Multilevel Latent Profile Analysis Research Articles

Determining the factors affecting the ICT use of Turkish students attending PISA-2018: Multilevel profile analysis approach

AbstractThis study aims to classify student profiles based on the type and frequency of Information and Communication Technologies (ICT) usage. Each profile exhibits homogeneous characteristics and heterogeneous characteristics compared to other groups. Additionally, the study investigates whether covariates at the school and student levels create differences in these profiles. For this purpose, a mixed-method analysis (Latent Profile Analysis (LPA), Multi-level Latent Profile Analysis (MLPA) and Multi-level Logistic Regression Analysis (MLRA)) were used to identify ICT usage profiles. The study analysed data from 6,890 Turkish students across 186 schools participating in PISA 2018 (Program for International Student Assessment), revealing three distinct ICT usage profiles: Profile-1 (Low-Level ICT Users (19.06%)), Profile-2 (Intermediate-Level ICT Users (71.10%)), and Profile-3 (Advanced-Level ICT Users (9.84%)). The MLRA showed that students with low ICT usage have less access to ICT resources. Students in the Low-Level ICT Users profile also had lower scores on ICT-related motivation factors (interest, competence, social interaction, and autonomy). No significant effect of ICT usage on mathematics, reading, and science scores was observed. Variance differences at student and school levels could not be fully explained by covariates at the school level (type of school, ratio of computers connected to the internet per student, lack of educational materials and personnel). However, the ratio of computers per student effectively explained these differences. According to an MLPA, schools were categorised into two groups based on ICT resources: 'high' and 'low'. In schools with high ICT resources, the proportion of students in Profile-2 was found to be 45.8%, while those in Profile-3 accounted for 9%. In contrast, in schools with low ICT resources, the percentage of students in Profile-2 was 22.45%, and in Profile-3, it was 1.82%. These findings indicate that the level of a school's ICT resources significantly impacts students' technology usage skills. A significant finding of the research is that most Turkish students use ICT at an intermediate level at home for school lessons. The COVID-19 pandemic has further increased the importance of studying at home using ICT resources.

Typologies of secondary school student academic resilience in science with classroom and school context predictors

ABSTRACT The education system strives to help students from low-income families achieve academic success. Academic resilience is related to not only individuals but also classrooms and schools. This study aimed to construct a comprehensive resilience model in science domains that presents the image of resilient students and describes the mechanisms by which they arise. This goal was achieved through a parametric multilevel latent profile analysis with level 1 and level 2 covariates. Latent profile indicators included variables reflecting motivation, subjective well-being, and engagement in science of secondary school students. Variables related to science classroom and school contexts constituted the levels 1 and 2 covariates. We used secondary data from Japan and Macao (China) collected in the 2015 Programme of International Student Assessment (PISA). The sample sizes were 1594 and 1114, respectively. We identified three resilience levels among students. The emergence of resilient students benefited from science classroom discipline, teacher-directed instruction, inquiry-based teaching, adjusting instruction, and teacher support. In Japan, the most robust predictor was inquiry-based teaching, while in Macao, it was teacher-directed instruction. Therefore, science teachers play a crucial role in building students’ resilience. It also suggests that a small class decreases the likelihood of students entering the high-level resilience group.

Students' regulation of anxiety and hope-A multilevel latent profile analysis.

Academic examinations are highly emotional for university students, making ER essential for preventing or reducing the negative consequences of negative emotions on well-being and academic performance. Initial theorizing and research suggest that flexibly applying combinations of strategies can promote successful ER. However, studies using intraindividual approaches to examine ER strategy use in specific contexts across multiple occasions, are lacking. Moreover, the combinations of strategies used by students within different contexts, and the adaptiveness of different strategies for regulating different emotions, remain unexplored. To address these gaps, we conducted an experience sampling study to identify patterns of students' momentary ER and to examine how context (achievement-related vs. nonachievement-related), emotions (anxiety vs. hope), and academic performance function as potential covariates. Over 200 university students rated their anxiety, hope, and use of eight ER strategies over a 7-day period, six times a day, prior to an important exam in 2016 and 2017. Results of a two-level latent profile analysis revealed distinct profiles of ER that differed on both levels. Intraindividually, ER patterns differed as a function of type of emotion and context experienced. More specifically, momentary use of multiple strategies tended to be associated with greater anxiety, while in the achievement context this association was reduced. Interindividually, students' tendencies to use different ER patterns were not related to test performance. Our findings suggest that ER strategy selection depends on both context and emotions, and advance ER research by considering intraindividual strategy use in concrete achievement situations. Yet the sample is not sufficiently representative. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Adolescents’ self-regulated and affective learning, teacher support and digital reading literacy: A multilevel latent profile approach

