Conference Program

Download the Book of Abstracts


Access to the virtual conference rooms requires a password (meeting code). The password was sent by e-mail to all registered participants on June 4th. If you have lost the password or do not get access, please contact us at



Tuesday | June 8, 2021
09:00 am -
09:15 am

Welcome and Opening of the NEPS Conference
Christian Aßmann (LIfBi)
>> Room A: >>

09:15 am -
10:00 am

Opening Lecture: Modeling cluster-level constructs with individual-level measures
Suzanne Jak (University of Amsterdam, NL)
>> Room A: >>

10:00 am -
10:15 am
10:15 am -
11:15 am

Symposium 1: It’s about time: Using logdata-based response times to address measurement issues in surveys & assessments
Chair: Tobias Deribo | Support: Jana Welling
>> Room A: >>

  • Studying the missing data mechanisms behind filtering rapid guessing in cognitive assessments (Deribo, Kröhne & Goldhammer)

  • Validating ability-related time components in reading tasks with unit structure (Engelhardt, Kroehne, Hahnel, Deribo & Goldhammer)

  • Mixture IRT identification of rapid responding using response times in questionnaires (Kroehne, Buchholz & Goldhammer)

Session 1
Chair: Eren Özberk | Support: Anna Scharl
>> Room B: >>

  • Modeling the dynamics between maths and reading skills with continuous-time models (Jindra, Sachse & Hecht)

  • Applying continuous-time modeling to PISA data: An illustration (Lohmann, Zitzmann & Hecht)

  • Using subgroup discovery and latent growth curve modeling to identify unusual educational trajectories (Kiefer, Langenberg, Lemmerich & Mayer)

Session 2
Chair: Anika Bela | Support: Eva Zink
>> Room C: >>

  • Company? Yes please! Using the NEPS to analyse why apprentices with Abitur decline university (Preböck & Annen)

  • The effect of teacher characteristics on students’ science achievement (Sancassani)

  • How stable is the relationship between education and class in Germany? Empirical distributions, counterfactual worlds, and a configurational analysis of NEPS data (Glaesser)

Networking Platform
Support: Sören Reimers
>> Room D: >>

11:15 am -
11:30 am
11:30 am -
12:30 am

Session 3
Chair: Timo Gnambs | Support: Jana Welling
>> Room A: >>

  • Comparing original and marginal trend estimates in large-scale assessment studies: Analytical derivations and a simulation study (Robitzsch & Lüdtke)

  • Analyzing PISA Mathematics 2000-2012: Evaluating the effects of the choice of calibration samples, item samples, estimation method, and linking method (Heine & Robitzsch)

  • The impact of measurement error for causal effect analysis: An illustration with NEPS data (Sengewald & Mayer)

Session 4
Chair: Ariane Würbach | Support: Anna Scharl
>> Room B: >>

  • Challenges of representativeness in survey research: An evaluation of the ERiK Surveys 2020 (Schacht, Gedon & Gilg)

  • Adjusting to the survey: How within-survey interviewer experience relates to interview duration (Pirralha, Haag, & von Maurice)

  • SES and gender bias in teacher assessments and their consequences for achievement inequality (Olczyk & Schneider)

Session 5
Chair: Christoph Homuth | Support: Eva Zink
>> Room C: >>

  • Labour market returns of ICT competences - status quo and methodological perspectives (Thürer & Annen)

  • Hidden Figures – Profiles and potentials of returns to education in working German adults (Reinwald & Annen)

  • A distributional analysis of gender gaps in wages and numeracy skills (Battisti, Fedorets, & Kinne)

Session 6
Chair: Claus H. Carstensen | Support: Sören Reimers
>> Room D: >>

  • Increasing test efficiency of the teacher knowledge survey module of TALIS 2024 through multidimensional adaptive testing (Fink & Frey)

  • The performance of item selection algorithms in Mokken scaling for various sample size: An empirical example with intelligence scale (Özberk)

  • Distributed Leadership and Alignment Optimization: A comparative, cross-cultural perspective across 40 countries (Eryilmaz)  +++ CANCELED +++

12:30 am -
01:00 pm
Lunch Break
01:00 pm -
01:45 pm

Announcement of winners of the NEPS Publication awards and short presentations of the award winners
Laudator: Christian Aßmann (LIfBi)
>> Room A: >>

01:45 pm -
02:00 pm
02:00 pm -
03:30 pm

Keynote Presentation: Longitudinal modeling with discrete and continuous latent variables
Matthias von Davier (Boston College, U.S.)
>> Room A: >>

03:30 pm -
04:00 pm
04:00 pm -
05:30 pm

Symposium 2: Using process data for investigating test-taking behavior
Chair: Esther Ulitzsch | Support: Jana Welling
>> Room A: >>

