+++ Registration closed +++ Late registration via Contact +++
The titles, authors and abstracts of all paper presentations are listed below the program overview and can be accessed directly by clicking on the sessions. Later changes to the schedule cannot be completely ruled out.
Welcome
New developments in the NEPS study
Christian Aßmann & Cordula Artelt (LIfBi)
Session 1
Motherhood Penalty & Partnership Trajectories
Chair: Ilka Wolter
Session 2
Effects of Parental Practices & Involvement
Chair: Thorsten Schneider
Session 3
School-to-Training & Work Transitions
Chair: Corinna Kleinert
Lunch Break with Poster Presentations
NEPS Publication Award
Announcement of the winner(s) and honorary presentation
Christian Aßmann (LIfBi)
Session 4
School-Related Long-Term Effects
Chair: Kamila Cygan-Rehm
Session 5
School-Related Problems & Grievances
Chair: Andreas Horr
Session 6
VET-Related Group Differences
Chair: Annalisa Schnitzler
Session 7
Longitudinal Perspectives on School Tracking
Chair: Christoph Homuth
Session 8
External Influences on School Performance
Chair: Lydia Kleine
Session 9
Transitions to VET & First Jobs
Chair: Katharina Weßling
Session 10
Family Background & Educat. Performance
Chair: Jan Skopek
Session 11
Educational & Occupational Aspirations
Chair: Melanie Fischer-Browne
Session 12
Dropouts, Contract Refusals & Job Stress
Chair: Stefanie Gäckle
Session 13
Family & Reform Influences on Study Behavior
Chair: Sebastian Franz
Session 14
Attainments in Higher Education
Chair: Christina Haas
Session 15
Effects of Digitization & Task Change
Chair: Michaela Sixt
Keynote & LIfBi Lecture [PDF]
Subsidized small jobs and maternal labor market outcomes in the long run
Regina T. Riphahn (FAU, Erlangen-Nuremberg)
Session 16
Political Issues & Trust in Media
Chair: Saikat Ghosh
Session 17
Studying, Teaching & Informal Learning
Chair: Jacqueline Kroh
Session 18
Further Education and Work-Related Training
Chair: Markus Nester
Virtual Get Together
Informal Conference Closing and "Meet the Research Data Center"
Previous research documents motherhood penalties for wages, career progression and hiring. Despite a rising literature indicating the importance of work content to employees’ career outcomes, implications of motherhood for job tasks are unknown. Hence, this study is the first to examine whether there is motherhood penalty regarding work content with a longitudinal perspective. In addition, we investigate how the motherhood penalty in job tasks varies by length of leave, level of education and occupational mobility. Drawing on the literature on job tasks and on motherhood penalties, we formulate hypotheses on the consequences of motherhood for advantageous job tasks such as nonroutine analytic and managerial tasks compared to less advantageous job tasks, e.g. routine manual job tasks. We use data from the Adult Cohort of the German National Educational Panel Study (NEPS; N=1,395) that include a job-tasks module in a four-year interval (2011/12, 2015/16 and 2019/2000). Results from individual fixed-effects regressions show that after childbirth, women execute considerably less nonroutine analytical and interactive as well as routine analytical job tasks. For tasks which are in the middle of the task distribution we also find clear declines. These patterns are exacerbated for mothers who took lengthy leaves, who are occupationally mobile upon return to the labor market and have lower levels of educational attainment. At the low end of the spectrum—routine interactive and routine manual job tasks—there is no motherhood penalty in job tasks. Overall, our findings provide additional evidence for—possibly long-term—consequences of motherhood.
Previous comparative cross-sectional studies suggest gender differences in job-related training participation, particularly in countries with few family policies supporting maternal employment (e.g., Boll and Bublitz, 2018; Dämmrich et al., 2016; Dieckhoff and Steiber, 2011). However, without longitudinal data, these studies cannot examine whether and how childbirth alters mothers' participation in job-related training in different policy contexts. Furthermore, none of these studies can directly test the relevance of specific policies such as public childcare provision. This study examines whether higher levels of state-subsidized childcare provision are positively linked to mothers’ participation in job-related training. It combines individual-level data from the National Educational Panel Study for Germany (NEPS-SC6 adult cohort, N=5512, 2008-2020) with annual administrative records on county-level childcare coverage. In line with economic and sociological theories on labour supply and childcare take-up, the results from fixed-effects regression models provide evidence that higher childcare levels reduce the negative childbirth effect on mothers’ job-related training. Nevertheless, motherhood training penalties exist even in contexts with higher childcare coverage levels, especially in West Germany. The findings highlight the importance of supporting family policies to reduce motherhood training penalties and associated gender inequalities in the labour market.
It has long been recognized that life course experiences strongly influence quality of life; but long-term influences of partnership experiences on life satisfaction (from a holistic perspective on trajectories) have not yet been analysed within the literature. This study aims to full this gap by analysing effects of partnership trajectories among men and women within twenty years after the end of a partnership trajectory. Contrary to prior research (focusing on mid age only), we could show that women’s partnership experiences until the age of 40 influences their quality of life at higher ages and that these effects even increase with age. We also found stronger and more consistent effects of early partnership trajectories among men than prior research, which however strongly decrease with age. Over-all effects of partnership trajectories need time to unfold among women, while the decrease with time among men. Based on different strands of theory, we conclude, that gender roles as well as cumulation of secondary effects of partnership trajectories are responsible for this effect: Women invest more in partnership formation and during partnerships and suffer more from partnership dissolutions. They can therefore invest less into individual interests (e.g. adequate position in the job, leisure activities, friendships aside from family). These lower investments result in lower quality of life on the long run (secondary effect), depending on the course of the partnership. We thus contribute to the literature on mechanisms of evolvement of social inequality with respect to quality of life between men and women (and within the latter group) over the life time and its cumulation at higher ages.
back to top
At the onset of the COVID-19 pandemic, school shutdowns were one of the main measures to curb the spread of the virus in Germany. We ask if parents’ mental strain, learning support, and changes in working conditions during the shutdown affected students’ development of mathematical skills in primary school. Cohort comparisons show learning losses due to shutdowns, especially in mathematical skills and for younger students. Some studies also point to an increase in achievement gaps by students’ social origin. Research on parents report social differences in stress, as well as in learning support during the shutdown. Furthermore, changes in working conditions, such as working from home or reduction in working hours were common. Previous research has not systematically focused on the consequences of parents’ mental strain, learning support and changes in working conditions on students’ skill development during the shutdown. We approach this topic referring to the Family Stress Model (Conger et al. 1992) and Family Investment Model (Conger & Donellan 2007) and derive our hypotheses therefrom. We investigate these hypotheses drawing on data from Starting Cohort 1 of the German National Educational Panel Study (NEPS), which interviewed parents and tested children before and during the pandemic. Mathematical skills of n= 1,512 8-year-old children, almost all in second grade, were tested after the first shutdown. In previous waves, these students already took part in tests on mathematical skills, vocabulary and basic cognitive skills. For information on parents' mental strain, parental support and changes in working conditions, we utilize data from a questionnaire designed to capture the situation during the first shutdown. Again, there is information from previous waves on mental strain and learning related activities. Additionally, the data contain extensive socio-demographic information. We analyze the research question in a value-added framework using ordinary least squares regression and multiple imputed data. Results show evidence for a negative impact of parents’ mental strain during the pandemic on students’ mathematical skills after the shutdown. This effect is robust even when controlling for prior mental strain. If the parents worked from home due to the pandemic, students’ test results are also lower. This effect is at least partly mediated by the perceived mental strain. These results correspond to the Family Stress Model. Parental investment in child’s learning during the shutdown does not show any impact on test results. Parent’s reduction in working hours is associated with higher test results, concurring with both theoretical approaches.
Theoretical framework: Not only influential theoretical frameworks (Epstein, 2001; Hoover-Dempsey & Sandler, 2005), but also strong empirical evidence (Hill & Tyson, 2009; Dettmers et al., 2019) suggest that students report greater academic competences and greater wellness outcomes, when parents are actively involved in their education. A lack of parental involvement tends to increase children’s mental health problems (Wang et al., 2019). Due to the COVID-19 pandemic, adolescents and youth in many countries reported that the pandemic had a negative impact on their mental health (UNICEF, 2020). To date, it remains unclear whether parental involvement before the pandemic would have a long-term effect on individuals’ mental health during the pandemic. Furthermore, previous research indicated inconsistent results on how competences and mental health are associated (Suldo et al., 2011). To shed more light on this issue, we anticipated that student competences would mediate this relationship. Purposes: First, we examined the associations between parental involvement and school competences three before the pandemic, and mental health conditions during the first national lockdown. Second, we examined whether student competences would mediate these connections. Methods: Analyses of the present study were based on the data from the NEPS study (NEPS Network, 2021). A total of 959 participants (60% female; 14% immigrant background) had participated in two waves of measurement: t1: Grade 12 (Fall/Winter 2017); t2: May-June 2020. In t2, 66% of the participants pursued their vocational or higher education. Four items assessed the quantity of parental involvement (NEPS, 2021; alpha = .77). We used standardized competence tests: reading (Gehrer et al., 2013), mathematics (Neumann et al., 2013), and ICT (Senkbeil et al., 2013). Single-item measures for mental health conditions (NEPS, 2021): social belonging, social trust, self-related mental health, and loneliness. Immigrant background (German usage) and parental education were included in the models. Results: Contrary to our expectations, but not surprisingly, SEM results demonstrated that parental involvement was negatively associated with student competences in three domains even when controlling for immigrant background and parental education. Four mental health conditions during the lockdown were associated with at least one of the predictor constructs before the pandemic: The greater the parental involvement, the greater the social belonging the participants reported. The greater the mathematical competence, the lesser the mental health and the lesser the loneliness the participants reported. A mediation analysis confirmed a full mediation by mathematical competence. Results and implications of the present study were discussed.
Parents from higher social classes often have more (informal) knowledge about the functioning of educational institutions and use this knowledge, for example, when it comes to coaching their children for a scholarship interview. At the same time, more parental knowledge about the educational system is associated with higher educational success of children. While there is already much evidence in Germany that social background and educational success are closely linked, little is known about the micro-level mechanisms within the family that lead to this condition and what role parental knowledge about the educational system plays. Additionally, the parental guidance and support practices that affect students' (school) lives on a daily basis have so far been studied only in parts (e.g., homework control). This paper addresses this research gap. It looks at a wide range of parental practices and relates them to social background, parental occupation and knowledge about the educational system. Using Data of the NEPS Starting Cohort 3 (secondary school students), cluster and latent class analyses (LCA) were performed to identify types of parental practices (n = 3.273). Using LCA, three types can be identified that differ fundamentally in their practices. Type 1 uses a wide range of parental practices such as talking about subjects that were discussed in class, having conversations about arts and culture, or buying additional learning materials. These parents have a high level of knowledge about the educational system and a relatively high social status. The same applies to type 2, which is additionally strongly involved in the school, e.g., in the parents' council or helps at events. Type 3, on the other hand, offers the least support, has the lowest social status and the least knowledge about the educational system. In a next step, when parents are clustered by occupation, there is no longer a difference in educational knowledge among university graduates, but there is a difference in practices. For example, parents who are teachers talk to their children about books significantly more than engineers do, but engineers in turn are significantly more involved in the school than teachers. The results show not only the large differences by social background and occupation, but also that the NEPS provides a valuable opportunity to gain a fine-grained insight into the life realities of students.
