Motivation: A predictor of PISA's mathematical competence beyond intelligence and prior test achievement

Assessing the predictive validity of expectancy theory for academic performance

BACKGROUND

Despite expectancy theory's widespread appeal and influence as a framework for motivation in organizational and educational settings, studies that have examined the theory's validity for performance-based outcomes, particularly with academic performance as the criterion, have been characterized by inconsistent results. Given numerous methodological concerns associated with past studies (e.g., prevalence of between-person rather than within-person design), we examined the predictive validity of expectancy theory for academic performance using methods that were consistent with the theory's original conceptualization. Additionally, we assessed the validity of the theory for students' study effort.

METHODS

The final sample included 123 undergraduate students who reported their final grades in four courses. Study effort and other variables were measured with self-report surveys. Because course grades were nested within each person, multilevel modeling was used to test study hypotheses.

RESULTS

Both the valence model and the force model predicted a student's current study effort, but contrary to expectations, neither model predicted a student's final course grades. In contrast, both valence for academic success and the simplified force model (based only on valence and expectancy) predicted current study effort, final course grades, and explained incremental variance beyond cognitive ability. Furthermore, the predictive validity of this force model was relatively stable across the 11 weeks of the study.

CONCLUSIONS

Based on methods congruent with expectancy theory's original framework, we find that the force model does not predict academic performance. An alternative version of the model, however, predicts course grades and has incremental validity over cognitive ability. Our results have several significant theoretical and practical implications.

Conscientiousness, Students' Goal Orientation, and Reasoning Ability: Significance for Educational Standards

Previous studies show that students' goal orientation and conscientiousness are related to academic performance. Few studies, however, allow conclusions to be drawn about the factor structure of goal orientation and its distinctions from conscientiousness. In a study with N = 145 secondary school students (M = 13.9, SD = 0.85; 41% male), we investigated if the residuals of latent factors of goal orientation are still meaningfully correlated with academic performance and reasoning. Based on structural equation models, we have replicated the theoretically derived four-factor structure and showed that conscientiousness explains 29% of the variance in learning goals and 40% of the variance in work avoidance. Furthermore, we show that the residuals of goal orientation are mainly not significantly related to reasoning or educational standards (only work avoidance correlated with reasoning, and performance goals correlated with educational standards). Educational standards were highly correlated with reasoning. Implications for school practice and possible interventions are discussed.

Motivation to succeed is not enough: motivated students need to know how to plan/organize their steps on their way to success

The study is based on dispositional (career motivation) and social-cognitive (generalized self-efficacy) theories of personality, further on the expectancy-value theory of achievement motivation and future time perspective theory (task value, time, and study environment). The study aimed to explain the mechanism of the prediction relationship between motivation and students' performance. It was assumed that skills of planning and organizing (operationalized as generalized self-efficacy and learning strategies) mediate the prediction of motivation (career motivation and task value) on students' success (operationalized as academic achievement and employability). In two studies (N = 313, N = 219), the hypotheses of the mediation models were supported by structural equation modeling. Generally, the skills of organizing/planning fully mediated the students' performance, measured as academic achievement and employability (number of employers). The results show the importance of combining dispositional motivation characteristics with dynamic planning skills on the way to students' success. Traditional psychological predictors of performance, like general mental ability and conscientiousness, were not controlled. Higher education institutions could support motivated students on their way to success by teaching them how to plan and organize specific steps on their way to success.

The relationship between achievement motivation and college students' general self-efficacy: A moderated mediation model

Objective

This study focused on the relationship between achievement motivation and college students' general self-efficacy, and aimed to explore the mechanism of action between achievement motivation and general self-efficacy.

Methods

Through convenience sampling, 1,076 college students were investigated from Anhui Province in China. Achievement motivation, general self-efficacy, perceived social support, and sports participation were evaluated using standard scales. For data analysis, Pearson's correlation analysis, structural equation model test, and bias-corrected percentile Bootstrap method were carried out.

Results

Common method biases can be accepted in this study. (1) Achievement motivation can directly affect general self-efficacy and make a positive prediction; (2) Perceived social support plays a mediating role between achievement motivation and general self-efficacy, that is, achievement motivation can indirectly affect general self-efficacy through perceived social support; (3) Sports participation plays a moderating role in the first half of the mediating path of "achievement motivation → perceived social support → general self-efficacy." The interaction between achievement motivation and sports participation affects perceived social support, and then indirectly affects general self-efficacy. In this moderated mediation model, The predictive effect of achievement motivation on perceived social support is significantly different among individuals with different levels of sports participation.