  • Under pressure: Measuring cognitive abilities under instruction-induced time pressure (Alfers, Gittler, Ulitzsch & Pohl)

  • How do examinees balance response time and accuracy on cognitive tests? Using mixture modeling to Investigate different test-taking strategies (Beverly, Loken & Weissman)

  • A multilevel mixture IRT framework for modeling response times as predictors or indicators of response engagement in IRT Models (Nagy & Ulitzsch)

  • Using sequence mining techniques for understanding incorrect behavioral patterns on interactive tasks (Ulitzsch, He & Pohl)

Session 7
Chair: Marie-Ann Sengewald | Support: Anna Scharl
>> Room B: >>

  • The achievement gap in reading competence: The effect of measurement non-invariance across school types (Rohm)

  • Unfairness, rater bias and further applications for DIF-approaches in large-scale assessment (Gürer & Draxler)

  • Measurement invariance: Dealing with the uncertainty in anchor item choice by Bayesian model averaging (Schulze & Pohl)

  • Borrowing historical information for the analysis of large-scale assessments (Kaplan)

Session 8
Chair: Sina Fackler | Support: Eva Zink
>> Room C: >>

  • Indicating information deficits at the start of university: A novel method to measure student’s level of informedness (Mouton & Ertl)

  • Social origin - Key concept and blackbox? (Siegel, Gröschl & Wohlkinger)

  • How can we measure young adults’ reading behavior: Is the title-recognition-test (TRT) a good alternative? (Pfost, Schnabel & Locher)

  • Comparing the construct validity of trait estimates of different response formats in the measurement of learning strategies (Tupac-Yupanqui, Heine, Schiepe-Tiska & Reiss)

Networking Platform
Support: Sören Reimers
>> Room D: >>



Tuesday | December 8, 2021
09:00 am - 10:30 am

Session 7
Chair: Christoph Homuth

Session 8
Chair: Florian Wohlkinger

Session 9
Chair: Gundula Zoch

10:45 am - 11:00 am

Announcement of the winners of the NEPS Publication Award
Christian Aßmann (LIfBi)

11:00 am - 11:45 am

Short presentation by the award winners

11:45 am - 12:30 am

Lunch Break

12:30 am - 02:00 pm

Session 10
Chair: Lena Nusser

Session 11
Chair: Anika Bela

Session 12
Chair: Annalisa Schnitzler

02:15 pm - 03:45 pm

Session 13
Chair: Claudia Traini

Session 14
Chair: Jacqueline Lettau

Session 15
Chair: Corinna Kleinert

04:15 pm - 05:30 pm

Roundtable „Meet the Research Data Center“

Session 1

Monday | December 07, 2020 | 11:15 am - 12:45 am (CET)

  • Germany-born children of immigrants: A mediation analysis of the language and socio-emotional skill pathway in intergenerational inequality
    Alejandra Rodríguez Sánchez (University of Cologne / Humboldt University Berlin)

    Previous studies suggest that migration background related differences in the verbal abilities and the socio-emotional development between the Germany-born children of immigrants and the children of German natives result from deficits in the parental investments of the former group. This assumption has led to the development of initiatives to improve the parenting skills of immigrant parents. However, whether parenting mediates the effects of growing up in often low-resource contexts and minority language homes is still open to question. Sociological and sociolinguistic perspectives contend that both parenting, verbal skills, and socio-emotional development are markers of parents’ social contexts and, therefore, a consequence of their socioeconomic circumstances, and not of parental deficits. Furthermore, the development of socio-emotional skills is fundamentally dependent upon language. This paper uses data from the National Educational Panel Study Starting Cohort 1, a random sample of n=3481 infants born between 2012 and 2013 in Germany, with a follow-up of over six consecutive years. Data from a standardized language assessment, and of a standardized test of executive control, an important dimension of socio-emotional development, are used in combination with various measures of parenting styles, practices, and investments. Joint mediation analysis is used to estimate, respectively, the direct effects of migration background on language and the mediated share going through parenting, as well as the mediated effect of both language and parenting on socio-emotional development. Results suggest parenting mediates a small share of the effects of migration background on children’s behavior and an even smaller one of the effects on language development. However, language is the most important mediating pathway between children’s migration background and executive function. Evidence suggests parenting interventions may be limited in their capacity to reduce gaps in socio-emotional behavior, unless the verbal ability gap among preschoolers, that are not mediated through parenting, are attended to.