A considerable body of research has demonstrated the pivotal roles played in digital reading by self-regulated learning (SRL) and affective learning. Yet, whether teacher support facilitates SRL and affective learning has hitherto remained unclear. This cross-cultural study adopted multilevel latent profile analysis (LPA) to establish latent profiles of adolescents’ SRL and positive affect, explored how such profiles were differentially correlated to digital-reading performance, and examined how perceived teacher support predicted profile memberships, at both student and school levels. Using data collected from 34,000 students nested in 1,100 schools from three Asian and three Western educational systems by the Programme for International Student Assessment 2018, single-level LPA revealed three or four student profiles, in associations with disparate digital reading. While self-regulatory strategies were decisive in profile assignment, emotions did not vary significantly across profiles or predict performance. The least self-regulated learners often fell into a vicious circle of poor emotional well-being and low performance, but highly self-regulated and capable digital readers did not necessarily report the most positive emotions. Multilevel LPA identified two or three school profiles based on relative frequency of student profiles, in relations to different school-mean digital reading. Students’ perceived teacher support significantly predicted student-profile memberships; greater support increased likelihoods of being stronger self-regulated learners. But such support did not predict school-profile memberships. East-West cultural variations existed in all the influential patterns.

How do working students fare? A person‐centric approach to understanding patterns of work–school conflict and facilitation

AbstractStudent workers are largely understudied in organizational research, yet they represent an important part of the workforce. Their numbers are expected to rise as tuition continues to increase, and many adult workers are returning to school. The current study utilizes a person‐centric approach to investigate latent patterns of work–school conflict (WSC) and work–school facilitation (WSF) as well as the antecedent (i.e., supervisor work–school support) and outcomes (i.e., job satisfaction, well‐being, and school outcomes) tied to these latent patterns across school weeks. In a pilot study (n = 178 student workers), we first confirmed whether classes of WSC and WSF could be identified using latent profile analysis. Then, using 347 weekly WSC and WSF experiences, multilevel latent profile analysis identified four distinct profiles of WSC and WSF at the week level. Also, this modeling demonstrated that individuals tend to fluctuate between at least two profiles across weeks. Supervisor work–school support was found to predict weekly profile membership. Furthermore, the week‐level profiles exhibited distinct relations with weekly outcomes (i.e., job satisfaction, well‐being, and school preparedness). The results provide important theoretical and practical insights into the joint experience of interrole conflict and facilitation among student workers.

Multilevel Latent Profile Analysis of Global Competence in PISA 2018: Small Country-Level Differences

We examined the global competence of 15-year-old students based on PISA 2018 data ( N = 355,579 from 57 countries). Our multilevel latent profile analysis identified two groups at both individual and country levels, resulting in a total of four groups (2 x 2; “competent” and “less competent”). While individual-level differences were pronounced and consistent across all seven global competence indicators, the country-level differences were small overall, with only “attitude towards immigrants” and “respect for people from other cultures” showing somewhat salient group differences. There were more globally “less competent” students in 52 of 57 countries, suggesting the needs to improve students’ global competence in virtually all participating countries. Global competence was related to family socioeconomic status, academic achievement, and countries’ standing on the conservatism/liberalism dimension. We considered different ways to enhance global competence among students and society as a whole.

Profiles of teaching practices and teacher co-operation: A person-centered approach in multilevel context

In this study, teachers’ profiles in terms of their instructional and professional practices were defined by using multilevel latent profile analysis. For this study, Turkish national data from TALIS 2018 (N = 6861 teachers from 443 schools) were utilized. Five teacher profiles that differed especially in terms of “clarity of instruction” emerged. Members of Profile E are the teachers using instructional practices most, and the members of Profile A are the ones using these practices least frequently. Results indicated the important role of both the “school climate for innovativeness” and school location as contextual factors in predicting profile memberships.

Working and working out: Decision-making inputs connect daily work demands to physical exercise.