Conclusion

Perceived social support plays a part of mediating role between achievement motivation and college students' general self-efficacy, which is moderated by sports participation.

Types of Intelligence and Academic Performance: A Systematic Review and Meta-Analysis

The concept of intelligence has been extensively studied, undergoing an evolution from a unitary concept to a more elaborate and complex multidimensional one. In addition, several research studies have focused their efforts for decades on the study of intelligence as a predictor of academic performance of students at different educational stages, being a stable and highly relevant predictor along with other variables such as executive functions, social context, culture or parental guardianship. Thus, the present study, based on a systematic review and meta-analysis, includes 27 studies with a total sample of 42,061 individuals. The main objective was to analyse the relationship between intelligence and academic performance using different predictive models that include moderating variables such as country of origin, type of intelligence, gender and age. The findings of this research highlight the significant, positive and moderate relationship between intelligence and academic performance (r = 0.367; p < 0.001), highlighting the predictive capacity on school performance when the type of intelligence (general and implicit; 35%) or the country of origin (45%) is taken as a moderating variable, with the explanatory models on age or sex not being significant. Therefore, it can be concluded that intelligence, in addition to being a good predictor of academic performance, is influenced depending on the type of intelligence or theoretical model taken as a reference, and also depending on the country or culture of origin.

Perception of educational environment as a predictor of academic performance in physiotherapy students

BACKGROUND

Students' perception of their academic environment has a substantial influence on their actions, academic accomplishments, satisfaction, goal attainment, and behavior. The study was carried out to evaluate the perception of academic environment in physiotherapy students, and to identify the specific domains which could act as predictors of academic performance.

MATERIALS AND METHODS

This descriptive, cross-sectional research was conducted over 12 months using total population sampling. Three hundred and forty-three (139 postgraduate and 204 final year) final year and postgraduate physiotherapy students from five colleges of Mumbai and Navi Mumbai, who conformed to the inclusion benchmarks, participated in the study. Perception of educational environment was evaluated using Dundee Ready Education Environment Measure (DREEM). Academic performance was evaluated using the percentage obtained in the last university examination. Data were entered in MS Excel (©Microsoft, USA) and converted to Stata Version 15.1© StataCorp, College Station, for further analysis. The mean values and standard deviations for perception of educational environment were calculated and linear regression models were used to conduct multivariate analysis for each item in the questionnaire.

RESULTS

The mean total DREEM score of physiotherapy students was 129.97 ± 22.72. Students whose Students' Academic Self-perception scores were in the "Feeling more on the positive side" category showed a 4.17 point increase in percentage scores as compared with those who reported feeling like a total failure/had many negative aspects to their academic self-perception (95% confidence interval [CI]: 2.46, 5.89; P < 0.001). Students whose Students' Perception of Learning (SPoL) scores were in the "Teaching highly thought of" category showed a 2.75 point increase in percentage scores as compared with those who viewed teaching poorly/negatively (95% CI: 0.44, 5.06; P = 0.02).

CONCLUSION

Physiotherapy students had a "more positive than negative" view of the academic environment; however, a few "problematic areas" need to be remediated. Efforts to improve students' academic self-perception may improve their academic performance since this was the strongest predictor of their academic performance, followed by total DREEM scores and SPoL.

Anxiety, Motivation, and Competence in Mathematics and Reading for Children With and Without Learning Difficulties

Knowledge of the relations among learners' socio-emotional characteristics and competencies as they engage in mathematics and reading is limited, especially for children with academic difficulties. This study examined the relations between anxiety, motivation, and competence in mathematics and reading, within and across domains, in an academically-diverse set of 8-13-year-old learners (n = 146). To measure anxiety and motivation across domains, we paired existing measures of math anxiety and reading motivation with researcher-developed analogs for reading anxiety and math motivation. Participants completed standardized assessments of mathematics and reading, anxiety and motivation surveys for math and reading, and a measure of nonverbal cognitive ability. Results showed high internal consistency for all anxiety and motivation scales (Cronbach's alpha = 0.76-0.91). Pearson correlations showed that within and across domains, participants with higher competence had lower anxiety and higher motivation. Higher anxiety was also associated with lower motivation. Regression analyses showed that for both math and reading, within-domain motivation was a stronger predictor of competence than anxiety. There was a unidirectional across-domain relation: socio-emotional characteristics for reading predicted math competence, after accounting for nonverbal cognitive ability, age, gender, and within-domain anxiety and motivation. Results contribute to knowledge of the socio-emotional characteristics of children with and without learning difficulties in association with reading and math activities. Implications of a unidirectional socio-emotional link between the two domains can advance research and theory of the relations among socio-emotional characteristics and competence for academically-diverse learners.