  • Relationships between linguistic skills and mathematical learning in pre- and primary school age under consideration of the special role of working memory − Comprehensive results of data from Starting Cohort 2 (SC 2) of the National Educational Panel Study (NEPS)
    Nurit Viesel-Nordmeyer (TU Dortmund University)
    Ute Ritterfeld (TU Dortmund University)
    Wilfried Bos (TU Dortmund University)

    Based on comprehensive research results, linguistic skills can be assumed as direct prerequisites for mathematical learning (Schröder & Ritterfeld, 2014). Furthermore, connections between linguistic skills and mathematical learning can be suspected via working memory processes (Prado, 2018). Even though the involvement of individual components of working memory in mathematical (David, 2012) and linguistic (Archibald, 2016) learning processes could be identified, the relationship between all three components − linguistics, mathematics, working memory – in the context of mathematical learning seems to be not yet well understood. Especially, there is a lack of information detecting the relationship between all three components within the important developmental age span between pre-school and end of primary school age.
    Hence, we used longitudinal NEPS data from starting cohort 2 (age 4-10) to address this research gap by different sub-studies. First, we computed ANCOVAS to investigate the extent to which pre-school requirements in the domains of linguistic and/or mathematical skills have an influence on the development of both areas of competence. Further, the influence of working memory on the development of linguistic and mathematical skills was taken into account. Second, we calculated path and mediation analysis to discover underlying interdependencies between linguistic skills, mathematical skills and working memory within the mathematical learning process.
    The results of the different sub-studies point to a complex interplay between linguistic skills, mathematical skills and working memory within the mathematical learning process between pre- and primary school age. Especially the influences of linguistic skills on mathematical learning were manifold: On the one hand, linguistic skills seem to be a direct prerequisite for mathematical learning. On the other hand, linguistic skills appear to be necessary to convey the influence of working memory on mathematical learning. Ultimately, the generation of mathematical knowledge may even be supported by linguistic skills relieving the working memory performance.

  • The developmental trajectory of intrinsic reading motivation: Measurement invariance, group variations, and implications for reading proficiency
    Ai Miyamoto (University of Freiburg)

    The goal of the present study is to extend previous research on the developmental trajectory of intrinsic reading motivation during early adolescence. Using large-scale panel data on secondary school students in Germany, we examined: (1) the longitudinal measurement invariance of intrinsic reading motivation, (2) the generalizability of the developmental trajectory of intrinsic reading motivation across students’ gender, parental socioeconomic status (SES), and school tracks (academic vs. vocational), and (3) the associations between the developmental trajectory of intrinsic reading motivation and the developmental trajectory of reading proficiency. The scale we used to measure intrinsic reading motivation showed the (strict) measurement invariance across six occasions of measurement from Grades 5 to 10, indicating the high structural similarity (e.g., factor loadings, intercepts) of intrinsic reading motivation during early adolescence. Our analyses of latent growth curve models also confirm previous findings that students tend to experience a steady and significant linear decline in intrinsic reading motivation from Grades 5 to 10. This developmental decline also seems to be more pronounced in size than previously reported (∆ = −0.772, p < .001). The developmental decline in intrinsic reading motivation was observed irrespective of students’ gender, parental SES, and school tracks. Male students expressed lower mean-levels of intrinsic reading motivation across the waves and exhibited a steeper motivational decline compared to female students. Despite mean-level differences across the waves, students showed similar degrees of a motivational decline across parental SES and school tracks. Finally, the larger decline in students’ intrinsic reading motivation was associated with the smaller growth of their reading proficiency from Grades 5 to 10. Our study provides further support for the high prevalence of the developmental decline in intrinsic reading motivation during early adolescence, its generalizability across students’ demographic characteristics, and its implications for the development of reading proficiency.

back to top

Session 2

Monday | December 07, 2020 | 11:15 am - 12:45 am (CET)

  • Health inequality among college students. First results from the project “Understanding the institutional context of health inequalities among young people. A life stage approach”
    Christian Deindl (Heinrich-Heine-University Düsseldorf)
    Marvin Reuter (Heinrich-Heine-University Düsseldorf)
    Paula Matos Fialho (Heinrich-Heine-University Düsseldorf)
    Nico Dragano (Heinrich-Heine-University Düsseldorf)
    Claudia R. Pischke (Heinrich-Heine-University Düsseldorf)

    Socioeconomic differences in health among college students require explanations. One is that health inequalities at that age are partially explained by different exposure to health-related risk factors at earlier stages of the life course. Or, they may reflect preexisting inequalities in chronic conditions. But there are also processes and mechanisms that occur during this particular life stage. We argue that three general pathways underlie the association between socioeconomic differences and health at the individual level: a) behavioural, b) psychosocial and c) material. However systematic research on these pathways in this particular age group is incomplete, and even more so for Germany. We therefore answer the question: Do individual determinants (behavioural, psychosocial and material factors) contribute to the explanation of health inequalities among young adults aged 16 to 24 years?
    We apply random effects models to the National Educational Panel Study (NEPS) to disentangle the health inequalities among roughly 12‘000 college students (Ø28 years, 40% men) over a period of seven years (2010-2017). First results show that the majority in our sample is quite healthy. Nearly 90 % report being on very good or good health. The mean number of days students were affected by health problems was only 2.5. However, we still found substantial differences with regard to socioeconomic position, health behavior and material resources. Greater success and a higher satisfaction with one’s studies lead to better health. Of special relevance is the result that a higher occupational position of parents has a positive impact on student’s health. We therefore argue that social inequality matters even in a group with high social standing like college students.

