Young People's Educational Careers in England and Germany

Integrating Survey and Interview Analysis via Qualitative Comparative Analysis

J. Glaesser

• Publisher: Palgrave Macmillan
• Year: 2015
• ISBN:
  9781137355508
• Topic: Education
• Subtopic: Education Theory & Practice

In this comparative study of young people's educational careers in England and Germany, individual factors, social class, school and country characteristics are shown jointly to shape these careers through mutually reinforcing processes. Qualitative Comparative Analysis (QCA) is combined with interview analysis to explore generative mechanisms.

1

Introduction

The motivation for this book is twofold. One aim is to contribute to sociological research concerning young people’s educational careers and the interconnected roles of individual characteristics, home, school, and educational system. The other is to contribute to methodological development in the social sciences by describing how a combination of large n based cross-case configurational analysis with in-depth interviews in two countries can contribute to theoretical development as well as being able to provide explanations for the outcomes of individual cases. In this introductory chapter, I will briefly comment on the background to both these aims before describing the project funded by the Economic and Social Research Council (ESRC) which underpins the research for this book. Since the research presented in this book potentially has policy implications, this will be followed by some comments on policy. The chapter closes with a summary of the book’s remaining chapters.

1.1 Substantive aims

Education forms a central part of young people’s lives. While gaining an education has intrinsic value, leading to a fulfilling life, it is also often seen, more extrinsically, as a means to a rewarding occupation. For this reason, it is often a topic for public debate. Some of the issues discussed are social inequality in education (in other words, the extent to which educational opportunities and achievements are linked to social background), the way an educational system should be organised, a country’s educational performance compared with others in terms of economic competitiveness, education and employment chances, and migrants’ chances in their new country’s education system, to name but a few. Some of these topics have been debated for some considerable time, but interest in them remains high, especially in the face of social change and associated new challenges for all education systems. In this book, I focus on just some of these issues, in particular on the links between social background and educational opportunities and achievements. Some of the other points will be shown to have a bearing on these. I draw on empirical data from two countries, England¹ and Germany. While these two countries share many characteristics, their education systems differ in important ways, and this will enable me to study to what extent these system differences shape educational outcomes, and to what extent social processes override any system differences and shape outcomes regardless of these.

The phrase ‘social mobility’ is frequently used in Britain in the media and by politicians as a shorthand way of describing the idea that people should have equal life chances regardless of social background, and that people from the bottom of the social hierarchy should be able to move up, given appropriate qualifications and skills.² Education is commonly assumed to have a key part to play in making social mobility possible. However, while both social class of origin and social class destination have been linked to education for the individual (for example Goldthorpe, 2003), it does not follow that simply improving individual young people’s educational outcomes will improve social mobility rates for the whole society (Goldthorpe, 2013).³ This claim is based on the observation that the increase in absolute social mobility, that is, the improved chances of attaining a high social class position, observed during the period following the Second World War, was largely the result of a change in occupational structure. More jobs became available in higher social class positions during this period, but the strength of the connection between educational qualification and social class position hardly changed. It is true that over this period, more people achieved higher qualifications, and were therefore able to obtain higher social class positions, but this was ‘demand driven: that is, by the growth of professional and managerial employment that far outstripped the supply of highly qualified personnel’ (Goldthorpe, 2013, p. 442) and not the result of educational reform. So, for an individual, higher qualifications do increase the chances of obtaining a higher social class position, but within the population, an increase in the rate of high qualifications will not, on its own, necessarily lead to an increase in the overall rates of mobility (absolute or relative). Nevertheless, the aim of improving social mobility is often used as a justification for school reforms and an interest in outcomes and educational targets in England.

Likewise, there is a concern with the link between social background and educational outcomes in public debates in Germany, as can be seen, for example, from the debate regarding Germany’s performance in the first Programme for International Student Assessment (PISA) study (Baumert & Schümer, 2001). While this debate was largely focussed on Germany’s performance in the mid-range of participating countries (rather than near the top, as had been hoped for and expected both by politicians and the general public), another significant part of the discussion concerned the finding that in Germany, social background and immigration seemed to have a particularly strong influence on performance. Various reasons were suggested for this finding, among them the existence of a selective tripartite secondary school system. Criticism of selection at the end of primary school has been a feature of educational debates since well before the PISA studies, as is evident, for example, from the debates regarding comprehensive schooling in the 1970s and the introduction of comprehensive schools in some parts of Germany around that time. While comprehensive schools did not become the norm, debates concerning the ideal secondary school system have continued to this day, along with ongoing reforms in all of Germany’s 16 federal states (Länder). A particular concern is with the fate of children who attend Hauptschule, the least academic of the three hierarchically ordered types of secondary school. Originally, this was the most commonly attended type of school, offering a qualification which enabled pupils to take up an apprenticeship. Now, it is attended by around 10 per cent of a cohort, and the prospects of its school leavers are much worse than those of any other school type.⁴ This has led to the Hauptschule being called the ‘Restschule’, or the school for leftovers. This rather cynical term is used to indicate a concern with the social problems that can be present in a Hauptschule and that may be created by a school type whose intake largely consists of young people who are not seen as academically able or motivated, many of whom are immigrants whose knowledge of German varies considerably, and whose parents either cannot or will not seek a ‘better’ type of school for their children.