The dual vocational education system in Germany is characterized by a high degree of content specificity and combines theoretical content in a vocational school with work in a training company, which enables the acquisition of subject-specific work experience (Bliem et al., 2014) and improves the employment opportunities of young adults on the labor market (Bol & van de Werfhorst, 2013). The process of career orientation is therefore essential for the development of individual educational biographies. In this regard, career choice maturity results from an interconnection of information, self-assessment and external factors (e.g. family, friends, school, career counseling) as well as working experience. Thereby, an active exchange between the individual and the environment proceeds (Höft & Rübner, 2017). The social-cognitive theory of career choice focuses on the great importance of the interaction of self-efficacy and outcome expectations, individual interest development, direct experiences, and job performance (Hirschi, 2008). School internships, which provide a realistic insight into occupational fields, represent one of the most important opportunities for career orientation (Brüggemann & Rahn, 2013; Ulrich et al., 2019). Despite the legal obligation to participate in an internship in the final classes of general education schools, it is not always realized. Studies show that young people who complete an internship are more likely to enter an apprenticeship. However, it is largely unexplored to what extent the school internship can specifically contribute to the transition into vocational training by the experiences gained, and how these experiences are related to the duration of the transition period. This is particularly relevant with reference to the effects of a fractionless transition on further educational and employment biographies. The aim of this study is to examine the experiences gained during the internship and their impact on the transition period. Therefore, data from the fourth starting cohort of the National Educational Panel Study (NEPS Network, 2021) is used, which represents a comprehensive documentation of the educational and employment trajectories of pupils first surveyed in the eighth grade (Blossfeld & Roßbach, 2019). Using event analytic models (Kaplan-Meier, Cox regression), the analysis focuses on pupils who aspired to an apprenticeship after completing the ninth grade. First, the results show that pupils with an internship have both a higher transition rate and a higher proportion of successful transitions. Moreover, the following factors emerge as important success factors: the function of the school internship to provide young people with information about the world of work, to support career choice, and the opportunity to get to know relevant people to contact in the search for an apprenticeship.
Young people’s choice of an apprenticeship is determined by various factors such as their interests, social expectations, requirements in the VET system and market conditions (Ahrens et al., 2021). Furthermore, the transition to different VET positions is socially selective (e.g. Beicht & Walden, 2015). This paper investigates whether social selectivity also exists with regard to adolescents’ work values, i.e. the importance individuals place on different job characteristics (Mortimer & Lorence, 1979; Johnson, 2001). Sociodemographic aspects explain differences in these work values: A lower family SES is associated with higher extrinsic work values, whereas a higher SES relates to intrinsic and autonomy work values (e.g. Johnson, 2002). Research results regarding the effect of education are not consistent. Research also shows that work values undergo changes due to transitioning to work (Lechner et al., 2017). This paper investigates 1. whether individuals from a lower social background have different work values from those from a higher social background and 2. which other (not only individual) factors besides social background determine individuals’ ability to realize the fulfilment of their work values in their current job. Analyses based on 1,800-1,900 tenth-graders from wave 3 of NEPS starting cohort 4 (2011/2012) reveal differences in work values according to parental background. T-tests show that pupils with an academic parental background attribute e.g. a higher value to being their own boss, and a lower value to a good working atmosphere as compared to pupils whose parents did not attend university. We also compared these ratings to the rating of the same values in wave 9. In the next step, we looked at those who finished an apprenticeship and compared these groups now with regard to the degree of fulfilment of their values in their current job. Here we used the individual rating of the values in wave 3/9 and looked at the degree of fulfilment of these aspects in wave 11 and 12. Only a tendency emerged with regard to good working hours (favouring those with an academic background), but linear regressions showed effects of other socio-demographic variables on the degree of fulfilment for some of the values assessed. In a further step, we expanded the regressions by characteristics of the occupation participants work in, which revealed e.g. effects of sector and company size. In a last step, we will compare the results for this subjective outcome (participants’ ratings of the job aspects) with objective outcomes like income.
School-to-work transition is a well-studied outcome in social research concerned with education and the labour market. Transitions are key objectives in the individual life course and are typically associated with a change in status. The transitions after compulsory school are particularly important because they can have lasting consequences for individuals’ occupational career paths. Factors influencing these transitions are ability, performance, social background, ethnicity migration, but also personality. Institutional factors, such as the social composition of school classes or schools, are known to be of relevance for post-school transitions. Furthermore, regional characteristics play a role. (Inter-)relations of effects at different levels (e.g., individual, institutional, regional) on individual outcomes are manifold and often contradicting. To untangle some of these (inter-)relations in effects, we use ANNs to analyse individual-level and context-level effects on transitions from school to training. To that end, we use NEPS Data enhanced with contextual information, i.e., regional and institutional data. Machine learning (ML) and artificial neural networks (ANNs) use algorithms to extract knowledge from large and heterogeneous data sets. Since the non-linear models are difficult to interpret and as they cannot be used straightforwardly for hypothesis testing, their implementation in the social science has been tentative. However, they are highly useful in generating hypothesis directly from the data and provide an interesting set of benefits: First, the derived results are agnostic to researchers’ believes, existing models and explanation approaches from different sciences. Second, it is possible to uncover new relations that have not been hypothesised about. Third, we are not bound to functional forms in the assumption of our models, like for example in regression analysis. By training multiple artificial neural networks (ANNs) on a well-suited dataset: the Starting Cohort 4 of the National Educational Panels Study (NEPS-SC4), we obtain a sufficiently accurate mathematical model for predicting individual choices of: (1) Entering dual VET after completing general education with a lower, intermediate or upper-level secondary diploma. (2) Entering school-based VET after completing general education with a lower, intermediate or upper-level secondary diploma. (3) Continuing in general school after lower- or intermediate-level secondary school or entering HE after obtaining Abitur. We derive hypotheses from the trained classification by analysing our neural network with decompositional and pedagogical methods, proving a ranking of the importance of input variables per classification and their interactions.
In the field of early childhood education and care (ECEC), it is often argued that early institutional childcare has beneficial effects on children’s abilities and thus promotes social equality. Accordingly, at young stage, children’s cognitive competences develop rapidly and immensely, increasing the returns of early educational investments (Beisly et al., 2020). Boudons (1974) primary effects are seen as the central mechanism in this context as the parental influence declines when they enrol their child (earlier) in formal childcare. However, most studies focus on rather short-term effects on different cognitive or non-cognitive abilities in the following years after Kindergarten, providing mixed evidence (Bassok et al., 2019; e.g., Fort et al., 2020; Klein & Becker, 2017; Santín & Sicilia, 2018; Skopek, 2017). One explanation for these inconsistent results is the challenge of eliminating unobserved heterogeneity, especially in cross-sectional studies. Due to such methodological constraints, just a few studies can observe causal effects on children’s temporary skills (Cornelissen et al., 2018; Felfe & Lalive, 2018; Kühnle & Oberfichtner, 2020). However, despite their methodological strengths, these studies are not able to identify whether the positive ECEC effects remain persistent during the lifespan, and hence, the importance of early education for the children’s future. Lastly, for the case of Germany, only small regions have been considered so far without including all federal states. Consequently, the following study is needed to achieve, as far as possible, causal long-term effects of ECEC all over Germany. This paper aims to contribute to the topical debate, by analyzing the effect of kindergarten attendance on long-term educational attainment, namely the type of the first school leaving certificate. The study links data from the National Educational Panel Study (NEPS), starting cohorts 3 and 4 (N=10.000), with administrative information on small-scale regional childcare coverage rates provided by the Federal Statistical Office (1994-2007). This time and region-varying information serves as an instrument in a special version of an Instrumental Variable Approach, the Marginal Treatments Effects Approach (MTE) (Carneiro et al., 2011; Cornelissen et al., 2016). By allowing for unobserved heterogeneity in treatment take-up, this method of modern causal analyses allows to extract a more causal effect, thus, providing valuable insight into the positive long-term relationship between institutional childcare and educational attainment in Germany.
Research question and relevance: This paper investigates the effect of school starting age (SSA) on noncognitive skills. School start is a major life event and school entry laws rank among the most common public policies worldwide. They typically define a particular cutoff in birth date such that all children born before this date are scheduled for enrollment in a particular school year, and children born thereafter start compulsory education one year later. The cutoff date largely determines a child’s SSA, and substantial research has studied its consequences for human capital development. Research gap: So far, the literature has predominantly focused on cognitive abilities and typically documented that a later school start yields a significant advantage in terms of early student achievement. Nevertheless, these disparities seem to fade out over time. Much less is known about the consequences of the SSA for the formation and persistence of noncognitive skills. Several studies have documented desirable effects of a later school start on students’ behavior and socioemotional development in the first years after enrolment. However, the evidence is far from being conclusive and does not offer insights into longer-run effects. Theoretical framework: From the theoretical perspective, persistent effects of SSA on noncognitive outcomes would be much in line with the dynamic model of skill formation (Cunha and Heckman, 2007, 2008), i.e., even if some differences initially arise solely due to the advantage in chronological age, the self-productivity and dynamic complementarity of skills create mechanisms that might perpetuate the effects into the teenage years and beyond. Data and methods: In this paper, we address the existing research gap by using plausibly exogenous, institutionally induced variation in SSA to provide a comprehensive analysis of its mid and long-run impacts on noncognitive skills, such as the Big Five Inventory of personality traits. Specifically, we apply a fuzzy regression discontinuity design to large-scale survey data from the National Educational Panel Study (NEPS-SC6) and the German Socio-Economic Panel (SOEP). Results: Our preliminary results suggest that SSA leaves significant imprints on the Big Five personality traits, some of which persist into late adulthood. Specifically, a higher SSA has an long-term impact on emotional stability.
The level and type of education that young people receive in the educational system is highly consequential for their prospective labor market chances. This applies both in terms of monetary returns as well as returns in other labor market dimensions such as employment chances or access to favorable positions in the occupational structure (e.g., Andersen & van de Werfhorst, 2010; Hanushek et al., 2017). However, education is not only connected to inequalities in the labor market. Based on Social Production Function theory, which stresses subjective well-being (SWB) as an “ultimate goal” (Ormel et al., 1999), life course research has also revealed inequalities with respect to SWB (Hadjar & Kotitschke, 2021; Patzina, 2022). In Germany, students with special educational needs (SEN) are at a particularly high risk of leaving school with low levels of education. They either attend segregated special schools or inclusive mainstream schools. Most often, students with SEN leave school without any certificate or only a lower-level school certificate (Holtmann et al., 2019). Students from special schools are especially disadvantaged in accessing vocational training (Blanck, 2020; Menze et al., 2021), resulting in an increased long-term risk of exclusion from the labor market (Menze et al., 2022). Moreover, students with SEN are also at risk with respect to SWB: Very early on, students with SEN report lower school-related SWB compared to students without SEN (Goldan et al., 2022). Qualitative research indicates that students with learning disabilities do not feel fully prepared for the transitions after school and experience anxieties during that time (Young-Southward et al., 2017). The presentation addresses the following questions: Does the type of school setting (segregated vs. inclusive) have long-term effects on the SWB of students with SEN after leaving school? To what extent is this effect mediated by differences in the transition to vocational training and the labor market that students with SEN experience after leaving different school settings? To answer these questions, we use rare longitudinal data on students with SEN from Starting Cohort 4 of the NEPS. We focus on students classified as having learning disabilities, i.e. the largest group among students with SEN in Germany, attending a segregated or inclusive school setting. Using matching techniques to render the two groups comparable in a wide range of observable characteristics (including cognitive skills), we aim to identify the effect of the school setting on two indicators for SWB: life satisfaction and feeling of social belonging.
Over the past few decades, educational scientists have enhanced our knowledge on students’ dismissive attitudes and emotions and its enormous consequences for students’ successful educational trajectories on one side and school dropout on the other side (Archambault et al., 2009; Pekrun et al., 2017). A recent conceptualisation understands school alienation as a multi-domain concept describing generalised negative attitudes towards at least one of three domains: teachers, classmates, and learning (Hascher & Hadjar, 2018). To date, this conceptualisation has been applied in empirical studies both in Switzerland and Luxembourg (Morinaj & Hascher, 2019; Scharf, 2018). However, much less is known about school alienation in the German school context. The aim of the present study was to examine school alienation in German secondary schools. Three research questions were addressed: (a) which alienation profiles can be found in German secondary schools, (b) are there any differences in the alienation profiles between seventh and tenth graders, and (c) how are alienation profiles associated with further individual characteristics, namely gender, immigrant background, social background as well as school related worries, school enjoyment and attention in class? Data were collected in autumn 2020through an in-class questionnaire survey in German secondary schools. A total of 152 seventh graders (M = 12 years, 62.9% male) and 98 tenth graders (M = 16 years, 54.3% male) participated in the NEPS developmental survey. School alienation was measured by the School Alienation Scale (Morinaj et al., 2017). The analytical procedure involves the following steps: After data preparation with SPSS28, Latent Profile Analysis (LPA) with Mplus 8.8. was carried out in order to examine school alienation profiles. Next, using multivariate analysis of variance linkages between selected profiles and students characteristics were assessed. Cross-sectional LPA revealed a four-profile solution for year 7 and a three-profile solution for year 10. The difference between the school years lies in the existence of a highly alienated profile from learning, teachers, and classmates. The results of the present study provide various insights into the characteristics of school alienation and its relations with students’ well-being in the German context. Implications of the present study will be discussed.