  • Understanding university students’ dropout intentions: The impact of physical and mental health
    Tim Baalmann (Carl von Ossietzky University of Oldenburg)   - CANCELED -

    In the OECD countries, every third student leaves university without completing their degrees. Against the background of potential far-reaching negative effects of students’ dropouts on the individuals’ future career prospects, the search for causes and influencing factors has been going on for decades. One area in this complex web of causes is that of the non-cognitive prerequisites the students bring along. In addition to motivation, the students’ physical and mental health as a finite resource can be seen as a key factor. Theoretically, this paper builds on the value-expectation theory and argues that poorer health reduces the probability of obtaining a targeted degree as well as the expected benefits of studying. Moreover, to the extent that the burdens of studying itself contributes to poor health, the perceived costs of studying may increase too. Thus, this paper hypothesizes that a deteriorating health in the course of studies shifts the cost-benefit-considerations of studying towards a more negative perspective which should lead to higher dropout intentions.
    The empirical analysis is based on data from the LAST study (Life Course Perspective and Dropout from Higher Education). Starting in 2017, this four-wave panel survey analyzes the influence of different life domains on academic success and dropout intentions of undergraduate students at the Carl von Ossietzky University of Oldenburg, Germany (n = 7,169). The empirical analysis covers two steps: on a cross-sectional basis, OLS regressions with the pooled dataset are calculated. To examine longitudinal effects, Fixed Effects models are implemented. Results suggest that poor physical and mental health correlate with stronger intentions to leave university prematurely. This can be observed both cross-sectionally and longitudinally. In particular, emotional role functioning and mental health have significant impacts on students’ dropout intentions even when considering central control variables such as the social integration or the average school-leaving grade.

  • Academic and social integration and their relations to dropout in teacher education compared to other study programs
    Sebastian Franz (LIfBi - Leibniz Institute for Educational Trajectories, Bamberg)
    Jennifer Paetsch (University of Bamberg)

    Due to teacher shortage in many countries (Klemm und Zorn 2017; Kim und Corcoran 2018), it would be desirable that most of student teachers complete their studies successfully and then start working as teachers. In fact, dropout of teacher education students is a problem as it leads to fewer transitions into the teacher profession (Dupriez et al. 2016).
    An established explanation model for dropout in higher education originates from Tinto (1975). Following his student integration model, dropout is the result of a malintegration into the social and academic system of the university. Whereas contact to other students can be seen as an indicator for social integration; the subjective perspective of one’s own achievement in the study program is an aspect of academic integration. Although the model has been proven to explain the decision to drop out in many different contexts, it is yet little known if the effects are the same for teacher education students compared to students in other study programs (see Bohndick 2020).
    Teacher education students differ from students in other study programs concerning personal characteristics and their study conditions (e.g., Henoch et al. 2015). These differences might cause different associations of social and academic integration with dropout. Knowing more about the differences of this relation can help to reduce the dropout rate in teacher education.
    We draw on data from the student cohort of the national education panel study that includes an over-sample of teacher education students (Blossfeld et al. 2011). The analytical sample contains teacher education students and students in other study programs in similar field of studies (e.g., students in medicine and law are excluded). We use structure equation modelling with multi-group analysis to investigate differential effects of social and academic integration on the intention and the actual decision to drop out.

back to top

Session 3

Monday | December 07, 2020 | 11:15 am - 12:45 am (CET)

  • The effect of formal, non-formal and informal training on job tasks and the probability of automation: Evidence for workers at risk
    Birgit Zeyer-Gliozzo (Ruhr-University Bochum)

    The automation of job tasks due to technological change increases the pressure on employees whose workplaces consist largely of such activities. In this context, politics and science attach great importance to further training, although the benefits in terms of adapting job tasks for affected workers have hardly been investigated. Building on considerations of human capital theory and the task-based approach, this study examines the effect of training on job tasks and the probability of automation for employees who are at risk of automation. These are workers who a) work in routine intense jobs or b) have a high probability of automation, based on the expert assessment of the automatability of occupations by Frey and Osborne (2017). The study uses data from the National Educational Panel Study (NEPS), which provides the opportunity to compare the returns to different forms of further training, namely formal, non-formal and informal training. Fixed-effects models are estimated to account for unobserved heterogeneity. The results show that non-formal and informal training in the form of media use actually helps to reduce the intensity of routine tasks. The effects of training on analytic, interactive and manual tasks as well as the probability of automation differ depending on the type of training but are in many cases not significant. There is some evidence that high-skilled workers benefit more from non-formal training, while lower-skilled workers tend to obtain higher returns from informal learning. In addition, the degree of computerisation and job changes also seem to play a role in the extent to which training affects job tasks. Taken together, these findings suggest that policies to promote training in the context of technological change should take into account the different types of training, target groups and contexts.