Work demands can undermine engagement in physical exercise, posing a threat to employee health and well-being. Integrating resource theories and a novel decision-making theory called the decision triangle, we propose that this effect may emerge because work stress changes the energetic and emotional processes people engage in when making decisions about exercise after work. Using diary-style data across two workweeks (N = 83 workers, 783 days), we used multilevel latent profile analysis to extract common decision input profiles, or daily configurations of energy and affect as key decision-making resources. Consistent with the decision triangle, three profiles emerged: visceral inputs (low energy/high negative affect), automatic inputs (low energy/low negative affect), and logical inputs (high energy/low negative affect). Daily job demands were highest among the visceral profile. In turn, the daily visceral profile related to the lowest likelihood of and intensity of physical exercise after work, especially relative to the daily logical profile. Whether or not those in the daily automatic profile exercised depended on their health orientation, or trait-level value of maintaining personal health. Our results support decision-making as a promising mechanism explaining the link between work demands and healthy leisure choices. Organizational interventions can target work stress, health orientation, or logical decision-making to promote frequent and vigorous employee physical exercise. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

'Prosociality' in Daily School Life and Early Adolescents' Peer Aggression: A Multilevel Latent Profile Analysis Approach.

We aimed to typify prosocial characteristics of aggressive youth. We classified early adolescents based on daily configurations of prosocial behavior and autonomous prosocial motivations (performing prosocial behavior for identified and intrinsic reasons) and controlled prosocial motivations (performing prosocial behavior for external and introjected reasons) and explored the links between the obtained sub-groups and peer aggression. The sample included 242 Israeli six-graders [Mage = 11.96 (SD = 0.18), 50% girls] and their teachers. At the daily level, adolescents self-reported on prosocial behaviors and their autonomous and controlled prosocial motivations for ten consecutive days. At the trait level, adolescents reported on global, reactive, and proactive peer aggression. Teachers reported on adolescents' global peer aggression. Using multilevel latent profile analysis, we identified four day-level profiles of prosociality: 'high prosocial autonomous' (39% of days), 'low prosocial' (35%), 'average prosocial controlled' (14%), and 'high prosocial bi-motivation' (13%). At the adolescent level, we identified four sub-groups, each characterized by one dominant daily profile: 'stable high autonomy' (33% of adolescents); 'stable high bi-motivation' (12%); 'often average controlled' (16%); 'often low' (39%). Higher self-reported aggressive adolescents, particularly proactive aggressive, had the least chance of being in the 'stable high autonomy' sub-group of all sub-groups. Teacher-reported aggressive adolescents had the least likelihood of being in the 'stable high autonomy' sub-group and the most likelihood of being in the 'often low' sub-group. In sum, peer aggression is a function of the configured phenomenology of prosocial behavior and motivations, with high prosocial autonomously motivated youth being the least aggressive.

Daily receiving and providing of social support at work: identifying support exchange patterns in hierarchical data

ABSTRACT The norm of reciprocity proposes that one who receives support feels obliged to return obtained benefits. Yet how employees regulate the mutual flow of social support with one another does not always follow a reciprocal dynamic, which may have different implications for employee outcomes based on whether social exchanges gain, drain or conserve resources. To better understand support exchange patterns and their relationship with basic need satisfaction (for autonomy, competence and relatedness) and emotional exhaustion in employees’ daily lives, we performed a multilevel latent profile analysis (N = 124 employees, 894 days). At the day level, we identified four support exchange profiles differing in low vs high received and provided social support (i.e., low-reciprocating, under-reciprocating, over-reciprocating and high-reciprocating days). At the person level, we identified three support exchange classes of employees, differing in the degree to which employees showed certain support exchange profiles over time (i.e., over-reciprocators, low-reciprocators and high-reciprocators). Pairwise comparisons with multinomial logistic regression revealed that over-reciprocating days were associated with the highest levels of need satisfaction for autonomy and relatedness and the lowest levels of emotional exhaustion. Moreover, over-reciprocators were most likely to satisfy their needs for autonomy. We discuss theoretical and practical contributions of our findings.

Exploring Teachers’ Instructional Practice Profiles: Do Distributed Leadership and Teacher Collaboration Make a Difference?