Performance decline in a low-stakes test at age 15 and educational attainment at age 25: Cross-country longitudinal evidence

INTRODUCTION

Educational attainment is associated with important life outcomes including labour market performance, health status, well-being, civic and political participation. An important question is whether it is possible to identify early those students who lack the achievement motivation that is needed to complete a higher education degree.

METHODS

Longitudinal follow-ups of representative samples of participants in the 2000 and 2003 Programme for International Student Assessment (PISA) from Australia, Denmark and Switzerland (N = 3110; 1130; and 1962; age = 15 to 27; % females 51%, 51%, 49%; ethnicity/race unknown) were used to identify the association between a measure of effort on a cognitively demanding low-stake task at age 15 - performance decline during the test - and educational attainment at age 25-27.

RESULTS

A one SD difference in performance decline was associated with a 5-6 percentage point difference in the probability of obtaining tertiary-level qualifications (r = -0.15 in Australia; -0.11 in Denmark and -0.11 in Switzerland). We find no evidence of differences in this relationship across genders, socio-economic status and baseline levels of ability in the three countries. The association between performance decline and educational attainment is homogeneous across these groups. Self-reported measures of achievement motivation were not predictive of educational attainment in the three countries.

CONCLUSIONS

Our work contributes new longitudinal evidence to the body of research in education employing behavioural measures of motivation and engagement. It can be used to understand the potential long-term consequences of disparities in students' preparation to sustain effort over cognitively demanding tasks.

How do self-efficacy and self-concept impact mathematical achievement? The case of mathematical modelling

Background

According to the self‐enhancement perspective, self‐efficacy and self‐concept are shaped by prior achievement and have a crucial impact on future development. Their role in improving performance on challenging tasks, such as mathematical modelling (i.e., solving realistic problems mathematically), has barely been studied.

Aims

We investigated patterns of self‐efficacy and self‐concept and their predictive effects on mathematical modelling while taking into account school grades as measure of prior achievement and reasoning to reveal cognitive and motivational effects on achievement.

Sample

N = 279 secondary students in Grade 8 or 9 from 16 classes and 6 schools participated in the study.

Method

The multi‐informant design consisted of teachers’ reports of school grades, students’ reports of self‐efficacy and self‐concept (questionnaire‐based), and assessment of students’ reasoning and mathematical modelling.

Results

Using random‐intercept models, we found that the predictive effect of self‐efficacy on mathematical modelling withstood taking the school‐classroom‐related nested structure into account, whereas self‐concept lost its predictive value. Further, self‐efficacy fully mediated the effect of school grades on mathematical modelling.

Conclusions

In line with the self‐enhancement perspective on self‐efficacy, our findings highlight the strength of motivational effects on mathematical modelling. When we take the nested structure into account, our results indicate an impact of school grades via self‐efficacy on mathematical modelling independent of students’ cognitive level or classroom. Given the diverse challenges such complex tasks present, important pedagogical and didactical recommendations, such as targeting the enhancement of students’ self‐efficacy by teachers and educational decision makers, can be drawn.

Anxiety and Social Support as Predictors of Student Academic Motivation During the COVID-19

In this study we examined whether parents’ perceptions of students’ anxiety as well as perceived support from both teachers and classmates were predictive of changes in students’ academic motivation during the first wave of COVID-19. To this end, we used a retrospective pretest-posttest design together with a latent change score model to analyze our data. From April to May of 2020, 394 Portuguese parents of students in grades 1–9 participated in this study. Our results showed that students’ anxiety and teachers’ social support, as perceived by parents, were highly significant predictors of academic motivation changes. Specifically, we found a negative effect of anxiety and a positive effect of teachers’ social support on students’ academic motivation. Our results did not show, however, a significant predictive role of classmates’ social support. This study provides an important contribution to further understand the intrapersonal and interpersonal factors that are associated with the decline of students’ academic motivation during the COVID-19 pandemic. The pivotal role of teachers in sustaining students’ academic motivation and other relevant educational implications for the ongoing pandemic are discussed.