  • Technological change and formal further education: Does automation risk influence aspirations and plans to pursue a new educational certificate?
    Martin Ehlert (WZB - Berlin Social Science Center)

    Recent technological developments suggest that many workers need retraining to maintain their employability. Automation using digital technologies may lead to changes in job tasks and even the disappearance of whole occupations. Especially in the latter case, workers need to learn new skills later in life to be able to move to emerging occupations. However, changing occupations is difficult in Germany because many occupations require specific formal certificates. Therefore, especially formal further education is important to help workers move from declining to emerging fields. Formal further education is defined as the attainment of a generally recognized educational certificate such as a vocational or university degree after entry into the labor market. Currently, only few adults on the German labor market engage in such programs despite a large number of occupations at risk of automation. This paper aims to explore whether the automation risk in the present occupation influences decisions to enter formal further education. To do this, I use the distinction between idealistic aspirations (unconstrained wishes) and realistic (constrained) intuitions to pursue formal further education. I aim to find out whether automation risk influences the idealistic aspiration to aim for a further formal certificate. Furthermore, I use the mismatch between idealistic and realistic aspiration to find out about the barriers to participation, especially among those who are most likely to be affected by automation. The analyses are based on NEPS SC6 data from wave 8 matched with data about occupation specific automation risks from the IAB. Preliminary results show that about 30% of employed adults report the idealistic aspiration to pursue a new certificate. However, descriptive statistics and multivariate regressions suggest that this is not influenced by automation risk. Furthermore, a mismatch between idealistic and realistic aspirations is more likely among poorer households, families with children, and less-educated workers.

  • Can we measure the digitization of the world of work? Evidence from the NEPS
    Teresa Friedrich (IAB - Institute for Employment Research, Nuremberg)
    Marie-Christine Laible (IAB - Institute for Employment Research, Nuremberg)
    Reinhard Pollak (GESIS – Leibniz Institute for the Social Sciences, Mannheim)
    Sebastian Schongen (WZB - Berlin Social Science Center)
    Benjamin Schulz (WZB - Berlin Social Science Center)
    Basha Vicari (IAB - Institute for Employment Research, Nuremberg)

    Technological change and automation has increased substantially in recent years and they have not only altered the communication and information systems, but also large parts of the working world. These fundamental changes also bring about a change in the training needs of employees; they create new employment opportunities, cause excessive demands or raise fear of job loss. The National Educational Panel Study (NEPS) have started to interview its adult respondents (NEPS-SC4 & NEPS SC6) with a survey instrument designed to measure the changes in the world of work, such as the use of digital technologies, and their perceived consequences, such as cognitive stress or fear losing employability.
    But how well can the new questions on measuring digitalization address the entire scope of the world of work? Which social groups are better or worse represented in terms of the consequences? We are investigating these questions in our methodological study. On the one hand, our analyses focus on criterion and construct validity of the new instruments. On the other hand, we focus on the group differences in terms of gender, age, education, ICT-literacy and the main task type of employment of the respondents. We show how these groups differ related to the subjective perception and change in the digitization of their workplace, their need for further training and their assessment of opportunities and risks caused by digitalization.
    Finally, we show which questions of social inequality can be examined empirically with this new instrument and how studies on the polarization of the world of work can be supplemented with these data for Germany.

back to top

Session 4

Monday | December 07, 2020 | 02:00 pm - 03:30 pm (CET)

  • No Matthew effects in competence growth over the school time, but the social origin achievement gap widens: A Bayesian multilevel approach using three German student cohorts
    Richard Nennstiel (University of Bern, Switzerland)

    The extent to which achievement gaps become wider or narrower in the course of schooling is a topic that is widely discussed in educational research and in the public. This study examines whether there are path-dependent and status-dependent cumulative advantages in competence development over the course of the school career.
    In other words, I try to answer the following questions. First, is the gap between the good and bad achieving pupils widening? Second, is the gap between social status groups widening?
    I investigate competence growth of three German student cohorts (National Educational Panel Study (NEPS), N = 12,311) within different stages of their school career (primary school (SC2), lower secondary school (SC3), and upper secondary school (SC4)).
    In order to take into account, the data structure and the federal organization of the German school system, I apply Bayesian multilevel models using weakly informative priors to estimate the effects of prior achievement and social origin on competence growth.
    Results consistently suggest that there is a strong negative association (-0.4 – -0.5) between prior achievement and subsequent growth. Additionally, I find that high SES pupils achieve higher competence gains compared to low SES pupils. Results favor the status-dependent cumulative advantage mechanism, but not the strict path-dependent cumulative advantages mechanism. The gap between the formerly good and bad achieving pupils is narrowing slightly. However, the gap between social classes is widening.