Purpose: While the literature includes multiple studies on the relationship between school leadership and instructional quality, they often use instructional practice as a continuous variable, assuming that a teacher would perform all sub-dimensions of instructional practice at a similar rate and failing to link distributed leadership to classroom teaching. Addressing these gaps in the literature, this study aims to identify teacher- and school-level latent profiles of teachers’ instructional practices and to investigate how distributed leadership predicts teachers’ membership in different instructional practice profiles, with the mediating role of teacher collaboration. Research Methods/Approach: The study employed a cross-sectional survey design using Türkiye's TALIS data for lower secondary education. Multilevel latent profile analysis with mediation modeling was conducted on data from 3,223 teachers in 192 schools. Findings: This analysis yielded four teacher profiles: laissez-faire, typical, controlling, and versatile; and two school profiles, high controlling and high laissez-faire. Findings indicate that distributed leadership promotes professional collaboration in lessons among teachers, which could, in turn, play a critical role in determining both individual teacher- and school-level profiles. Implications: This study provides practical contributions to understanding the nature of classroom teaching, suggesting that future studies should use instructional practice profiles instead of a single construct of teaching.

Profiles of Chinese preschoolers' academic and social-emotional development in relation to classroom quality: A multilevel latent profile approach.

Linking classroom quality to separate domains of child development might neglect the transactional interactions across developmental domains. This research utilized latent profiles across academic and social-emotional development to explore which aspects of classroom quality can predict children's profiles at the classroom level. Data were drawn from 96 preschool classrooms and 547 children (3-5 years old) in China in 2020. Multilevel latent profile analysis identified three profiles (entitled low-, average- and high-level development at the individual level), and two classes (entitled average and below-, average and above) at the classroom level. Multinominal logistic regression analyses revealed that instructional quality in math, science, and diversity, and the interactional quality in supporting children's learning and critical thinking, predicted children's profiles.

Does classroom matter? - A longitudinal multilevel perspective on students' achievement goal orientation profiles during lower secondary school

The present study adds to earlier person-oriented research by investigating differences in students' achievement goal orientation (AGO) profiles and their development using a simultaneous consideration of classroom patterns with longitudinal multilevel methods. The sample of almost 10,000 lower secondary school students, representing over 600 classrooms, was surveyed on their AGOs in the 7th and 9th grade. Multilevel latent profile analyses (MLPAs) and transition analysis (MLTA) revealed similar student profiles in AGOs in both grades: success-oriented, moderate multiple goals and avoidance-oriented, as well as two classroom types: success-oriented and mixed orientation classrooms with varied relative proportions of different student-level profiles and patterns of likely transitions. Stability of profiles was more typical than change. Maladaptive transitions were related to lower, and stable and adaptive transitions to higher GPA in the end of 9th grade. In success-oriented classrooms, it was more common to maintain or adopt the success-orientation across lower secondary school compared to the other classroom type.

Mentoring first-generation college students: Examining distinct relationship profiles based on interpersonal characteristics, support provision, and educational capital.

The goal of this study was to identify distinct profiles among first-generation college students' mentoring relationships based on interpersonal characteristics, provision of college-related support, and mentors' educational capital.First-year, first-generation undergraduates (n = 176) identified up to six mentors, rating various relationship characteristics, the types and degrees of college-related support each mentor provided, and each mentor's level of educational attainment. Ratings were used as indicators in a multilevel latent profile analysis among mentoring relationships (n = 254), accounting for clustering of multiple mentors within individual participants.A 3-profile solution best fit the data. Profiles were distinguished by closeness, frequency of contact, degree of college-related support provided, and mentors' educational attainment. Mentors with high educational attainment ("High-Capital Mentors") provided the most support for college-related issues, even with relatively infrequent contact.During their transition to higher education, first-generation college students appear to receive more active mentoring from adults with educational capital, although other adults may serve important functions not captured by the college-specific measures used in this study.

Supplemental Material for Students’ Regulation of Anxiety and Hope—A Multilevel Latent Profile Analysis

Supplemental Material for Students’ Regulation of Anxiety and Hope—A Multilevel Latent Profile Analysis

Early childhood educator's implementation readiness and intervention fidelity: Findings from a person-centered study

Research has indicated that early educators’ intervention fidelity is a significant predictor of children's responsiveness to classroom interventions. To improve understanding of predictors of intervention fidelity in early childhood settings, this study adopted a person-centered approach to identify profiles of “implementation readiness” in 1,192 Danish educators, and to examine relations with implementation fidelity. Multilevel latent profile analyses including setting-level characteristics as well as characteristics particular to the individual educator, showed reliable profiles of general and intervention-specific implementation readiness, which were associated with proportion of fulltime educators, employees with a teaching-related pre-service education, and investment in professional development. Higher and more positive implementation readiness predicted implementation dosage and adherence to early childhood interventions.

