mitigate carbon affect this relationship. Using a sophisticated two-stage hierarchical linear modeling approach, we analyze panel data covering 155 countries from 2011-2022. We explore the sophisticated relationships between per capita CO2 emissions and average life satisfaction. First-level random effects identify significant factors affecting individual life satisfaction within each country. Second-level fixed effects examine the implementation of carbon mitigation policies, such as carbon taxes and carbon capture and storage technologies, at the national level in relation to life satisfaction. The results show the significant impact of life expectancy, average years of education, unemployment rate, and social peace on life satisfaction. Notably, there is a positive correlation between an individual's carbon footprint and a 0.038 increase in life satisfaction. In addition, the sustained implementation of carbon tax policies and emissions trading systems (ETS) contributes to a 0.037 increase in life satisfaction. This study provides valuable insights into the complex interplay between environmental policies and global subjective well-being, and highlights the potential long-term effectiveness of such policies in improving life satisfaction.
Article
This study draws on three well-established streams of leadership research: transformational leadership, servant leadership, and five-factor personality theory and applies them to the labor union context. The study tested the relationship between these theories and both leader and union effectiveness. The sample consisted of 240 respondents who rated the leadership of 60 participants in the 2018 and 2019 Harvard Trade Union Program. Results indicated that the best predictors of both leader effectiveness and union effectiveness were the transformational and servant leadership styles. In contrast with other research, the five-factor personality variables (extraversion, conscientiousness, emotional stability, openness, and agreeableness) had no significant relationship with leadership effectiveness after controlling for the leadership style models. Only two of the five personality variables, extraversion and emotional stability, were related to union effectiveness after controlling for the effects of transformational and servant leadership. Discussion includes implication for labor educators.
Article
Full-text available
Two studies were conducted to test the convergence of mass and interpersonal media processes and their effects on YouTube. The first study examined the influence of interpersonal interactions on video enjoyment. The results indicated that positive comment valence affected participants’ identification with the content creator, which then affected enjoyment of the video. To investigate the effects of convergence from a macro-level perspective, the second study tracked and recorded data from 32 YouTube videos for 34 days and recorded the following data for each video: number of views, likes, and comments/responses. The results indicated that the more content creators and users interact, the more likes the video receives. However, user-to-user interactions are associated with a decrease in the number of likes a video receives.
Article
Full-text available
Cambodian upper secondary education is divided into two tracks: science and social science. Students make their track selection in grade 10 and begin their enrollment in grade 11. The enrollment in the science track has witnessed a steep decline from 96% in 2014 to 34% in 2020, while the enrollment in the social science track has seen a remarkable surge from 4% in 2014 to 66% in 2020. This tendency poses a significant challenge to the government’s endeavors aimed at promoting workforces in science, technology, engineering, and mathematics (STEM) fields. This survey study aimed to examine factors influencing students to leave science-track classes. The survey was administered to 696 grade-12 students from 20 upper secondary schools. The two-level hierarchical linear modeling (HLM) was used for data analysis. The findings revealed that at the individual level, age, ease of national examination, high passing rates, preference for good grades, expense on private tutoring, STEM major choices at the territory level, attitudes toward science, parental advice, and family income significantly influenced students’ decision to opt for social science track over science track. At the school level, school location was a significant predictor of track choices. The findings were discussed with practical implications to improve enrollment in the science track at Cambodian upper secondary schools.
Article
Green entrepreneurial orientation (GEO) plays a pivotal role in reconciling economic growth with environmental sustainability. This research focuses on the individual green entrepreneurial orientation (IGEO) of mid‐level managers. In particular, it aims to explore how individual, environmental, and organizational factors foster IGEO. The data were collected from a total of 102 small‐ and medium‐sized enterprises (SMEs), as well as 510 mid‐level managers operating within Taiwan's environmental protection sector. Subsequently, the gathered data were analyzed using hierarchical linear modeling (HLM) techniques. The study found that green opportunity identification and entrepreneurial motivation notably influence various aspects of IGEO. In addition, elements like a green institutional environment, corporate green resources, and corporate environmental policies significantly boost IGEO. However, no significant links were found between a green institutional environment and green innovativeness or between corporate environmental policies and green proactiveness. This study provides a valuable contribution to GEO research, specifically focusing on SMEs and environmentally sustainable initiatives within developing economies.
Article
Purpose: To determine if a consumer co-designed infographic increased knowledge of physical activity and self-efficacy for exercise after total knee joint replacement surgery. Methods: Forty-four adults with primary knee joint replacement surgery were recruited from a public and a private hospital in Melbourne, Australia. Participants were randomly allocated to an experimental or control group. The experimental group received a consumer co-designed infographic. All participants received usual care. Primary outcome measures were knowledge of physical activity and self-efficacy for exercise. Outcomes were administered at baseline, week 1 and week 6. Semi-structured interviews with experimental group participants explored the acceptability, implementation and efficacy of the infographic. Results: There were no between-group differences for knowledge of physical activity at week 1 (MD -0.02 units, 95% CI -0.9 to 0.9) or week 6 (MD 0.01 units, 95% CI -0.9 to 0.9). Self-efficacy for exercise increased at week 1 (MD 14.2 units, 95% CI 2.9-25.4) but was not sustained. Qualitative data showed that the infographic was embraced by some participants but not by others. Conclusions: A consumer co-designed infographic did not improve knowledge of physical activity but may have had a short-term positive effect on self-efficacy for exercise after knee joint replacement. Trial registration ACTRN12621000910808.
Article
Full-text available
The article presents insights into holiday travel and its determinants in Poland. The purpose of the study was to analyze Polish citizens’ modal split and its determinants. Raw data from a pilot survey conducted in 2015 were used as the source material. To identify the determinants of travel mode choice for holiday trips, a multilevel multinomial logit model was utilized. This approach made it possible to include the hierarchical structure of the data, in which respondents are clustered within municipalities. The results reveal that apart from the decision-maker’s socioeconomic characteristics and household attributes, trip characteristics significantly determine Polish citizens’ choice of holiday travel mode. Moreover, the inclusion of municipality-level predictors substantially improved the accuracy of the model. The analysis revealed that the severity of the environmental consequences of motorized transport perceived by respondents also significantly influences their travel mode choice for holiday trips.
Article
Conversational agents (CAs) can analyze online conversations using natural language techniques and effectively facilitate group discussions by sending supervisory messages. However, if a CA makes imperfect interventions, users may stop trusting the CA and discontinue using it. In this study, we demonstrate how inaccurate interventions of a CA and a conversational repair strategy can influence user acceptance of the CA, members' participation in the discussion, perceived discussion experience between the members, and group performance. We built a CA that encourages the participation of members with low contributions in an online chat discussion in which a small group (3-6 members) performs a decision-making task. Two types of errors can occur when detecting under-contributing members: 1) false-positive (FP) errors happen when the CA falsely identifies a member as under-contributing and 2) false-negative (FN) errors occur when the CA misses detecting an under-contributing member. We designed a conversational repair strategy that gives users a chance to contest the detection results and the agent sends a correctional message if an error is detected. Through an online study with 175 participants, we found that participants who received FN error messages reported higher acceptance of the CA and better discussion experience, but participated less compared to those who received FP error messages. The conversational repair strategy moderated the effect of errors such as improving the perceived discussion experience of participants who received FP error messages. Based on our findings, we offer design implications for which model should be selected by practitioners between high precision (i.e., fewer FP errors) and high recall (i.e., fewer FN errors) models depending on the desired effects. When frequent FP errors are expected, we suggest using the conversational repair strategy to improve the perceived discussion experience.
ResearchGate has not been able to resolve any references for this publication.