  • Understanding the response to high stakes incentives in primary education
    Maximilian Bach (ZEW – Leibniz Centre for European Economic Research, Mannheim)

    In this paper we provide the first causal evidence on how high-stakes incentives inherent in the German school system affect early educational achievement by inducing greater study effort. Specifically, we examine whether primary school students respond to incentives arising from early ability tracking–the allocation of students to different secondary school tracks based on previous achievement. We address the following questions: Does the need to perform well to qualify for a better school track increase study effort, and if so, does it translate into achievement gains?
    We use two complementary research designs and two separate data sources (IQB Bildungstrend and NEPS) to estimate the response of primary school students to tracking incentives. The first is a difference-in-differences (DiD) design that exploits the fact that two states have repealed binding track assignment policies relatively recently and gave parents free track choice. In a second research design, we leverage the fact that student performance is not graded until third grade and track recommendations are only based on performance in grade 4. As a result, the performance incentives from binding track assignment in secondary school are less salient in the lower grades of primary schools. To estimate effects, we thus compare test score gains between grades 2 and 4 in states with and without binding track assignment.
    We find across both research designs that tracking incentives have a substantial impact on academic achievement of primary school students. With free track choice, student test scores in grade 4 are reduced by between 0.10-0.14 standard deviations in math, 0.06-0.08 in reading, 0.09 in listening and 0.2 in orthography. Achievement effects can be found throughout the entire ability distribution.

  • Teacher effects in Germany: Evidence from elementary school
    María Daniela Araujo P. (University of Bamberg)
    Johanna Sophie Quis (University of Hannover)

    Identifying high-quality teachers who substantially contribute to student learning has been one of the main challenges faced by policy makers and researchers in education in recent decades. The value-added to student achievement model –derived from the education production function literature– has become a key tool for estimating the effects of individual teachers not only on students’ short-term academic success but also on later-life outcomes. We use data from the first elementary school grades of the Starting Cohort 2 of the German National Educational Panel Study (NEPS) to estimate teacher value-added to mathematical and early reading competence development. We specify our teacher value-added model with fixed as well as random effects. Both model specifications apply empirical Bayes shrinkage to adjust the teacher value-added estimates by their level of precision. Our results show substantial teacher effects or individual contribution to math and early reading competence development in the first grades of the German elementary school, and also substantial variation in this contribution. One standard deviation increase in teacher effectiveness or value added is associated with at least 17 percent of a standard deviation increase in student mathematical competence score, and at least 16 percent of a standard deviation increase in early reading competence score, over a year of instruction. In addition, we find that only few teacher observable characteristics modestly explain the teacher value-added estimated.

back to top

Session 5

Monday | December 07, 2020 | 02:00 pm - 03:30 pm (CET)

  • Personality trait differences between pre-service teachers with different majors
    Florian G. Hartmann (Bundeswehr University Munich)
    Bernhard Ertl (Bundeswehr University Munich)

    Pre-service teachers’ personality traits are seen as crucial for developing into a successful teacher since they determine how pre-service teachers are able to use the learning opportunities given by teacher education (Roloff Henoch et al., 2015). More general, personality characteristics such as the Big Five have been found to be associated with academic success and subjective as well as objective career success (e.g., Vedel, 2014). Here, some personality traits are especially important regarding positive career outcomes. Therefore it is important to investigate which entry characteristics can be observed in teacher education. Since personality traits also drive major choice it is likely that pre-service teachers with different majors differ in their personality and therefore have different resources that are available for mastering their studies. Previous studies have grouped majors only into a few broad categories when analyzing teacher students’ personality possibly blurring group differences within STEM and non-STEM fields (Ertl & Hartmann, 2019).
    The current study uses data of the NEPS first-year students starting cohort (SC5) to examine the Big Five of pre-service teachers with different majors in comparison to students not aspiring the teacher profession. The sample comprises 3305 female and 1684 male students across eight different majors (German language, Social Sciences, Economics, Mathematics, Physics, Chemistry, Biology, Geography). Of the sample 46% aspire the teacher profession.
    Results indicate that personality traits are associated to gender, teacher education (vs. non-teacher education) and major. While neuroticism and conscientiousness differentiate between female and male students, openness is associated to belonging to certain majors and extraversion clearly distinguishes between pre-service teachers and other students with pre-service teachers showing higher scores.
    The results are discussed with respect to the recruitment of pre-service teachers and regarding the development of strategies to compensate lacking favorable personality characteristics.