Grade retention has a long tradition in most school systems and is widely accepted by politicians, teachers, parents, and students. As of 2018, up to 148.000 students per year repeat a grade in Germany (Statistisches Bundesamt (Destatis) 2018). The effects of grade retention on repeaters’ academic achievement have been studied extensively, but results are only conditionally transferable to other contexts. The most frequently cited studies focused on the US context and showed mixed results, like neutral to positive effects of grade retention on achievement (for an overview, see Allen et al., 2009; Xia & Kirby, 2009). Overall, there is a lack of research for the German context with its hierarchically structured school system and an increasingly diverse student body. Against the backdrop of the commonly shared critical life event (cf. Filipp & Aymanns 2010) of the Corona pandemic and subsequent debates about the use of grade retakes to close potential gaps in individual learning development, empirical evidence on the effectiveness of grade retakes is highly relevant. In light of the theory of differential learning and developmental environments, (see Baumert, Stanat & Watermann 2006) of different educational tracks, the question arises if grade retention works differently in Gymnasiums, compared to the other school types. In this study, a subsample of the NEPS Starting Cohort 3 (N = 6,313, Blossfeld, Roßbach & von Maurice, 2011) was used. To address the effects of grade retention appropriately a propensity score matching was applied, which enables to compare students who repeated grade 7 and comparable students who were promoted to grade 8. With this matched subsample as well as the full sample various regression analyses were carried out. Doing so, further potentially remaining imbalances were accounted for in order to receive doubly robust analyses (Stuart 2010). Preliminary results show a significant effect of retention in a Gymnasium on math achievement (b = .65, p < .000) in the matched sample. Once math competence prior to retention is taken into account, the effect vanishes. No effects were found for reading competence as well as grades in German and mathematics. In view of these results, the practice of class repetition should be reflected critically.
Because of the pivotal role school grades play for transitions in the education system as well as from school to work and, therefore, in the allocation of life chances, it is imperative that they reflect student performance as objectively as possible. However, based on previous research, the extent to which school grades are assigned objectively and thus fairly according to students' performance is questionable. The aim of this paper is to examine the extent to which disadvantages in grading based on the ascriptive characteristics of gender, migration background, class and body weight can be identified in the German school system. We examine the grading of students in German, mathematics, physics, chemistry and biology. We are particularly interested in the extent to which there are potential intersections of various sources of disadvantage, e.g., overweight (BMI over 25) boys with an immigrant background versus girls without a migration background and with a weight considered “normal” (in terms of BMI: 18.5—24.9). To investigate these relationships, we use NEPS data (SC4; N = 14,005). These data allow us to account for standardized achievement tests, socioeconomic background variables, context factors, psychological factors such as the big five, and classroom behaviors. To investigate whether we can find grading bias based on ascriptive characteristics in the German school system, we take a stepwise approach. First, we estimate multilevel regression models (to account for the clustering of students in school classes) with standardized school grades as the dependent variable and one ascriptive characteristic as the independent variable (e.g. gender), controlling for domain-specific standardized test scores. In a second step, we include all other ascriptive characteristics and the control variables in the models. Furthermore, to illustrate potential intersections between different ascriptive characteristics, we calculated predictive margins (all other variables at the mean) for a combination of particularly advantageous and disadvantageous ascriptive characteristics. Our results indicate that we can detect grading bias by ascriptive characteristics in all subjects, which can only be partially explained by psychological factors and classroom behaviours. It is striking that the most significant effects are found for gender. Interestingly, we find overweight bias for all subjects apart from mathematics. Furthermore, a bias against Turkish students is only observable for German. Looking at the intersections between ascriptive characteristics, we see that chubby Can indeed gets lower grades than skinny Sophie - however this difference is only substantial for German.
My paper sheds light on the sources of matching inefficiencies between firms and apprentices in the German vocational education and training system with a special interest in gender differences, namely the gender gap in matching. The German system has been characterized by problems of fit for some time now (BIBB, 2021). At the same time, female students, although better qualified at school, have been less successful in entering vocational training than male apprentices (Imdorf, 2013). Nevertheless, research on the match quality between school leavers and their apprentice firms is scarce. My approach is novel in the way that I am applying the rank-based method of Campbell, Macmillan, Murphy, and Wyness (2021) to the German apprenticeship market. I rank school leavers and firms based on the job-competition model of Thurow (1975) and Reskin and Roos (1990), using the rich NEPS-SC4-ADIAB data set (Bachbauer, Wolf, & Fuß, 2021). While the respective labor queue consists of school leavers ranked by their final grades, their apprentice firms are ranked by four different quality measures: the firm’s size representing long-term non-monetary career prospects and benefits, the firm’s average wage level as a measure of monetary returns, the apprenticeship salary as a measure of short-term monetary benefits, and a multi-dimensional measure which combines these aspects. The quality of match is measured as the distance between the relative rank of the firm and the relative rank of the school leaver. By doing so, I provide answers to how efficient the matching on the apprenticeship market is and whether female apprentices are systematically mismatching. I find that there exist sizeable ineffiencies in the matching process and women are more affected by them. Female apprentices work in lower quality firms, regardless of the firm quality measure looked at and conditional on background characteristics. Women utilize their educational achievement less than men in terms of potential future earnings, career opportunities and the received apprentice wage. The multidimensional measure reveals the largest gender match gap. The gender gap is particularly prevalent for high attaining female school leavers. Lastly, I explore potential mechanisms driving the gender gap in matching. My findings suggest that a stronger preference for high occupational prestige, even if this reduces career prospects according to economic indicators, is a possible explanation, as could discriminatory hiring behavior by employers. I could not conclusively verify either of these with the data available to me.
To become a baker, a technician, or a bank clerk requires learning different tasks and acquiring different competencies. In the German vocational education and training (VET) system, young people get training in more than 300 training occupations. (Training) occupations differ in multiple ways. One decisive characteristic is how cognitively demanding it is to learn and to perform occupations. Getting a training place for an occupation, which is cognitively more demanding is associated with better opportunities regarding further competence development and future returns. While numerous studies investigated transition chances to VET in general, we contribute to the still developing research focusing on occupational heterogeneity within VET and its implications for social inequalities. More specifically, we ask whether young people with more advantaged social backgrounds and with higher cognitive abilities have higher chances of getting training places in more demanding occupations compared to those with less advantaged backgrounds and with lower cognitive abilities. Theoretical accounts on the intergenerational transmission of educational and occupational advantages rest on the ideas of cultural reproduction (Bourdieu 1973), and/or class‐specific decision making (cf. Boudon 1974; Breen/Goldthorpe 1997). Employers in the dual apprenticeship system strongly rely on school-leaving certificates and school grades as selection criteria. These skill signals are also decisive for higher-level school-based programs. Once individuals leave the general schooling system, family resources might still facilitate accessing more demanding training occupations even if their school achievements are – perhaps unexpectedly – poor. We analyze NEPS-SC4 data and focus on school leavers who applied for a training place in the VET system and started their apprenticeship or school-based program latest in September, the year after graduation. To the training occupations, we match information on the cognitive requirement level; an indicator developed by the Occupational Psychology Service of the Federal Employment Agency to advise young people on the various training occupations (Friedrich et al. 2022). Preliminary linear probability regressions on 2,572 school leavers indicate that young people with higher school-leaving certificates have higher chances of entering a cognitively more demanding occupation – with (non-)cognitive abilities, school grades, background characteristics, and regional conditions as controls. Furthermore, comparisons within educational groups reveal that higher parental socio-economic status and higher cognitive abilities are associated with higher probabilities to enter more demanding training occupations for school leavers with intermediate certificates. This is not the case for those with lower or upper secondary school-leaving certificates. We cautiously interpret these findings as barriers to higher-levels of VET learning experience for lower SES-children and those with lower levels of schooling, and suspect ceiling effects for those with upper secondary certificates.
Few studies with NEPS data have so far resorted to perceived discrimination as a construct and mostly it was considered as an independent variable (e.g., Busse, 2020; Hunkler & Tjaden, 2018; Lindemann, 2020). For those affected, the experienced situation is often more important than statistical discrimination (Ette et al., 2021, p. 28). Thus, perceived discrimination should be subject for further research. This study asks the question how the perception of personal ethnic discrimination in adolescents differs depending on their origin group. Several studies have already found differences in perceived ethnic discrimination by origin groups (Diehl et al., 2021; Flores, 2015; Salentin, 2007; Steinmann & Strietholt, 2019). In contrast to them, the focus of this analysis is on the transition to vocational education and training. Productivity-theoretical assumptions are taken into account to explain differences between origin groups. Furthermore, gender differences in the perception of ethnic discrimination are examined in consideration of intersectional models: The Subordinate Male Target Hypothesis, the Additive Approach, the Intersectionality-Inspired Approach and the Model of Intersectional Invisibility (Veenstra, 2012; Purdie-Vaughns & Eibach, 2008). Some studies show that women perceive ethnic discrimination more strongly (e.g., Skrobanek, 2007, p. 32), whereas other studies demonstrate that women report less ethnic discrimination than men (e.g., Salentin, 2008, p. 521; Schaafsma, 2011, p. 789; Te Lindert et al., 2008; Venema & Grimm, 2002, p. 72) and anticipate ethnic discrimination in the labor market with a lower probability than men (McWirther, 1997, pp. 133-135). Due to the contradictory results, the question arises how perceived ethnic discrimination differs depending on gender. In which group of origin is the difference between young men and women most pronounced? Wave 5 of Starting Cohort 4 of the NEPS is used (N = 1.421). A sufficiently large sample for the analyses is available for migrants from Turkey (N = 307), the former Soviet Union (N = 296), Poland (N = 131) and the former Yugoslavia (N = 141). Logistic regression models and marginal effects (AMEs, MERs and predictive margins) are calculated to test the hypotheses. The results show that young people from Turkey perceive more personal discrimination when looking for an apprenticeship position than Polish, ex-Yugoslavian, and ex-Soviet adolescents. The probability that Poles will experience ethnic discrimination compared to Turks is lowest. No correlation between personal discrimination and gender is confirmed significantly, which is probably due to the small sample size. Nevertheless, a higher gender difference in the Turkish group than in the other groups of origin is noticeable. In the Turkish group, the probability is higher among women, but in the ex-Yugoslavian group it is higher among men.
This paper studies the long-term relationship between parental and child education in Germany, where children are tracked into academic and non-academic track schools at the age of 10. On average, children are more likely to attend an academic track school if their parents attended one. Estimating marginal treatment effect curves, we find that there is no effect for disadvantaged individuals, suggesting that educational policies attempting to improve the educational prospect of disadvantaged individuals may fail to reduce inequalities in the long run. Low labor market returns despite better education is the main explanation for the null effect for these individuals.