Students' Motivation and Affection Profiles and Their Relation to Mathematics Achievement, Persistence, and Behaviors

Objective

We aimed to explore profiles of subgroups of United States students based on their motivational and affective characteristics and investigate the differences in math-related behaviors, persistence, and math achievement across profiles.

Method

We used 1,464 United States students (male 743 51%, female 721 49%, age 15.82 ± 0.28) from PISA 2012 United States data in our study. First, we employed latent profile analysis and secondary clustering to identify subgroups of students based on motivational (math self-concept, interest in math, perceived control, and instrumental motivation) and affective factors (math anxiety). Next, we used regression to compare differences in math behavior, persistence, and achievement among all identified subgroups.

Results

We found five distinct groups of students with different patterns of motivation and affection. The subgroup of students with the lowest math anxiety and the highest motivation levels showed the highest math achievement and levels of persistence. The groups with high math interest, math self-concept, and instrumental motivation showed the most frequent math-related behaviors.

Conclusions

Our findings reveal the complexity of the students’ motivational and affective profiles. Our findings are significant for teachers and educators to understand the diversity of students and provide theoretical and practical support for individualized and differentiated instruction.

Parents' Perceptions of Student Academic Motivation During the COVID-19 Lockdown: A Cross-Country Comparison

The COVID-19 outbreak has ravaged all societal domains, including education. Home confinement, school closures, and distance learning impacted students, teachers, and parents’ lives worldwide. In this study, we aimed to examine the impact of COVID-19-related restrictions on Italian and Portuguese students’ academic motivation as well as investigate the possible buffering role of extracurricular activities. Following a retrospective pretest–posttest design, 567 parents (n_Italy = 173, n_Portugal = 394) reported on their children’s academic motivation and participation in extracurricular activities (grades 1 to 9). We used a multi-group latent change score model to compare Italian and Portuguese students’: (1) pre-COVID mean motivation scores; (2) rate of change in motivation; (3) individual variation in the rate of change in motivation; and (4) dependence of the rate of change on initial motivation scores. Estimates of latent change score models showed a decrease in students’ motivation both in Italy and in Portugal, although more pronounced in Italian students. Results also indicated that the decrease in students’ participation in extracurricular activities was associated with changes in academic motivation (i.e., students with a lower decrease in participation in extracurricular activities had also a lower decrease in motivation). Furthermore, students’ age was significantly associated with changes in motivation (i.e., older students had lower decrease). No significant associations were found for students’ gender nor for parents’ education. This study provides an important contribution to the study of students’ academic motivation during home confinement, school closures, and distance learning as restrictive measures adopted to contain a worldwide health emergency. We contend that teachers need to adopt motivation-enhancing practices as means to prevent the decline in academic motivation during exceptional situations.

Test-Taking Motivation in Education Students: Task Battery Order Affected Within-Test-Taker Effort and Importance

Different types of tasks exist, including tasks for research purposes or exams assessing knowledge. According to expectation-value theory, tests are related to different levels of effort and importance within a test taker. Test-taking effort and importance in students decreased over the course of high-stakes tests or low-stakes-tests in research on test-taking motivation. However, whether test-order changes affect effort, importance, and response processes of education students have seldomly been experimentally examined. We aimed to examine changes in effort and importance resulting from variations in test battery order and their relations to response processes. We employed an experimental design assessing N = 320 education students' test-taking effort and importance three times as well as their performance on cognitive ability tasks and a mock exam. Further relevant covariates were assessed once such as expectancies, test anxiety, and concentration. We randomly varied the order of the cognitive ability test and mock exam. The assumption of intraindividual changes in education students' effort and importance over the course of test taking was tested by one latent growth curve that separated data for each condition. In contrast to previous studies, responses and test response times were included in diffusion models for examining education students' response processes within the test-taking context. The results indicated intraindividual changes in education students' effort or importance depending on test order but similar mock-exam response processes. In particular effort did not decrease, when the cognitive ability test came first and the mock exam subsequently but significantly decreased, when the mock exam came first and the cognitive ability test subsequently. Diffusion modeling suggested differences in response processes (separation boundaries and estimated latent trait) on cognitive ability tasks suggesting higher motivational levels when the cognitive ability test came first than vice versa. The response processes on the mock exam tasks did not relate to condition.