  • The effect of instrumental support on prospective teachers’ job satisfaction
    Stefanie Gäckle (DZHW - German Centre for Higher Education Research and Science Studies, Hannover)

    Teacher preparatory service in Germany is the “most important formal learning environment” for prospective teachers to develop professional competencies (Kunter et al. 2011a: 60). One facet of teacher competencies subsumes aspects of self-regulation. Occupational well-being or more specific, job satisfaction, belongs to these aspects (Klusmann et al., 2008: 704). According to Richter et al. job satisfaction can be linked to social support and its subdimensions (2011: 40; 2013: 174). Regarding the preparatory service in Germany, there are different sources from which support can arise. These are mentors (at schools), instructors (of the obligatory study seminar) and co-pre-service teachers.
    The research questions are therefore: How does instrumental support as one specific dimension of social support (Richter et al., 2011: 43) influence job satisfaction of prospective teachers? And do these effects differ depending on the source that provides support?
    Analyses are based on data from Starting Cohort “First-Year Students” of the German National Educational Panel Study (NEPS; Blossfeld et al., 2011; doi:10.5157/NEPS:SC5:14.0.0). The analysis sample contains n=1850 pre-service teachers. Measurement scales reach high internal consistency (.83 ≥ α ≥ .72); a CFA confirms the factorial structure of instrumental support by the different sources. A SEM reached good model fit (RMSEA=.034, CFI=.985, TLI=.980). The effect of instrumental support differs between the sources and only the effects by mentors (β=.17) and instructors (β=.10) are significant. To control if the duration of the ongoing preparatory service influences the effects, measurement invariance tests were conducted successfully (between groups).
    It is planned to add selected control variables into the SEM model in further analyses. Details of these analyses will be presented and the meaning for preparatory service will be discussed.

  • The development of occupational self-regulation at the beginning of a teaching career: A latent transition analysis
    Claudia Menge (DZHW - German Centre for Higher Education Research and Science Studies, Hannover)

    Occupational self-regulation is crucial for dealing effectively with stress at work. As a facet of teachers’ professional competence, it is considered to be a learnable skill (Baumert and Kunter 2013). However, previous research has shown a high stability of this attribute (Roloff Henoch et al. 2015). Against this background, the proposed contribution aims to investigate the stability over time of four distinct patterns of occupational self-regulation (H: healthy-ambitious, U: unambitious, A: exessively ambitious and R: resigned; see e. g., Klusmann et al. 2008; Menge and Schaeper 2019).
    The analyzes are based on data of n=401 participants from two online surveys (2016 and 2018) of starting cohort “First-Year Students” of the German National Education Panel Study, who transitioned from the second part of teacher education (also called preparatory service; see Klusmann et al. 2008 and Dicke et al. 2016) to work between the two measurement points.
    Occupational self-regulation is measured by a short scale of the Occupational Stress and Coping Inventory (AVEM) from Schaarschmidt and Fischer (2001) with two subscales each for occupational engagement and resilience to occupational stress. All four scales show good internal consistency (for each measurement point α ≥ 0.77).
    The factor structure of the instrument was confirmed by factor analyzes (T1: RMSEA=0.057; CFI=0.959; T2: RMSEA=0.061; CFLI=0.961). Tests of measurement invariance over time demonstrated (partial) scalar equivalence for three and partial metric equivalence for one subscale. A latent transition analysis (LTA) showed that 68% of the sample retained their original self-regulation pattern, whereby people with risk patterns are more likely to transition into health-promoting patterns than vice versa.
    In a final step, which is still pending, predictors for transitions in and out of risk patterns will be tested and implications will be discussed.

back to top

Session 6

Monday | December 07, 2020 | 02:00 pm - 03:30 pm (CET)

  • Gender disparities in further education – the role of family formation
    Gundula Zoch (LIfBi - Leibniz Institute for Educational Trajectories, Bamberg)