Since the end of the 1990s, following Boudon (1974), an increasing number of empirical tests of the main arguments of RCT on class-related educational decision at the branching points in the educational system have been published (Stocké 2019). However, most of them provide an indirect test of RCT on educational decision-making (Brüderl 2004). They seek to indicate the impact of the class-specific expectation of the costs and benefits of different educational alternatives, as well as the expected probability of successfully attaining an educational degree in regression models. However, the specification of such an additive model are notoriously incomplete, since it does not meet the logic of the rational choice of options; namely the maximization of the subjectively expected utility as the rule of the selection of an educational option. This theoretically inappropriate procedure is suspected of playing into the hands of “variable sociology” (Esser 1996). Therefore, it is insufficient to contribute empirically to the explanation of IEO (Becker 2022). Overall, it is the aim of this empirical contribution to test established explanations of educational differentials suggested by Esser (1999), Breen and Goldthorpe (1997) and Erikson and Jonsson (1996) directly. To explain educational inequalities by class, a broad version of a rational choice theory is tested using a direct test strategy. First, it makes use of theoretically adequate measured constitutional elements such as the benefits, costs, probability of success and the motive of status maintenance. Second, the logic of selection in terms of utility maximization will be specified in regression models in a theoretically adequate way – i.e. the process and mechanism of procedural rationality are indicated directly in the multivariate analysis. It is assumed that social mechanisms of individual educational decisions can contribute to the full explanation of the social inequality of educational opportunities. Third, longitudinal data and adequate statistical procedures are used for the empirical analysis. Panel data on the educational progress of young people in Switzerland are used as an example for the direct test. Fourth, according to Boudon (1974), educational differentials are explained by the systematic interplay of primary and secondary effects of social stratification. Using an innovative modelling of the core mechanisms – such as selecting one of the educational options – by a direct measure of the cost–benefit calculation and an evaluation of the subjective expected utilities (Becker and Glauser 2018), or by “hedging” (Tutić 2017), the impact of social origin on the educational decision has been described by longitudinal data. In sum, it is possible – in our view at least – directly to test predictions of human behaviour relating to continued education in an adequate way.
Research question: The current study aimed to investigate the influence of student and school level factors on school tracking in secondary education. We were especially interested in the association between student characteristics and school composition in Grade 3 and school track in Grade 9. Data source: Data were collected as part of the Luxembourg school monitoring programme “Épreuves Standardisées” (ÉpStan; Fischbach et al., 2014). The study cohort include all students enrolled in the Luxembourg public education system in Grade 3 in November 2013 combined with data from the same students in Grade 9 in November 2017-2019 for students following advanced or regular educational pathways, completed with data from November 2020 and 2021 for students that repeated once or twice (N≈3600). Theoretical approach: The study draws upon theoretical frameworks and empirical findings (e.g., Boudon, 1974; Bourdieu, 1984), that have demonstrated students´ socio-demographic characteristics are associated with (dis)advantages for specific groups of students in education systems as well as more recent work focusing on school composition (e.g., Baumert et al., 2006), especially as tracked school systems are known to be prone to social segregation (e.g., Hadjar & Gross, 2016). To date, most research on school segregation in tracked education systems such as Luxembourg has focused on individual student´s characteristics. However, with increasing heterogeneity of student cohorts and known differences in educational opportunities related to the social and ethnic composition of the school’s student body (e.g., Thrupp et al., 2002), the current research extents the existing literature by considering both individual (including prior academic achievement and socio-demographic characteristics) and school level factors (mean academic level and percentage of students from lower socio-economic and migration background) in predicting school track placement. Main findings: Results of a multilevel random effect logistic regression analysis in which we estimated marginal effects on the probability to be placed in the highest, middle or lowest track in Luxembourg show that even after controlling for student´s academic achievement, track placement is affected by the gender and socio-economic background of the student, whereby boys and students from low SES families have less chance to be placed in the highest track. The association with socio-economic background is not only visible on the student level but also on school level, whereby students attending primary schools with a higher percentage of low SES families have less chance to be orientated to the higher track compared to the middle track, regardless of the student´ individual academic performance.
Since various international and national educational large-scale studies (e.g., IGLU, PISA, VERA) repeatedly identified a substantial proportion of students with very low levels of reading literacy, these studies´ authors emphasized the urgent need for reading promotion programs throught extracurricular tutoring. Accordingly, numerous tutoring programs and related evaluative research have been launched in recent years. In addition to attempting to improve students´ reading literacy, these projects also provided important insights into the tutoring conditions that promote learning success in general (cross-curricular features, e.g., optimal scope, organization and tutor types, quality criteria) and the acquisition of reading literacy in particular (reading-related features, e.g., reading exercises). Nevertheless, large-scale studies to date have rarely examined the use, cross-curricular and reading-related features, and reading literacy-enhancing effects of such tutoring programs. Against this background, the present study investigates a) the extent to which students take advantage of different types of tutoring (e.g., in German), b) how these programs are characterized in terms of cross-curricular and reading-related features, and c) how they differentially affect reading literacy of students of various subgroups that differ in terms of gender, migration status, and support needs. The analyses are based on data of Starting Cohort 3 of the National Educational Panel Study (N = 5.113). Differential effects of tutoring on reading literacy in Grade 7 were examined using a multigroup structural equation model, while controlling for students´ prior reading literacy in Grade 5 and additional covariates. First, the results show that only a relatively small percentage of students used tutoring programs (mainly in German, less frequently in reading), which are also described in terms of their cross-curricular and reading-related features. Second, differential effect analyses related to participation in tutoring in German (compared to non-participation) revealed no overall effects on students´ reading literacy and only minor variations for specific subgroups of students. Generally, the study highlighted that the most commonly used tutoring in German does not seem to be suitable for meeting support needs and promoting students´ reading literacy, especially not in the same way for all students. Instead, we expect interventions to be more effective if they target reading literacy and are implemented adaptively to target students based on their baseline characteristics. Therefore, it seems promising to implement more specific tutoring programs in the future and study their effects on helping students with low reading literacy to achieve age-appropriate reading literacy standards.
Climate change and global warming have already significantly shaped the 21st century with global temperatures even further on the rise with each of the last four decades being successively warmer than the one preceding. These changes have resulted and will continue to result in profound alterations for humankind and natural systems in manifold ways: wildfires, floods, pollution, storms to only name a few. Northern regions were partly expected to benefit in the medium term with damages being more likely to manifest at higher levels of warming and at low latitudes. This perception, however, has shifted to a more nuanced discussion in recent years, bringing a new stance in economics to life that focusses not only on agricultural outputs and extreme weather events but on societal implications due to a changing climate such as mortality & morbidity, productivity and human capital. Within these emerging research interests, human capital – and with-it education – are recognized to hold significant importance for cross-generation development, economic competitiveness and growth, their evaluability, however, poses a challenge to researchers. The envisaged paper seeks to contribute to this quest and aims to answer the question whether rising temperatures in Germany are already impeding students´ cognitive performance. The empirical analysis will employ a panel method executing the hypothesis that students score lower following a hotter time period relative to their own score immediately following a cooler time period. The observation of individuals across time and their responses to changing climatic conditions allows the assumption of fixed factors as fundamental characteristics that are unlikely to change within the period under review. To inform the analysis, data will be obtained from the National Educational Panel Study (NEPS). The selected Starting Cohort 3 follows students from class 5, the first year of secondary schooling, through Lower Secondary School up to 8 years starting in the autumn/winter term 2010. The second major source of data in this research project is data on local temperature. It is envisaged to use data from the COMSO-REA6 reanalysis system warranted by the Germany's National Meteorological Service, the main repository for weather data in Germany.
The transition to vocational education and training (VET) is an important developmental goal in adolescents’ life course (Buchmann & Kriesi, 2011). However, in recent years the match between applicants and training positions has become increasingly difficult, which hightens the risk of being unemployed (Baas & Philipps, 2017; Baethge et al., 2007). Therefore, it is important to identify factors that impact adolescents’ transition to VET. We argue that individual regulation strategies increase transition probabilities but that this influence is limited by contextual constraints. Th work is based on Brandtstädter’s dual-process framework of goal pursuit and goal adjustment. The model postulates that in goal striving two complementary modes of regulation exist. In the assimilative mode, individuals tenaciously pursue their goal; in the accomodative mode, they adjust their goal to situative constraints. While the former is adaptive when goals are attainable, the latter is adequate when goal achievement becomes unrealistic. It is well established that individuals have stable preferences regarding their activation. People high in tenacious goal pursuit (TGP) tend to stay in the assimilative mode even when goal attainment becomes difficult. Individuals high in flexible goal adjustment (FGA) switch to assimilative processes early in goal striving situations (Brandtstädter & Renner, 1990; Brandtstädter & Rothermund, 2002). Both TGP and FGA are positively associated with (subjective) success (e.g. Haratsis et al., 2016). However, no study has investigated their influence on transition to VET. This is one aim of the study. Secondly, we focus on their dependence on contextual constraints: We hypothesize that TGP is more successful in favorable situations, whereas FGA is more effective in unfavorable contexts. Opportunity structures that are especially relevant for the transition to VET are local labor markets (Weßling et al., 2015). Therefore, we will analyze how local labor market structures moderate the relationship between individual strategies and transition success. Transition success is operationalized threefold, namely as a) attainment of any VET position, b) attainment of the preferred VET position and c) satisfaction with the obtained position. The analyses concentrate on students from intermediate-level secondary schools (Realschüler*innen), who will be targeted until one year after obtaining a school diploma. To that end, we use data from the National Educational Panel Study (NEPS) – Starting Cohort 4 (9th graders; Blossfeld & Roßbach, 2019) and combine it with regional information on the supply and demand of training places (Bundesagentur für Arbeit, 2011).
Motivation: Youths’ spatial mobility is, on the one hand, an important means of solving matching problems in training and labor markets, which vary widely across regions (Matthes & Ulrich, 2018). On the other hand, spatial mobility is associated with higher monetary returns at the individual level and therefore a source of social inequality (Lemistre & Moreau, 2009). While aggregate data analyses show substantial spatial mobility among students in vocational education and training (VET) (Bogai et al., 2008; Jost et al., 2019), research on its determinants is scarce. Therefore, this study analyzes the role of individual occupational orientations, social ties, and regional opportunity structures for the spatial mobility of youth in the transition from school to VET. Theory: Prevailing theories explaining spatial mobility originate from the neoclassical rational choice framework, which views job- or training-related mobility as economic cost-benefit decisions (Cadwallader, 1989). Extending this perspective, we integrate theoretical approaches that understand individual mobility potential – also called motility (Kaufmann et al., 2004) – to result from agentic motivation and structural constraints and emphasize the importance of individual, social, and regional binding and push factors (Haldimann et al., 2021; Windzio, 2008). Data: We used longitudinal data from the German NEPS (Starting Cohort 4, N=3,143) (Blossfeld & Roßbach, 2019; NEPS Network, 2021) to analyze the predictors of spatial mobility in students entering VET within two years of leaving the general school system. To map the regional opportunity structure, we merged regional data at the level of the school location using NUTS-3 regional delineation (BA, 2022; BBSR, 2022). Method: We estimated multinomial logistic regression models to examine the role of youth occupational orientations (status aspirations; search duration), social ties (satisfaction with friends, and the family; participation in organizations), and regional structure (commuter balance; share of the aspired occupational segment) for spatial mobility. Drawing on the concept of functional labor market regions (Kosfeld & Werner, 2012), we distinguished between spatial mobility within and across regional labor markets. Results: Our results show a strong prevalence of spatial mobility among youth transitioning from school to VET: 16% are mobile within and 23% across regional labor markets. The results show that individual orientations, social ties, and regional opportunity structures are decisive for youths’ spatial mobility. Most notably, the regional opportunity structure is a key determinant of whether youth start VET in their home region or become mobile, even across labor market regions.