Transfer effects of mathematical literacy: an integrative longitudinal study

Mathematical literacy (ML) is considered central to the application of mathematical knowledge in everyday life and thus is found in many comparative international educational standards. However, there exists barely any evidence about predictors and outcomes of ML having a lasting effect on achievement in nonmathematical domains. We drew on a large longitudinal sample of N = 4001 secondary school students in Grades 5 to 9 and tested for effects of ML on later academic achievement. We took prior achievement in different domains (information and communication technology literacy, scientific literacy, reading comprehension, and listening comprehension), socioeconomic status, and gender into account and investigated predictive effects of math grade, mathematical self-concept, reasoning, and prior achievement on ML. Using structural equation models, we found support for the importance of integrating multiple predictors and revealed a transfer effect of ML on achievement in different school domains. The findings highlight the importance of ML for school curricula and lasting educational decisions.

A Randomized Field Experiment Using Self-Reflection on School Behavior to Help Students in Secondary School Reach Their Performance Potential

Recent policy reports documented that a growing group of students in secondary education could perform better given their expected performance. Studies showed that school performance is related to a range of social–emotional factors, including self-awareness, self-management, social awareness, and responsible decision making. However, experimental studies in schools on the relation between these factors and school performance are scarce, and results are mixed. This study used a randomized field experiment to examine whether self-reflection on school behavior of underperforming secondary school students affected their school performance [grade point average (GPA)], school engagement, and self-concept. The sample comprised 337 ninth-grade students (M = 15.74 years old; SD = 0.58) from 18 secondary schools in Netherlands. The intervention was designed in co-creation with teachers, to be as close to school practice as possible. Underperformance was measured using achievement test scores from both primary and secondary school, supplemented with teacher and parental assessments. Different model specifications were estimated to perform the analyses and test for robustness of findings. The results showed that, for treatment compliance, students with higher school motivation were approximately 29% more likely to comply. Students who reported higher levels of self-concept of school tasks were 17% less likely to comply. No significant effects of the treatment were observed on students’ GPA, school motivation, hours spent on homework, or self-concept of school tasks. The treatment showed a negative effect on self-concept of leadership skills.

Longitudinal relations between South Korean adolescents' academic self-efficacy and values in mathematics and English

BACKGROUND

Students' motivation generally declines over time. Some researchers have suggested that the parallel decline in academic self-efficacy and values may be as a result of the longitudinal reciprocal relations between these two motivational constructs. However, little empirical evidence has supported this speculation. Further, all prior evidence has been provided based on samples of students from Western countries (Europe, United States).

AIMS

The current study was designed to examine the reciprocal relation between academic self-efficacy and values with a sample from another culture, namely South Korea.

SAMPLE

We used nationally representative longitudinal data of 6,908 students in seventh grade (M_age  = 12.83 years).

METHODS

We analysed the data tracking our sample from 7th grade to 11th grade. Latent cross-lagged models of academic self-efficacy and values in mathematics and English for 5 years were tested, while controlling for gender, achievement, and family income.

RESULTS

In both mathematics and English domains, there emerged significant unidirectional paths from prior values to later self-efficacy from Grades 8 to 11. For English, significant unidirectional paths from prior self-efficacy to later values additionally emerged from Grades 8 to 9. That is, significant reciprocity between self-efficacy and values was found for English from Grades 8 to 9.

CONCLUSIONS

Relatively consistent paths from prior values to later self-efficacy were identified among Korean adolescents, distinct from prior work focusing on students from Western countries. The results underscore the importance of considering different educational contexts and suggest the critical role of values in the development of Korean adolescents' academic self-efficacy.

The Importance of Students' Motivation for Their Academic Achievement - Replicating and Extending Previous Findings

Achievement motivation is not a single construct but rather subsumes a variety of different constructs like ability self-concepts, task values, goals, and achievement motives. The few existing studies that investigated diverse motivational constructs as predictors of school students’ academic achievement above and beyond students’ cognitive abilities and prior achievement showed that most motivational constructs predicted academic achievement beyond intelligence and that students’ ability self-concepts and task values are more powerful in predicting their achievement than goals and achievement motives. The aim of the present study was to investigate whether the reported previous findings can be replicated when ability self-concepts, task values, goals, and achievement motives are all assessed at the same level of specificity as the achievement criteria (e.g., hope for success in math and math grades). The sample comprised 345 11th and 12th grade students (M = 17.48 years old, SD = 1.06) from the highest academic track (Gymnasium) in Germany. Students self-reported their ability self-concepts, task values, goal orientations, and achievement motives in math, German, and school in general. Additionally, we assessed their intelligence and their current and prior Grade point average and grades in math and German. Relative weight analyses revealed that domain-specific ability self-concept, motives, task values and learning goals but not performance goals explained a significant amount of variance in grades above all other predictors of which ability self-concept was the strongest predictor. Results are discussed with respect to their implications for investigating motivational constructs with different theoretical foundation.