    In Germany, participation in non-formal further education courses is common, as these are mainly offered by employers. Men participate more frequently in such courses but these gender differences are often explained with other socio-demographic characteristics, such as education or place of residence. However, previous studies seem to have largely overlooked whether and to what extend gender disparities in further education are linked to family formation and the associated changes in maternal employment after childbirth. This study contributes to the literature on further training participation by (1) investigating the role of childbirth and associated changes in work and family life for gender inequalities in non-formal training activities from a long-term perspective; and (2) examining whether employment-oriented work-family policies are linked to gender disparities.
    Based on human capital theory, it is assumed that employers generally have less incentive to invest in firmspecific training for mothers because they expect interruptions in employment and more part-time work due to increased family responsibilities after the birth of a child. However, over the last decade, employment-oriented family policies have significantly increased maternal employment rates and shortened mothers’ family-related employment interruptions after child birth. It is therefore assumed that employers’ have increased their support for mothers’ participation in further education. Hence, gender disparities in further education should be smaller when supportive family policies are available.
    The analysis uses NEPS SC-6 data and applies fixed-effects panel models. Preliminary results show that in the first 10 years after child birth, mothers are less likely than fathers to participate in non-formal further training courses. These differences are particularly pronounced for job-related trainings. With increasing age of the child, mothers again participate more frequently – even more frequently than fathers. In addition, first results indicate that gender inequalities are smaller when employment oriented family policies enable an early return to work.

  • Does training beget training over the life course? On the gender-specific influence of true state dependence and unobserved heterogeneity on non-formal work-related further training participation among workers in Germany
    Sascha dos Santos (WZB - Berlin Social Science Center)
    Martin Ehlert (WZB - Berlin Social Science Center)

    Times of technological innovations require constant adaption to workplace- and occupational skill-requirements. Therefore, the importance of further training over the life course increases. However, previous studies showed that training participation is unequally distributed based on individual and workplace characteristics. Yet, it may be that previous training also plays a role because it facilitates and motivates further participation. So far, little is known about the dynamics of training participation over the life course. This paper investigates the persistence of job-related non-formal training participation over the life course among workers in Germany. The theory of skill formation by Cunha and Heckman (2007) predicts that previous educational investments should promote following educational investments. This is because skills attained at one stage augment later skill acquisition and thereby raises the productivity of skill investments. Therefore, further training participation at one stage may generate cumulative advantage because further participation is caused by previous participation (“true state dependence”). However, we assume that this differs between men and women because they follow different employment trajectories over the life course. We test these predictions using data from starting cohort 6 of the Germany National Educational Panel Survey (NEPS), which contains detailed information on learning and working trajectories of individuals born between 1944 and 1986 in Germany. We apply correlated dynamic random effects probit models (Rabe-Hesketh & Skrondal, 2013; Grotti & Cutuli, 2020, forthcoming) that allow to assess the causal effect of previous training participation on current training participation by controlling for unobserved heterogeneity. Our preliminary results reveal that there is substantial and statistically significant true state dependence for both genders, with a slightly higher effect size of previous training participation on later participation for women. Nevertheless, for both genders, unobserved heterogeneity remains the main driver concerning persistence in non-formal training participation.

  • Who cares when care closes? Care-arrangements and parental well-being during the COVID-19 pandemic in Germany
    Ann-Christin Bächmann (LIfBi - Leibniz Institute for Educational Trajectories LIfBi, Bamberg / IAB - Institute for Employment Research, Nuremberg)
    Basha Vicari (IAB - Institute for Employment Research, Nuremberg)
    Gundula Zoch (LIfBi - Leibniz Institute for Educational Trajectories, Bamberg)

    Around the world, the COVID-19 pandemic has led to the closure of formal childcare facilities and schools so that childcare has fallen back into the responsibility of families. This study examines the short-term consequences of this closure for working parents in Germany – a country dominated by a modernized breadwinner model. By applying multinomial logistic regression models to novel panel data from the National Educational Panel Study (NEPS-C), we analyse (1) the chosen informal care-arrangements among parents under control of pandemic-related altered working conditions and (2) resulting changes in parental well-being.
    To draw a comprehensive picture of the informal care arrangements in Germany during the first months of the crisis (May/June 2020), we focus our analyses on three types of families: working mothers with 14-years-old schoolchildren (NEPS-SC2), working mothers and fathers with under-14-years-old school and pre-schoolchildren (NEPS-SC6), as well as highly educated working mothers and fathers with under-14-years old school and pre-schoolchildren (NEPS-SC5). Based on the neoclassical economic theory and the resource-bargaining perspective, we expect that altered working conditions in the course of the COVID-19 pandemic will influence the ad-hoc division of family work. Specifically, we expect that parents will have greater bargaining power and thus be less involved in parental childcare if they work in a system-relevant occupation, have to work more hours or are unable to work remotely.
    Our analyses indicate that mothers continue to play a key role in the ad-hoc care-arrangements during the COVID-19 pandemic confirming the traditional division of family work in German couples. Moreover, the results illustrate the importance of working conditions, especially the possibility of remote work, in the bargaining processes of parents. However, contrary to our assumptions, parents’ well-being was not influenced by the chosen care arrangements during those turbulent months.

back to top

Conference Host

Leibniz Institute for Educational Trajectories (LIfBi)

Wilhelmsplatz 3
96047 Bamberg