In recent decades, Germany has turned away from regular employment for the non-established, with employment entrance becoming the most insecure phase of a career. The public-sector employment regime, with its specific institutional setting, has been seen as a "model employer” for disadvantaged groups, but reforms and restructuring are increasingly calling this special status over the private sector into question. In this article, we examine whether the public sector provides more stable employment for entrants than the private sector. We use the matching approach by Sørensen and Kalleberg (1981) and the occupational closure framework by Bol and Weeden (2015) which explain under which institutional conditions jobs are more (un-) stable, adapt it on the situation of labour market entrants and debate why the occurrence of these institutional conditions is more likely in the public sector. We identify three potentially stabilizing institutional factors and two potentially destabilizing factors. In a khb-decomposition-analysis, we test the potentially stabilizing (working in a service relationship, high occupational closure, participation in further education) and the destabilizing institutional factors (fixed-term and part-time employment) to explain the stability of first jobs in the public sector. We use the labour market entry cohorts 1995 to 2012 of the retrospective life history data of the Starting Cohort 6 of the National Education Panel Study (NEPS) and combine it with data from SC4 to account for potential differences in professional and family preferences. The results show that the public sector is offering more stable first jobs, and stabilizing factors, like further education and a higher level of licensing as part of occupational closure are influential. However, destabilizing factors, like the intensive use of fixed-term contracts, have a significant negative impact. This suggests that the stabilising institutional structure of the public sector, while functioning, is under pressure and thus offers potential for polarisation.
Parental educational attainment and their economic resources, often jointly referred to as social origin, can be considered as a practically indispensable concept in educational research. Economic argumentation dominate the scientific debate about different students' performance with respect to parental socioeconomic status. For this reason, the Index of Socio-Economic Status (ISEI) is an important and frequently used empirical measurement tool in educational research. In addition to economic resources, cultural capital, often measured by parental educational qualification as representation of its institutionalized form, is another relevant factor in resource-orientated explanations of primary effects on pupils´ performance. However, when trying to account for socioeconomic and educational influences jointly in linear regression models, relevant theoretical and methodological issues arise. During the analysis of social-origin-specific achievement differences between pupils in 4th grade using data from the National Educational Panel Study (NEPS-SC2), this contribution will discuss nonlinearity, collinearity and inconsistent variances as well as some distributional assumptions of OLS-Regression, which occur when regress ISEI and parental educational background on pupils mathematical competencies. Besides the fact that misspecification violates the requirements of linear ordinary least squares regression and lead to wrong interpretations, it also masks underlying mechanisms and structures. In contrast, the explicit consideration of theoretical premises and assumptions about the functional form of a correlation, as well as their methodological implication in the analysis can help to avoid statistical model violations and to develop statistically adequate, more informative models. This empirical and methodological discussion of theory-sensitive modeling and interpretation strategies will illustrate how the use of non-linear, piecewise linear and non-parametric regression techniques can be used to implement empirical models which explicit respect the functional form of theoretical arguments. In addition to avoiding the demonstrated statistical issues, this approach improves the fit between theory, methods and data and consequently leads to more precise and meaningful empirical models as well as to deeper insights into the mechanisms behind the container of social origin. This is supposed to revitalize the discourse for a more theory-sensitive operationalization of Social Origin in educational research.
As theorized by Boudon (1974), there is broad evidence that students with a high socioeconomic status (SES) perform better and obtain higher school-leaving certificates than their low-SES peers (e.g. Stephens et al., 2015). At the same time, following Wigfield and Eccles’ (2000) expectancy-value theory, malleable motivational factors promote educational attainment (e.g. Wang & Finch, 2018). However, students’ motivation depends on their family background and also mediates the relationship between family background and educational attainment (e.g. Ditton et al., 2019). Using data from the German National Educational Panel Study (NEPS) starting cohort Grade 9 (SC4) (NEPS Network, 2021) and estimating three structural equation models, this study sheds light on the associations between family background, motivation and educational attainment, addressing three questions: Do the known relations between family background and obtaining Abitur hold for students in Gymnasium? How are the investigated motivational factors related to obtaining an Abitur? Is the expected association between family background and obtaining Abitur mediated by motivational factors? Extending previous research – which mainly focused domain specific contexts and performance – we took a more general perspective to uncover mechanisms promoting obtaining an Abitur. Data were collected between November 2010 (T1) and August 2013 (T5). We focused on students enrolled in Gymnasium (N = 5,409 students; M = 14.6 years; SD = 0.65; 52.4% female; 77.4% obtained an Abitur). Family background was operationalized by parental educational level, parental occupational status, and immigrant background (T2, data provided by parents). Concerning motivational factors, students answered questions on their expectation of success (T3), value of education (T4), effort (T5), and performance-oriented motivation (T5). Additionally, we controlled for math and reading performance, basic cognitive skills, and gender (all T1). Also in a socially selective group (students who attended Gymnasium), the finding that a higher family background promotes educational attainment was reproduced: In Model 1 especially parental education stood out as main predicator among the investigated family background variables. The results in Model 2 indicated that particularly expectation of success promoted obtaining an Abitur compared to the other motivational factors included. Since parental education was found to have a significant relation with educational attainment, mediation analyses in Model 3 focused on the potential mediation of this relation. The direct effect from parental education to obtaining Abitur was still significant but smaller – this partial mediation was mainly due to the expectation of success. Results and practical implications for reducing educational inequalities are discussed.
This paper analyses 1) how children with a one-sided migration background (mixed origins by immigrant-native parents) perform in their natural science competencies and 2) which factors contribute to above-average performance in assessing scientific competencies among children from interethnic families. The proportion of interethnic partnerships in Germany is steadily increasing, but there are hardly any findings on the children from these partnerships. This knowledge deficit applies in particular to the observation of educational success with a focus on the natural sciences, which, in addition to linguistic competencies, are central to the occupational placement and social participation in the future life course. Based on theoretical considerations by Boudon (1974) and Bourdieu (1983) we analyse children's educational performance with a one-sided migration background based on NEPS data compared to a two-sided and without a migration background (N = 4.344). Moreover, we consider whether the father or the mother causes the one-sided migration background. Since the proportion of students with a migration background at the school level or in the school class could be relevant as a context effect, multi-level models are applied. The models control for language competence, cultural distance, and heterogeneity in the characteristics of the parental home (parents' education, marital status, employment status, etc.). Our findings show that the educational outcome for children from interethnic families can be located between the educational success of children without and with a two-sided migration background. However, this effect becomes smaller when controlling for different characteristics of the parental home and the child's language competence. Migrant fathers play a more important but negative role in their children's educational success. Moreover, our results support gender differences in natural science competence in favour of boys. Our findings contribute to the debate on economic integration and educational inequality of immigrants and their descendants in Germany.
The Wisconsin Social Psychological Model of Status Attainment is a prominent theory aiming to explain the development of aspirations as dependent on social background. A central assumption of the model suggests that students’ perception of their parents' aspirations shapes students’ aspirations, which then determines students’ educational and occupational attainment (Sewell, Haller, & Portes, 1969). The authors state clearly that it is students’ perception that is crucial in this process, not parents’ factual aspirations (Sewell & Hauser, 1980). In contrast to this, literature investigating the Wisconsin Model often operationalizes parents’ aspirations as measured by parents instead of taking students’ perceptions. One main argument for doing so is the possible problem of reverse causality. Instead of students’ perceptions of parental aspirations influencing students’ aspirations, it might be the other way around and students’ own aspirations are projected on their perceptions of what their parents wish for (Hoenig, 2019). In our paper, we ask if the substitution of students’ perception of parental aspirations with parents’ aspirations, directly asked from them, is reasonable and valid for the investigation of the Wisconsin Models’ mechanism. Two research questions are addressed. First, we ask if the mentioned potential problem of reverse causality can be found empirically. Second, we investigate whether the idealistic aspirations for an Abitur (binary coded) measured by parents can explain as much as the student's perception of these aspirations can do. NEPS data from starting cohort 3, wave 2, 4 and 5, are used for analysis. Concerning research question 1, a cross-lagged panel model (CLPM) with three measurement points shows statistically significant crossed effects. Thereby, the problem of reverse causality is empirically detected. Comparing effect sizes, it turns out that the effect of own aspirations on perceived parental aspirations is stronger, sometimes almost twice the effect size as in the other direction. Regarding research question 2, coefficients of linear fixed effect regressions and within-person R2 show students’ perceptions of parents’ aspirations to be more relevant for explaining their aspirations than factual parental aspirations. To conclude, one should avoid taking students’ perceptions of parents’ aspirations when trying to make causal inferences, since we empirically showed the assumed problem of reverse causality. Students do indeed project their own aspirations on the perception of their parents’ aspirations. However, using parental aspirations as measured in parents’ interviews, researchers should consider their analysis to underestimate the influence parents’ aspirations have through the perception of their children.
Aspirations have been shown to be highly predictive for young people’s educational and occupational attainment and they substantially regulate students’ behavior at the transition from school to work (e.g. Cochran et al., 2011; Rojewski, 2005). In line with prominent theories (Gottfredson, 1981; Super, 1980), a body of research emphasized age-related changes and individual differences in development of occupational aspirations, specifically during adolescence (e.g. Gao & Eccles, 2020; Lee & Rojewski, 2009; Miyamoto & Wicht, 2020). The results revealed also differences in the level of aspirations depending on social origin with an indicated disadvantage for adolescents from less privileged social origin. However, there is a lack of evidence regarding origin-specific development and whether social disparities increase over time. To explain individual differences in the development of occupational aspirations, research mainly focused on individual characteristics, but according to theory also the institutional context seems to have an impact on the development of students’ aspirations, especially in stratified school systems such as the German. Although some studies reveal mean level differences in occupational aspirations by school track (e.g. Lühe, 2018; Wicht et al., 2017), less is known about the influence of attended school type in Germany, particularly regarding the development of occupational aspirations depending on social origin. Therefore, the present study examines social disparities in the development of occupational aspirations during adolescence and aims to answer the question whether and to what extend attending a specific school type fosters (or hampers) developmental trajectories of occupational aspirations, particularly regarding origin-specific differences. To address these research questions, we use longitudinal data from Starting Cohort 3 of the German National Educational Panel Study (NEPS; Blossfeld & Roßbach, 2019), consisting of a sample of secondary school students (N = 3,347) from grades 8 to 10. Occupational aspirations are operationalized via continuous ISEI scale. Methodologically, we focus on multi-group latent growth curve modeling. The results indicate social disparities in the initial level and the development of occupational aspirations during secondary education, that are partially explained by attended school type: While the academic track seems to provide a selective but daunting environment for students from less privileged social origin, it makes a difference whether students attend a degree-linked or degree-flexible school type, with the latter seems to be a beneficial and disparity-reducing environment. Our findings extend the knowledge on social disparities in the development of occupational aspirations during adolescence and the contribution of the German secondary school system.
Across countries, young women are less likely to aspire STEM occupations than young men (Hägglund & Leuze, 2021). Since subsequent STEM occupations come along with higher wages (Kim et al., 2015), understanding what keeps women from entering STEM continues to be an important research endeavor. Previous research indicated various factors determining this gender gap: e.g. course-taking (Mann & DiPrete, 2013), life and career plans (Morgan et al., 2013) or significant others (Gabay-Egozi et al., 2015; Raabe et al., 2019). Only few studies investigated relations between STEM occupations and personality differences. Existing research, mostly focusing on self-efficacy and interests, reports first – partially contradictory – support for such an association (Glynn et al., 2015; Kang et al., 2019; Kurban & Cabrera, 2020). Specific personality traits have hardly been considered. Previous research indicates a negative association between Neuroticism, Agreeableness and Extraversion and STEM preferences, while results on Conscientiousness are partly contradictory and context specific (Coenen et al., 2021; Korpershoek et al., 2010; Korpershoek et al., 2012). There is also evidence that effects of personality traits on occupational aspirations differ between girls and boys (Korpershoek et al., 2012). Since no study on the German context – in which professional decisions have to be done late, but are difficult to revise – has yet investigated this relationship we do so. Based on Gottfredson's Theory of Circumscription and Compromise (Gottfredson, 1981), we question whether occupational stereotypes and specific personality traits are matched in adolescence. We test the derived hypotheses empirically by means of Starting Cohort 4 of German NEPS data (Blossfeld & Roßbach, 2019), using a cross-sectional sample of 4,680 students at German upper secondary school. Our central variable is the realistic occupational aspiration in grade 9, categorized in two STEM subfields (Mathematics/Natural Science; Engineering/Technology), since gender differences are more pronounced in the latter. These are compared to pupils with Non-STEM aspirations and those who cannot yet designate a clear occupational preference. To measure personality, we use the NEO-PI-R Big-5-scale by Rammstedt and John (2007). Results based on multinomial logistic regression show moderate effects of personality on STEM occupational aspirations. In the whole sample increases in Extraversion, Conscientiousness, and Openness are conducive to the formation of vocational aspirations. But, for women in particular, higher scores lead to a lower likelihood of indicating a natural science aspiration, in line with our hypotheses. While those students aspiring to tech show generally more pronounced personality traits, they hardly meet our theoretical assumptions.