Evaluation des Rechtschreibtrainings mit Dino

Zusammenfassung. In der vorliegenden Studie wird das neu entwickelte Rechtschreibtraining mit Dino hinsichtlich seiner Auswirkungen auf die Lese-Rechtschreibkompetenz sowie motivationale Variablen evaluiert. 111 Kinder der dritten und vierten Jahrgangsstufe wurden randomisiert einer Interventions- ( n = 60) bzw. Kontrollgruppe ( n = 51 ) zugeteilt. Mitglieder der Interventionsgruppe wurden in Kleingruppen über einen etwa dreimonatigen Zeitraum von Studierenden gefördert. Varianzanalysen mit Messwiederholung zeigten, dass die Interventionsgruppe sich im Vergleich zur Kontrollgruppe signifikant stärker in ihrer Rechtschreibleistung und den Fähigkeitsselbstkonzepten Schreiben und Lesen verbesserte. Auch die Leseleistung nahm signifikant zu, wurde in der Kontrollgruppe jedoch nicht erfasst. Modelle latenter Veränderungsmessung zeigten positive Effekte der Intervention auf die latente Differenz im Fähigkeitsselbstkonzept Schreiben und die Differenz in der Rechtschreibekompetenz. Zudem lag ein positiver Zusammenhang der beiden Veränderungsvariablen vor. Neben dem Nachweis der Effektivität des Rechtschreibtrainings mit Dino leistet die Studie durch die Fokussierung motivationaler Variablen einen innovativen Beitrag in der Evaluationsforschung von Lese-Rechtschreibtrainings.

Mathematical Competence Scale (MCS) for Primary School: The Psychometric Properties and the Validation of an Instrument to Enhance the Sustainability of Talents Development through the Numeracy Skills Assessment

The adequate development of the numeracy skills is a target of the fourth of the Sustainable Development Goals and is considered the basis for a financial literacy: both are competences needed for successful social and professional inclusion. Building on these goals, we carried out a unidimensional Mathematical Competence Scale (MCS) for primary school. The aim of this study was to present the psychometric properties and the validation process of MCS, designed basing on Item Response Theory. The final version of the scale, which measures different domains of mathematical knowledge (Data Analysis and Relationships, Geometry, Dimensions and Measurements, Numbers and Calculations), was validated on the entire population of 2935 fourth graders in Ticino Canton, Switzerland. The results reveal the high level of correlation between the six mathematical dimensions and confirm the assumption of a latent “mathematical construct”. However, even the multidimensional model could be considered a good model because it fitted the data significantly better than the one-dimensional model. In particular, the differences of the deviance between the two models are significant (χ2 (20) = 642.66, p < 0.001). Moreover, findings show a significant gender effect and a positive correlation between students’ actual school performance during the same academic year and MCS scores. MCS allows a reading of the learning and teaching process in the perspective of the psychology of sustainability and sustainable development and helps a teacher to sustain student talent through the development of numeracy skills; in fact, the scale is intended both as an assessment tool and an innovative approach for shaping the development of curriculum, and therefore has potential to serve as a bridge between empirical research, classroom practice and a positive (school and professional) career development.

Modeling competence development in the presence of selection bias

A major challenge for representative longitudinal studies is panel attrition, because some respondents refuse to continue participating across all measurement waves. Depending on the nature of this selection process, statistical inferences based on the observed sample can be biased. Therefore, statistical analyses need to consider a missing-data mechanism. Because each missing-data model hinges on frequently untestable assumptions, sensitivity analyses are indispensable to gauging the robustness of statistical inferences. This article highlights contemporary approaches for applied researchers to acknowledge missing data in longitudinal, multilevel modeling and shows how sensitivity analyses can guide their interpretation. Using a representative sample of N = 13,417 German students, the development of mathematical competence across three years was examined by contrasting seven missing-data models, including listwise deletion, full-information maximum likelihood estimation, inverse probability weighting, multiple imputation, selection models, and pattern mixture models. These analyses identified strong selection effects related to various individual and context factors. Comparative analyses revealed that inverse probability weighting performed rather poorly in growth curve modeling. Moreover, school-specific effects should be acknowledged in missing-data models for educational data. Finally, we demonstrated how sensitivity analyses can be used to gauge the robustness of the identified effects.