Intrapersonal and interpersonal psychosocial adjustment resources and achievement: A multilevel latent profile analysis of students and schools.

Intrapersonal and interpersonal psychosocial adjustment resources and achievement: A multilevel latent profile analysis of students and schools.

MULTILEVEL LATENT PROFILE ANALYSIS OF ESTONIAN SECONDARY SCHOOL STUDENTS' CAREER EXPECTATIONS WITH SCIENCE ACHIEVEMENT AND GENDER AS COVARIATES

Although much research has explored the personal, family, school, and social influences on adolescents' career expectations, little is known about the types of family-related career expectations of students and career expectation contexts in school. Using PISA 2018 test data, a multi-level latent profile analysis was conducted with indicators of parents' occupational status and children's career expectations. This research found that family-related career expectations of Estonian secondary school students can be divided into three types: enterprising, resilient, and disengaged. At the organizational level, contexts for career expectations in schools can be classified as enterprising, enterprising and resilient composite, and disengaged and resilient composite types. The research also found that science achievement and gender predicted latent profile memberships at the individual level. It is proposed that to value the science curriculum and its teaching is to invest in future human resources. It is suggested that families and schools should pay attention to stimulating male students' career aspirations. The career expectation types of secondary school students within and between schools are of great significance to the construction of national human resources, the development of schools, and the research of families. Keywords: career expectation types, multilevel latent profile analysis, PISA 2018 datasets, variations between schools, science achievement

Role of School Principals’ Technology Leadership Practices in Building a Learning Organization Culture in Public K-12 Schools

A principal’s technology leadership (TL) practices can facilitate information and communication technologies (ICT) integration into K-12 schools, alongside a drastic change in teachers' organizational behaviors to accompany this technological and cultural transformation process. However, there is little evidence of which TL practices contribute to such a holistic change in school culture; that is, the learning organization (LO) culture stimulating teachers' individual and collective professional learning orientations in tandem. The purpose of this paper is twofold. The first is to identify distinct profiles of LO culture in schools, based on teachers’ professional learning orientations. The second is to explore how a principal’s TL practices are related to teachers’ endorsement of these LO profiles. Data were drawn from 1105 teachers and 58 principals in 69 K-12 schools in Turkey. Multilevel latent profile analysis was used to uncover latent LO profiles with similar professional learning orientations. The resulting profiles were regressed on principals' TL practices, measured against the International Society for Technology in Education-Administrators (ISTE-A) standards. Teachers' professional learning orientations fell into high, moderate, and low-level LO profiles. The “systemic improvement” ISTE-A standard, including TL practices such as a principal’s data-driven decision making, school benchmarking, and technology-oriented staff employment, proved to be a significant predictor of a high-level LO profile. The school stage and teacher’s age were also found to be associated with that profile. Overall, we conclude that TL practices may have far-reaching implications for teachers’ professional learning orientations and school culture.

Individual- and school-level patterns of substance use and mental health symptoms in a population-based sample of secondary students: A multilevel latent profile analysis

BackgroundWhile substance use and mental health symptoms commonly co-occur among adolescents, few population-level studies have examined profiles of co-occurrence to inform tailored prevention and early interventions. MethodsA multilevel latent profile analysis was conducted on a representative sample of 11,994 students in 68 secondary schools to: 1) identify distinct profiles of co-occurring substance use and mental health symptoms; 2) identify types of schools based on student profiles; and 3) explore school correlates of student profiles and school types, including school climate, belonging, and safety. ResultsFive student profiles and three school types were identified. Among students, 57.6 % were in a low substance use and mental health profile, 22.5 % were in a high mental health but low substance use profile, 9.7 % were in a heavy drinking and cannabis use profile, 3.7 % were in a heavy drinking and smoking profile, and 6.5 % were in a high substance use and mental health profile. Positive school climate, belonging, and safety increased the odds of students being in the low profile, with belonging yielding larger effects among females. Among schools, 28 % had low, 57 % had moderate, and 15 % had high levels of student substance use and comorbid mental health symptoms. Rural schools were disproportionately represented in higher risk school types. ConclusionsThe identified student substance use and mental health symptom profiles can serve as targets for tailored prevention and early interventions. Results support examining school-based interventions targeting school climate, belonging, and safety with potential benefits to both substance use and mental health.