In the course of the discussion about the lack of qualified personnel, problems in the dual training system are gaining attention. In addition to supply and staffing problems, the proportion of unsuccessfully completed apprenticeships in Germany is rising almost steadily (26.9% in 2019) (Bundesinstitut für Berufsbildung, 2021). First evidence already indicates that high-quality information can reduce uncertainty in career choices (Beinke, 2011). The present study addresses this topic and investigates the following question: What is the impact of apprentices' level of information on dropout of vocational training? In the labor market, it can be assumed that information asymmetries exist on part of both the employer and the employee. More complete information about requirements and skills can lead to a better fit (Hinz & Abraham, 2018). This consideration can be transferred to the vocational market and thus the following main hypothesis can be stated: The more informed apprentices are, the lower the probability of dropping out of vocational training. Since information is derived from a wide variety of sources, the present study differentiates between strong and weak ties (Granovetter, 1973). Furthermore, additional sources of information are added that play a role in the context of vocational orientation and school-based vocational preparation. Corresponding hypotheses are tested using data of the start cohort class 9 (SC4) from the National Educational Panel Study (NEPS). The first findings show that a good state of knowledge about one's own apprenticeship makes it less likely that one will drop out prematurely. However, knowledge about alternative vocational trainings has the opposite effect. Information provided by parents or relatives seems to reduce the probability of dropping out. Still, this finding is not significant, as is the influence of information from friends, which seems to make dropout more likely. Information from weak ties, especially from teachers and acquaintances from an internship or previous jobs, significantly reduces the probability of dropping out. If the vocational preparation provided by the school is perceived of high quality, the probability of dropping out is reduced significantly. The study demonstrates that selected aspects of apprentices' level of information have a significant influence on the probability of premature dropout from vocational training. The better the level of information, the lower the probability of dropping out. However, information comes from a wide variety of sources that differ in their influence on dropout.
To improve training decisions and the value of companies‘ apprenticeship programs it is imperative to understand why talented apprentices chose to leave their training companies. This research investigates the following questions: (1) What role do personal links to a company play, alongside the subjective risk of leaving, in apprentices’ decision to (dis)continue employment. (2) What are apprentices‘ inherent attachments to their training companies, the communities they live in and their chosen professions? The latter correspond to the following macro-level embeddedness foci: organization, community and occupation. The common identification heuristic for micro-level embeddedness factors employs the categories of links, fit and sacrifice of working individuals. Framed by embeddedness theory (Ng and Feldman 2007; Holtom and Darabi 2018), the literature proposes various predictors to apprentices’ acceptance of post-apprenticeship employment offers from their training companies. These predictors are: the number of apprentices in the training company, task variety, apprenticeship satisfaction, perceived job security, as well as individuals’ perceived importance to continue working within the company, the profession and their place of residence. The NEPS data allow the development and the validation of a logistic regression model using independent samples across the German population within the predictive research paradigm (cf. Yuan et al. 2021). SC4 data are used to train this logistic regression model, while the SC3 data are used for the validation. ROC-AUC-scores reveal the predictive utility of a parsimonious model including the macro-level embeddedness foci of organization and occupation. A 3% likelihood of a completely embedded apprentice to decline a post-apprenticeship employment offer increases to 57% when the apprentice perceives the company and the profession as less important. Declining an employment offer reaches 74% when the apprentice is also risk seeking and was part of a company with few apprentices. The analysis tests embeddedness theory at the transition from apprenticeship to employment and proposes a guideline for predictions at that point. The results reveal community and task variety as insignificant predictors. Job security and company size, through the number of apprentices, partly influence refusals independent of embeddedness. We can recommend training companies to focus on organizational and occupational attachment starting with the recruitment of apprentices without offering them immediate job security. These conclusions are limited by the lack of objective information on company characteristics, possible reciprocal influences among apprentices as well as any intergenerational changes in attitudes and values that may affect embeddedness.
This contribution examines the development of occupational self-regulation during the transition from teacher training to employment as a teacher in Germany. Self-regulation, i.e. the ability to manage one's own resources effectively, is considered a skill of great importance for successfully coping with the demands of everyday working life (Klusmann 2011). Based on the conservation of resources theory (Hobfoll 1988), it is assumed that individuals with many resources are more likely capable of health-promoting self-regulation than individuals with less favorable resource endowment. In addition, they are expected to be at lower risk of deterioration in their ability to self-regulate during the transition from teacher education to employment. Resources are material and immaterial assets that can be used to equip oneself against potential stress (Buchwald & Hobfoll 2013). We include final grades (as proxy for cognitive performance), self-efficacy expectations and Big Five personality traits and control for contextual characteristics of the teaching profession such as the type of school or weekly hours of work. Analyses are based on data from n=401 participants from the NEPS Starting Cohort „First-Year Students“ (Blossfeld et al. 2011), who transitioned from teacher education to the teaching profession between two online surveys in 2016 (T1) and 2018 (T2). Self-regulation was measured by a short scale of the Occupational Stress and Coping Inventory (AVEM) by Schaarschmidt and Fischer (2001). It contains two subscales each for occupational engagement and resilience to occupational stress. All subscales have good internal consistency (for both measurements α≥0.77). The factor structure of the instrument was examined via confirmatory factor analyses (T1: RMSEA=0.057; CFI=0.959; T2: RMSEA=0.061; CFI=0.961). (Partial) Scalar invariance between the measurement times was proven. Based on the scales, four self-regulation patterns can be distinguished (H: healthy-ambitious, U: unambitious, A: exessively ambitious and R: resigned; see e. g., Menge & Schaeper 2019). Two patterns (A and R) are considered risk factors for the mental and physical health, especially due to their low levels of resilience to occupational stress. To examine the stability of these patterns over time, a latent transition analysis was estimated. It displayed that 68% of the sample retained their original self-regulation pattern after entering the teaching profession. Particularly the health-promoting patterns were characterized by a high degree of stability; transitions occurred more frequently from risk patterns to health-promoting patterns. The extent to which this can be attributed to different entry conditions of the participants will be analyzed and discussed.
In this article, we examine how the improvement of parental education across cohorts affects sons’ and daughters’ tertiary educational attainment in the process of educational expansion. Using data from the National Educational Panel Study (NEPS), we focus on West Germany and study whether the upgrading of the educational composition of origin families across cohorts contributes to daughters catching up with and even outperforming sons in tertiary educational attainment over time. In particular, we ask whether the rise of highly educated families, who are assumed to have stronger gender-egalitarian attitudes toward their children, has contributed to daughters faster increase in tertiary education compared to sons. Our empirical analysis shows that the long-term upgrading of families’ education across cohorts has in a similar manner increased tertiary educational attainment of both sons and daughters. Thus, women’s educational catch-up process cannot be explained by the greater gender-egalitarian focus of highly educated parents. Rather all origin families, independent of their educational level, are following the same secular trend towards more gender egalitarianism. We also examine to which extent highly qualified mothers serve as role models for their daughters. We find that graduate mothers do not serve as particular role models for their daughters. Rather mother’s education is equally important for both sons and daughters success in higher education. Finally, we show that the rising proportion of higher-educated families across cohorts is connected to a rising share of downward mobility for both sons and daughters. However, the share of upward educational mobility from families with intermediate education is higher for daughters than for sons.
Mismatch between one’s skills and job requirements imposes costs on individuals, employers, and society at large. This paper focuses on the mismatch between workers’ educational attainment (supply side) and the educational requirement of their occupation (demand side) (ILO 2014). More specifically, it adopts an intergenerational and comparative perspective to study risks of overqualification, defined as a situation of vertical mismatch where a worker has achieved a level of education higher than required in their current occupation (Erdsiek 2016). While research has documented many factors associated with the risk of overqualification, we know little about how overqualification may contribute to the reproduction of inequalities from one generation to another. Social stratification research finds a substantial relationship between parental background and child educational attainment (Breen and Müller 2020). In addition, some research has focused on inequalities in socioeconomic status attainment between university graduates from privileged social origins (Torche 2011; Witteveen and Attewell 2017). However, that research contributed little to our understanding of the sources of these differences, and more specifically of variations in the characteristics of career trajectories of university graduates by social origin. To address this shortcoming, we ask to what extent parental class background have an influence on one important career dynamic in the UK and Germany: the risk of overqualification and the probability of exit from overqualification. In both cases, family background is likely to influence the risk of overqualification and exit probabilities in various ways. First, family’s financial capital could influence overqualification, because graduates from wealthy families might have the opportunity to search longer for an adequate job. Second, the process of finding a job can be influenced by the parents’ social capital and networks (Jonsson et al. 2009). Third, the transmission of cultural capital from parents to children may contribute to higher chances of accessing the most elite jobs available to university graduates from higher social origins (Laurison and Friedman 2020; Rivera 2015). Finally, social differences in aspiration could also affect individuals’ occupational choices due to the will to prevent downward mobility (Jacob and Klein 2013). To test these hypotheses, this paper provides a comparison of the occurrence as well as the persistence of overqualification in UK and Germany between 2010 and 2020. The results derive from analyses of data from the UKHLS and the NEPS SC6. Exploratory results show that having a father in a higher-status occupation lowers the probability of overqualification.
Background: Numerous studies point to the great importance of the developmental environment in childhood and adolescence for the later life course, for example concerning the development of skills and abilities, income, or health (Gertler et al., 2014; Taylor, 2010; Tymms et al., 2018). Currently, however, few studies examine in more detail whether education policy reforms and resulting changes in the school learning environment during adolescence can have adverse long-term effects on learning behavior later in life. This paper addresses this issue in the context of the G8 reform. The G8 reform reduced the length of Gymnasium schooling from 9 to 8 years and increased the number of hours per week from 30 to 32 to 32 to 34 hours per week. Several studies suggest that students from G8 years had more stress-related health problems overall due to the intensified learning time (Huebner et al., 2022; Huebener et al., 2017; Marcus et al., 2020; Quis, 2018). Learning time is also central to college, particularly for student achievement (Andrietti & Velasco, 2015; Metcalfe et al., 2019; Schwerter et al., 2022). From a life course perspective (Becker & Schulze, 2013), we can therefore derive different scenarios regarding the long-term effects of the G8 reform on student achievement. For example, the reform might have led to habituation or compensation. Research question: This paper takes a closer look at the question of whether the G8 reform in Germany had long-term effects on study-related learning efforts. The learning effort was recorded using self-assessments in hours per week spent on lectures and self-study. Method: Data from the DZHW's German Student Social Survey (Apolinarski et al., 2021) were used to answer the research question. The pooled data set included a total of N = 71,426 students (59% female, MAlter = 23). Results and their significance: Using the difference-in-differences method proposed by Callaway & Sant’Anna (2021), we found negative effects of reform exposure on hours spent attending classes and on self-study. Thus, overall, the estimated reform effects tend to indicate a compensatory process. Thus, we show one reason G8 students' performance was worse (Doersam & Lauber, 2019), without any intelligence difference due to the reform (Dahmann, 2017). Our results support the assumption that the long-term effects of school reforms on individual life trajectories should be studied more systematically and considered when designing new reforms.