Akademische Selbstkonzepte und akademische Selbstwirksamkeiten: Interdependenzen und Beziehungen zu schulischen Leistungen

Zusammenfassung. Selbstkonzept und Selbstwirksamkeit sind zwei wichtige pädagogisch-psychologische Konstrukte. Über ihre gegenseitige Beziehung und diskriminanten und inkrementellen Vorhersagevaliditäten für schulische Leistungen ist nur wenig bekannt, weil sie zumeist separat analysiert wurden (entweder Selbstkonzept oder Selbstwirksamkeit). Untersucht wurden die Interdependenzen von drei schulfachspezifischen akademischen Selbstkonzeptskalen und drei schulfachspezifischen akademischen Selbstwirksamkeitsskalen sowie deren statistische Vorhersagekraft für Schulleistungen (Zensuren). Die Fächer waren Englisch, Geschichte und Mathematik. Untersucht wurden N = 825 Jugendliche, die chinesische Mittelschulen besuchten (Alter: M = 14.46 Jahre, SD = 0.79). Konfirmatorische Faktorenanalysen (CFA) belegten einerseits in jedem Schulfach zwei separate akademische „Selbst“-Faktoren (Selbstkonzept und Selbstwirksamkeit) und andererseits innerhalb eines jeden „Selbst“-Konstrukts drei Schulfachfaktoren (Englisch, Geschichte und Mathematik). Eine gemeinsame CFA über alle „Selbst“-Items führte zu sechs prägnanten Dimensionen: drei fachspezifische Selbstkonzeptfaktoren und drei fachspezifische Selbstwirksamkeitsfaktoren. Sparsamere Alternativmodelle waren weniger gut mit den Daten vereinbar. Entsprechend zusammengestellte Skalen hatten gute psychometrische Kennwerte. Multiple Regressionsanalysen und Kommunalitätenanalysen zeigten, dass die akademischen Selbstkonzeptskalen die erfassten Schulzensuren wesentlich besser statistisch vorhersagten als akademischen Selbstwirksamkeitsskalen. Die von beiden „Selbst“-Skalen gemeinsam aufgeklärten (d. h. nicht weiter aufteilbaren) Schulleistungsvarianzen betrugen 29.9 % (Englisch), 17.0 % (Geschichte) und 31.5 % (Mathematik). Die spezifischen (d. h. nicht mit der jeweiligen anderen „Selbst“-Skala konfundierten) Varianzaufklärungen lagen für die akademischen Selbstkonzepte bzw. akademischen Selbstwirksamkeiten bei 5.3 % bzw. 3.4 % (Englisch), bei 8.0 % bzw. 0.4 % (Geschichte) und bei 8.1 %. bzw. 1.8 % (Mathematik). Die Diskussion betont die große Relevanz der Schulfächer für die konzeptuelle und empirische Differenzierbarkeit sowie die Validität der beiden „Selbst“-Konstrukte.

Explaining Social Disparities in Mathematical Achievement: The Role of Motivation

This study examined the role of motivation as a mediator of the relationship between parents’ socio–economic status (SES) and children's standardized test achievement in math. We employed a one–year longitudinal approach using Programme for International Student Assessment (PISA) 2003 and a follow–up exam in 2004. The sample consisted of N = 6020 German students (mean age 15.5 years, SD = .55) who continued school after Grade 9 (PISA 2003) and were in Grade 10 at the time of PISA 2004. Children completed measures related to their parents’ SES, math–specific self–concept, task–specific and global self–efficacy, and interest, intelligence and mathematical competence. We found a small to moderate correlation between parents’ SES and children's achievement. All motivational constructs partially mediated the relationship between father's SES as well as a family index for SES (economic, social, and cultural status) and children's mathematical competence, but only math–specific self–concept and self–efficacy were significant mediators for mother's SES. Even when simultaneously considering the mediating effect of children's intelligence and prior achievement, the mediation effects of motivation remained significant. These results are important for our understanding of educational equality. Copyright © 2016 European Association of Personality Psychology