The institutional structure of study programs and official recommendations on how students should organize their studies likely influence students' credit acquisition cost function, planning and behaviour -- consequently, these institutional features can limit the potential of some students and affect time-to-degree. In this study, we examine whether institutional design, specifically the European Credit Transfer and Accumulation System (ECTS) leads to bunching in the number of credits per semester. We use data from various sources: administrative student-level data from the examination offices of two universities, as well as nationally representative data from the German National Educational Panel Study (NEPS) and the Beginning Postsecondary Students Longitudinal Study (BPS:12/17). We show that there is significant bunching at 30 attempted and obtained credits per semester in European higher education. We argue that this is due to institutional design based on the ECTS with the study structure fully aimed at a ``typical’’ student obtaining 30 credits. This structure promotes bunching, as it creates (i) a kink in the cost of obtaining credits beyond 30, and (ii) a reference point at 30 credits. The missing mass above 30 credits suggests that a substantial share of students is held back by these rigid features, increasing time-to-degree and causing potentially large individual and societal costs. For the US, where the institutional design is much more flexible, we do not observe comparable bunching. In a companion field experiment, we tested whether an informational intervention encouraging students to take more than 30 credits would change their bahavior, ultimately decreasing the magnitude of bunching in credits per semester. We find that the intervention had no significant effects on changing the behavior of the treatment group, likely because this does not sufficiently alter the ECTS cost structure and reference point influence.
In the long-standing research tradition on the value-added by attending particular higher-education institutions, selectivity has been a prominent measure of institutional quality that relates positively with favourable individual labour market outcomes (Dale/Krueger 2002, Long 2010). However, profiting from institutional quality requires that a degree is completed successfully, and this seems to also depend on the institutional context. International research indicates that selectivity of higher-education institutions increases degree completion rates at 4-year colleges in the US (Dillon/Smith 2020). There is evidence that peer effects, which are often mentioned as mechanisms behind the context effects, are advantageous for grades in shared courses (Lomi et al. 2011, Parker et al. 2010, Winston/Zimmerman 2004) and studies on group-specific variation by student ability, mainly referred to as under- or overmatching, yield inconclusive results on the corresponding profits (Dillon/Smith 2020, Wolniak/Muskens 2021). So far, research mostly stems from the US context. Since the German higher-education system differs significantly in structure as well as the corresponding admission and sorting processes, questions of institution-specific development and graduation chances remain open for this context. Moreover, the definition of selectivity might need to be refined. Institutional selectivity in the German higher-education system not only captures formal selection processes but also the self-selection of prospective students into fields of study and specific higher-education institutions. In contrast to international research where quality is typically measured for whole universities, in Germany fields of study need to be considered for that. Hence, we are interested in whether 1) the empirically observable selectivity of study locations within fields of study is associated with individual academic attainment in Germany, 2) there are group-specific effects depending on individual characteristics (own ability, social background), and 3) there are differences across structural dimensions (fields of study, type of degree). To test our assumptions, we use the NEPS Starting cohort 5 (Blossfeld et al. 2011). Thereby, a cohort of first-year students in winter term 2010/2011 has been followed through their studies. As dependent variable, we use whether a student graduated successfully within a particular time frame. The main independent variable is the institutional selectivity in form of the mean individual GPA aggregated at the cluster level of field of study and higher-education institution. At the individual level, we include the individual GPA of entrance qualifications and social origin as well as the field of study and type of degree for the structural dimensions (as well as relevant interactions). Results of multi-level survival analyses with field of study and higher-education institutions as higher levels of analysis will contribute to international research regarding institution-specific development and success in higher education.
Occupationally qualified individuals who have entered higher education without a formal entrance qualification (Abitur) remain a minority in Germany. They are, therefore, seldom on the radar screen of representative surveys. Consequently, this group cannot be distinctly identified in most surveys, or the available data derives from case studies and are of a cross-sectional design. The German National Educational Panel Study (NEPS, Blossfeld & Roßbach, 2019) is the only available dataset in Germany that provides a concise analytical definition of this minority group in different starting cohorts. Furthermore, the NEPS provides longitudinal data linked to the administrative data of the Institute for Employment Research (IAB, Bachbauer & Wolf, 2020). The proposed talk will explore the income differences of higher education graduates (1) with the German Abitur qualification and without a vocational and educational training degree, VET (direct qualifiers), with VET (dual qualifiers) and (2) an Abitur, and (3) without an Abitur. I use data from a linked data set of the student cohort SC5 of the National Educational Panel Study (NEPS) and the Institute for Employment Research (IAB), NEPS-ADIAB. The analytic sample contains 4,698 direct qualifiers, 1,315 dual qualifiers with an Abitur, and 213 without one. Longitudinal and pooled Poisson regressions are estimated to explore the income trajectories within the careers of dual qualifiers and for the comparison with direct qualifiers. Findings show first that dual qualifiers benefit from higher education. Second, the findings indicate differences in income in favor of dual qualifiers without an Abitur for the population of Bachelor graduates. Master graduates experience no income differences. However, this might be due to the short time I can observe them on the labor market.
Technological and digital change has led to far-reaching structural changes in the labor market, bringing new opportunities but also challenges e.g., increased demand for highly skilled labor (e.g., Acemoglu, 1998) and substitution and complementation of job tasks (e.g., Autor et al., 2003). It can be assumed that digitalization does not take place everywhere at the same pace, but that its spread is shaped by societal and regional structures and barriers. Moreover, digitalization has often been associated with general labor market outcomes, but its specific impact on vocational education and training (VET) remains unclear. Against this background, we ask: how does the level of digitalization in the region affect young people’s status attainment in VET? In this study, we focus on young people’s transition from secondary school to VET, using data from the National Educational Panel Study (NEPS; Blossfeld & Roßbach, 2019), combining information from starting cohort 3 and 4 (2010-2020). We capture the achieved social status in VET as a metric variable indicated by the occupation coded with the International Socio-Economic Index of Occupational Status (ISEI). To identify digitalization as a latent factor, we use the following regional indicators at the district level (NUTS-3) from administrative data sources: broadband coverage, German website domains, IT-capital of firms, the share of employees in scientific IT and service occupations, and registration of digitization patents. To predict status attainment among youth entering VET, we apply general linear models with cluster-robust standard errors at the level of training market regions (Kosfeld & Werner, 2012). The results of the confirmatory factor analysis (CFA) confirm a good fit between the regional indicators and regional digitalization as our latent factor. In line with our theoretical expectations, the multivariate results show a positive association between digitalization and status attainment in VET, with more precise predictions for urban districts. While females attain a higher VET status on average, boys are more likely to close that gender gap in regions with higher levels of digitalization. The situation is similar with education: in highly digitalized regions, young people with an intermediate school leaving certificate attain just as high a VET position as young people with a university entrance certificate. However, digitalization plays a less decisive role for young people with low levels of schooling. Overall, our results show that the degree of digitalization in the regional context has the potential to mitigate existing social inequalities among youth entering VET.
The COVID-19 pandemic has greatly changed the way people work. For example, the rise of remote work has led to a boost in digitization in selected occupations. Yet even before the pandemic, different occupations were characterized by varying prevalence of networked digital technologies. In addition to the degree of digitization, occupations differ in their job quality. Job quality includes not only classic distinction features such as pay, opportunities for promotion, working hours and reconciliation with private life, but also content-related aspects as autonomy or the amount of manageable information. For example, occupations in transport, logistics, protection and security showed particularly low proficiencies in both, the classic job qualities and the degree of digitalization in 2019, while the opposite was true for occupations in business organization, accounting, law and administration. For the pre-pandemic time, we find correlations between positive connoted job qualities and the degree of digitization. Therefore, in the present study we investigate which occupations experienced a pandemic-induced boost in digitization and whether this boost had an impact on job quality. For this, we use data from the Adult Survey of the National Educational Panel Study (NEPS-SC6) and compare the pre-pandemic wave 2019/20 with the wave 2020/21 conducted during the COVID-19 pandemic. Both waves provide a wide range of information on job quality and job satisfaction as well as on the use of networked digital technologies at the workplace. Preliminary results from analyses of variance and hierarchical change score models show that the boost in digitization has indeed affected specific occupations differently, but has so far had only a marginal impact on job quality. However, we find a clear connection between more digitized occupations and an increase in the amount of information, which can stress employees and reduce job quality.
Due to digitalisation and automation processes, work is changing considerably because some job tasks which were previously performed by humans are substituted by machines while others are complemented and gaining importance. This might influence employees in two different ways: on the one hand, automation and changing tasks might increase job strain and affect job or life satisfaction and health negatively through e.g., cognitive overload, loss of meaning, or the fear of technological substitution. On the other hand, automation and task changes might be a chance for employees if technology improves their work when e.g., undesirable monotonous tasks or hard physical work decrease, or by gaining more autonomy, which in turn is associated with better health. However, research on the consequences of task change on working conditions and workers' well-being is rather scarce. We contribute to fill these gaps by analysing longitudinal data from the German National Educational Panel Study (NEPS). We focus on job tasks according to the “Task-Based Approach” by Autor et al. (2003) and examine their impact on job strain and satisfaction with health, living standard, job, and life as measurements of well-being using fixed effects models, allowing for unobserved heterogeneity. The results show that a change in job tasks has a significant impact on working conditions. A reduction in substitutable routine and an increase in complementary analytical and interactive non-routine tasks increase performance and time pressure as well as job strain. This is accompanied by only partial improvements in wellbeing. However, performing less manual and more autonomous tasks increases well-being in many domains. These findings are highly relevant because the future of many jobs will be digital and automated. We offer insights into the impact of changing job tasks on workplace and well-being indicators, raising awareness that tomorrow's work does not inherently come with better working conditions and health, which can be used for e.g., occupational health management.
Description of the research question: School laws that determine the timing of a child’s school start exist in virtually all developed countries. From the literature it is known that school starting age (SSA) affects a variety of outcomes, such as cognitive abilities, educational attainment, fertility, or health. However, little is known about the relationship between SSA and civic engagement in terms of voting behaviour or political activism. This is surprising because civic participation is a key feature of any functioning democracy. This paper fills the gap in the literature by investigating the causal effect of SSA on an individual’s civic engagement in Germany. Data and empirical strategy: To address the potential endogeneity of an individual’s SSA, we exploit policy-induced variation generated by the statutory cut-off rules for school enrolment. Specifically, we apply a fuzzy regression discontinuity design (RDD) where an indicator for being born after the cut-off date serves as an instrument for the actual SSA. For estimations, we use a sample of individuals born in West-Germany between 1944- 1985 from two complementary data sets, the National Education Panel Study: Starting Cohort 6 (NEPS-SC6) and the German Socio-Economic Panel (SOEP). Theoretical approach: From a theoretical perspective, there are several potential channels through which SSA might affect civic participation in the adulthood. First, earlier evidence suggests that the relative older school entrants have a higher probability of taking a leadership role within their cohort, which can affect their opinion and/or willingness of civic engagement later in life. Second, extensive literature shows that older school starters perform better in school in terms of cognitive abilities and therefore they might have a better understanding of topics related to civic responsibility, separation of power within democracies, and the necessity of civic engagement. Additionally, a higher absolute SSA has been shown to improve the secondary school placement which implies that older school entrants are exposed to different peers and potentially build up different social networks. Such network effects might spill over to an individual’s general interest and knowledge about civil engagement. Main findings: Our preliminary results show that higher SSA increases the perceived importance of political and civic engagement. The effect is relatively stable over the entire life cycle.
Investigating the relationship between education and political attitudes is a broad interdisciplinary field of research. Studies from the USA have repeatedly found a positive effect of education on political attitudes and participation (e.g. Campbell et al., 1990), but causal factors remain unexplained (e.g. Bömmel & Heineck, 2020). From the various factors identified as relevant, we focus on two: First, the type of education is especially in segregated education systems, crucial for the development of political attitudes (Janmaat, 2022). Second, the effect of education is mediated by socioeconomic position within society (Persson, 2013). However, as for many studies in this research field, the validity of these results is limited because they rely only on cross-sectional analyses (Russo & Stattin, 2017). In previous research, political interest has emerged as a decisive factor in political attitudes, because of its central function for political behavior and participation (Prior, 2010). Considering this, it seems useful to analyze in more detail the extent to which schooling has an influence on the political interest of adolescents and young adults. Two questions are of interest here: Do the type of education affect the development of interest? Can these educational effects already be observed during or after secondary school? This is precisely where the data from the National Educational Panel come in. Cohort 3 offers the opportunity to analyze the political interest of approximately 4,300 11 to 29 year-olds in the course of secondary school and beyond. The dataset also includes a large number of control variables, especially on other relevant socialization instances (e.g. Koskimaa & Rapeli, 2015). Multilevel models are used to examine the influence of educational characteristics on the development of students' political interest over time. The models show that on average political interest increases during secondary school. Regarding the attendance of different streams, there is largely a parallel development, which does not indicate that the type or quality of education has a causal influence. The observed differences in levels of political interest seem to have manifested before secondary school. After finishing school, however, divergent development in political interest according to school stream can be observed. Thus, the effect of schooling may be mediated by the socioeconomic position achieved after school, as another study has shown (Persson, 2013). Future analyses with the dataset should start at this point and, moreover, separate selection and socialization effects more precisely.
We examine whether growing up or living in a socialist country, with its limited access to free media, has a lasting impact on trust in conventional media, press and television, and new media, such as social media, decades after the socialist regime was abolished. For identification, we exploit two features of German history as natural experiments. First, we use the separation and subsequent reunification of the two Germanies to compare East and West Germans, and thus differences in access to free media. Second, we compare East Germans with and without access to free West German media. The identifying variation arises from differences in reception of West German television, based solely on geographic characteristics, namely the shortest distance to the nearest televion tower on the East-West German border. Using data from the German National Educational Panel Study (NEPS), our results show that East Germans trust less in press or television than West Germans. In contrast, we find that East Germans are more likely to trust social media. Results of analyses based on Eurobarometer data for the EU-28 countries support this finding. The dataset also allows to study trust in radio and the Internet. We take advantage of this data and examine the differences between former socialist and non-socialist countries. Consistent with our earlier findings, individuals living in a former socialist country trust the press, television and radio less, but social media and the Internet more. Further, we explore the importance of trust in conventional and new media by investigating the role of trust in combating the COVID pandemic. We provide evidence that COVID vaccination rates in the EU-28 countries correlate positively with trust in conventional media (press, television, radio) and strongly negatively with trust in new media (social media, Internet). This is crucial, as a high vaccination coverage against COVID predicts better physical and mental health and thus less severe negative impacts on health care and the economy. We take this as suggestive evidence that experience with free media has important long-term consequences for individuals’ trust in media and the information they provide, even today.
ETheories like rational choice or the socio cognitive career theory postulate some kind of rational or interest driven study choice. These theories, no matter how different they are, presume that students show a certain kind of commitment to the study subject they chose. This presumption, however, implies that students are able to study some kind of study subject that they desire – or at least which they are willing to commit to. Yet, this may not always be the case, e.g., because of access restrictions, financial issues, or other kind of unavailability of a study subject. While most theories that focus on predictors for study outcomes implicitly assume that students study a subject that they are willing to commit to, less is known about students that study a subject that they don’t desire. The current study focuses on latter students and analyses following research questions: a) how far are students with an undesired study subject less successful than their mates? b) are there differences between various study subjects? c) are some subjects especially characterized as “alternative subjects” for an impossible subject choice? The study analyzed 13711 students of 39 study subjects of the NEPS first year student cohort SC5:14.0.0. Students were asked if the subject they study was their desired study subject and also which was their preferred study subject. The study outcome was analyzed by looking into episode data and assessing if the first study episode was finished successfully. Results show that students with an undesired study subject are descriptively less successful than their mates in a desired study subject: a higher proportion of these students failed and less students finished successfully. This observation was significant in 25 out of 39 study subjects. Furthermore, there are notable differences between the subjects in terms of how far they were desired. Subjects like social sciences (62%) or philosophy (68%) were at the lower edge of being named as desired while e.g., medicine was at the top with 93%. For most of the study subjects, less than 10% of the students headed for a specific alternative, but for social sciences, welfare, general law and economics, administration, industrial engineers, general engineers, and traffic each more than 10% to 20% intended to study a specific alternative like e.g., psychology, economics, or politics. These results highlight that the phenomenon of undesired study subjects should receive more attention in research, but also in curriculum design and counselling.
The aim of the present study is to investigate student development of social integration at the university and which role constructivist teaching (social negotiations) plays in this regard. Based on self-determination theory and social inference theory contrary effects can be postulated for the development of social integration as well as the effects of constructivist teaching. Taken together, this study focuses on 1.) the development of social integration during the bachelor studies and 2.) the effect of constructivist teaching on the development of social integration. Capturing students in the middle of their bachelor studies, two time points were used from the Starting Cohort First-Year students of the NEPS study (further details; Blossfeld, Roßbach, & von Maurice, 2011). Students at time point 1 were in the third semester and at time point 2 in the fifth semester of their higher education studies (one-year time-lag). This results in a longitudinal core sample of n = 7619 German students. True-Intraindividual-Change Models were applied to tap intra-individual social integration trajectories predicted by constructivist teaching. The results showed that social integration decreased significantly over the course of one year. Constructivist teaching showed significant positive effects on the level of social integration (intercept) and no effect on the development of social integration (change). The results are discussed in terms of theoretical models regarding group processes. This study contributes to the literature in several regards. First, the results of the study revealed that university student social integration decreased in the middle of the bachelor studies. This result extends previous results, which showed that student social integration in the first year of their bachelor studies increased. Second, the results revealed that constructivist teaching had positive associations with student social integration within one semester. But no effect were found for buffering the negative trend in student social integration during their bachelor studies. Thus, further investigations are needed to disentangle if other instructional methods can influence the decrease of student social integration.
For museums, as informal learning environments, detailed information on visitors and especially on nonvisitors is key to preparing adequate offers. However, representative or population-based audience studies are scarce, and even less information is available on nonvisitors in particular. Furthermore, many audience or visitor studies are focused on specific museums, they often lack nonvisitor information and are usually not comparable due to a heterogeneity of designs (Kirchberg & Kuchar, 2014; Tröndle & Awischus, 2019). Comparative studies of available data that does include nonvisitors, however, often lack appropriate information to find explanations outside the realm of demographic and socioeconomic domains. Even though NEPS is designed as an educational panel study, its sampling and the comprehensive survey program (Blossfeld & Roßbach, 2019) lead to an extensive database which also lends itself to analyses outside the general research directions of its educational core components. One of the rotating instruments, namely on cultural participation (Stocké et al., 2019), thus facilitates access to a specific but hard to survey population: nonvisitors of cultural institutions. In our paper, which has recently been accepted for publication in Poetics (Haag & Specht, in press), we have used NEPS SC6 (NEPS Network, 2020) to analyze cultural participation with a focus on museum visits. Based on NEPS wave 10 and a final analytic sample of n = 6,837, we use partial proportional odds models and average marginal effects to compare non-, casual-, and habitual visitors of museums. We analyze which factors distinguish between them and, most importantly, we describe characteristics of nonvisitors. In our models, we include three major sets of explanatory variables: (1) demographic and socio-economic aspects, (2) cultural-capital-related aspects, and (3) personality-related aspects. We find that particularly objectified and embodied cultural capital (Bourdieu, 1986) show the strongest effects on museum (non-)attendance. Even though nonvisitors are more likely to have lower than tertiary education, have difficulties in getting along on their income, live in smaller communities, and rate their satisfaction with life on a lower level, those aspects are far less relevant than objectified cultural capital, like cultural household possessions, or cultural participation, like leisure reading or attending other cultural events (i.e. cinema; opera, ballet, or classical concert; theater; or rock or pop-concerts). The personality trait openness to experience from the Big Five Inventory (Rammstedt & John, 2007) was the only one relevant for museum visits and significantly distinguishes between nonvisitors and habitual visitors.
The acceleration of the labor market requires both companies and employees to constantly adapt their own resources and skills. For workers, this development leads to a shorter half-life of initial education and requires a willingness to engage in lifelong learning. An essential part of such learning processes takes place in non-formal courses and training programs and is organized or arranged by their employer. Most of the financing is also provided by the companies themselves, e.g. costs for course participation. HR managers therefore have a gatekeeper function for this quantitatively most important continuing education segment. In addition to who participates in training, they also play a decisive role in deciding when. In Germany, the training participation is unequally distributed and mainly explained by theoretical arguments derived from human capital and/or segmentation theory. In particular, highly qualified workers, civil servants and employees of the public sector or of large enterprises, full-time workers, employees under the age of 50, without a migration background and formally underqualified persons show increased participation rates. This paper uses a sample of employees aged between 25 and 55 from the adult cohort (SC6, version 13.0.0.) of the NEPS and contributes to existing research in two ways. First, the theoretical perspective is enriched by considerations from the search and matching approach of Jovanovic. The relationship between the employee and workplace is added to the individual and firm level explanations of selective training involvement, which have so far been considered in isolation.These considerations allow to test for an additional matching-based mechanisms that drive training behavior. The analyses are based on multilevel logit regression models. Furthermore, the findings are used to reclassify previous work on determinants of selective training participation. In terms of the presented search and matching explanation, for instance, career stability following training would not be the result of training efforts, but would instead depend on unobserved (matching quality) heterogeneity: A good match between qualifications and job requirements as well as a good skill-match increase the chance of training, while at the same time counteract labor market mobility. Second, theory-based assumptions on the timing of non-formal training are developed and tested by non-parametric event analysis. Thus, it can be shown for the first time that the unequal distribution of training opportunities also has a temporal dimension, with advantages for underqualified and disadvantages for overqualified employees.
In the age of demographic change, skill shortage and digitalization, lifelong learning is often perceived as a solution to combat the negative consequences of these trends. However, despite a growing number of studies on the topic, it is still debated whether and how participation in further training helps individual workers. In this paper we focus on the returns to training among a vulnerable group: the temporary employed. This form of work is described in the literature often as either serving as a stepping stone into (permanent) employment or acting as a trap and leading to continuously precarious employment biographies. We address the important question whether non-formal further training promotes the transition to permanent work and thus helps to escape precarious employment trajectories. According to human capital theory work-related training increases workers productivity and above that might also act as a positive signal to employers as signaling theories suggests. At the same time, it is unclear if firms really invest in the skills of temporary workers when their perspective in the company is unclear. In this study we test several hypotheses addressing these theoretical assumptions using NEPS SC6 data and event history models. Results suggest positive effects of employer-financed training participation on transitions to permanent positions within the same firm. Thus, further training can be part of a stepping stone out of precarity. When it comes to self-financed training participation however, we find no association with transition probabilities.
The world’s population is ageing. Ageing has an impact on all aspects of life including the social, economic, and political domains (Boulton-Lewis 2010). The World Health Organisation (WHO 2002) has proposed a model of active ageing that focuses on optimizing opportunities for health, participation, and security in order to enhance quality of life for old people. Learning and education are both essential aspects of participation and have been proposed as a 4th pillar of active ageing in recent years (Moody 2004, Formosa 2019). Lifelong learning covers all ages, educational institutions and forms of learning (Schäfer 2017). It takes place in different, sometimes unconventional settings and is an open-ended process (Mercken 2010). While the WHO (2001), the OECD (2006) & the EU (European Commission 2006) have recognized the learning needs of older people, the fulfilment of these needs remains unclear. My PhD project focuses on retirement as a special life stage with regard to learning. Retired individuals enjoy a late freedom (Rosenmayr 1987), given that the circumstances are favourable (good health, sufficient income etc.). They are no longer the main care taker for children and free from professional obligations. Therefore, in my PhD project, I investigate different types of learning activities (informal, non-formal and formal; Allmendinger et al. 2019) and the participation of retirees in each activity. The focus of this presentation is on informal learning (e. g. reading non-fiction books, watching learning videos, visiting fairs) and non-formal classes such as sports, art or language classes. For my analysis, I use NEPS SC6 data and draw an analysis sample of about 1500 retired individuals (age range: 60-78 years). Descriptive analyses reveal that only 19 % participated in at least one non-formal course during the last 12 months, which is way below the average rate in the working population (BMBF 2018). At the same time, about 45 % learn in informal ways. The following questions arise from this observation: Who participates and who is excluded from non-formal classes during retirement and why? Which relation exists between non-formal and informal learning? What benefits do learners have, compared to non-learners?