These brief sketches of some of the debates in the two countries illustrate that the concern with the link between social background and education is widespread. I discuss the sociological, theoretical background to this debate in Chapter 4. Here, I merely gave a flavour of some of the issues debated in both countries which drive both educational policy and individual decision-making.

1.2 Methodological aims

The previous section explained the substantive background to this book. I now turn to the second set of aims, that of illustrating the usefulness of a particular methodological approach. At the core of the project described in this book is a concern with describing and explaining social phenomena and confirming and developing relevant explanatory theory. Causal explanation is a difficult and contentious aim in the social sciences (more on this in Chapter 2). Some argue that only cross-case analyses using large samples, statistical testing and control variables can establish causal models and provide a basis for generalisation, whereas others would see such models as a form of sophisticated description and suggest, alternatively, a stress on the understanding of generative or causal mechanisms based on theoretical knowledge and the in-depth analysis of a small number of cases.⁵ My view is that, on its own, either approach falls short of being able to provide causal explanation, and I therefore combine cross-case analysis of survey data with in-depth case studies in order to gain deeper insights than would be possible from either form of analysis on its own.

My method of choice for analysing survey data is Charles Ragin’s increasingly popular Qualitative Comparative Analysis (QCA) (Ragin, 1987, 2000, 2008) which I introduce in Chapter 3. QCA is a conjunc-tural method which relies on Boolean algebra and set theory. It aims to preserve the case as a configuration of characteristics as the basis of analysis rather than analysing variables as if they could be treated almost as actors and their effects on some outcome could be identified net of that of other influences. The configurational context of any factors under study can therefore be taken into account. As well as for the cross-case analyses, I use QCA to inform the selection of cases, via a typological analysis, for interviewing. Subsequently, the results of the analysis of these interviews enabled me to refine some of the QCA analyses.

Another feature of the research design is the use of two comparison countries, England and Germany. Given that, among other things, the organisation of secondary schooling and higher education in the two countries is different, the international comparison will allow a deeper understanding of adolescents’ careers through the analysis of cross-national similarities and differences. It will be possible to identify mechanisms and processes that are shaped by organisational features specific to one country, as well as those that are more universally applicable.

1.3 ESRC project

The research reported in this book was made possible through an ESRC-funded research fellowship I held from 2009 to 2011. It arose from joint work I had been undertaking with Barry Cooper (who also was my mentor during the fellowship) on QCA and large n analysis applied to British secondary survey data, with a substantive interest in the sociology of education (see for example Cooper & Glaesser, 2008; Glaesser, 2008). I sought the ESRC’s support because I wished to take this work further and gain deeper insights into sociological processes by combining QCA with process-tracing interviews (George & Bennett, 2005). To this end, I identified appropriate existing survey data from both England and Germany (for details see Chapter 4) and I undertook 79 interviews with 16 to 18 year olds in both countries (43 in Germany and 36 in England). As noted above, choosing to study two countries was done for both substantive and methodological reasons. In addition, I have some knowledge of both countries which I was able to draw on. I am German and, having received my own schooling and university education in Germany as well as having studied educational careers in Germany for my PhD thesis, I was able to draw on relevant contextual knowledge in devising the interview schedule for the semi-structured interviews and in responding to points raised by the interviewees. Likewise, I was familiar with the English system, having lived in England for five years by the time I undertook the interviews, and having worked in an education department in an English university (Durham) during this period.

1.4 Policy

My key concern in this book is with gaining methodological understanding, that is, an understanding of how causal knowledge may be acquired by social scientists, as well as substantive, sociological insights. I do not undertake any policy evaluation as such. However, I expect some of my findings to be of potential use to policymakers, in particular regarding the set-up of a school system and the implications of a system’s properties for social processes and outcomes. As indicated above, the debate on selective versus comprehensive schooling has been going on in the two countries under study (as well as many others) for some time and shows no signs of being resolved. The comparison I undertake might therefore prove useful to those having to make decisions in this respect. Such comparisons between the educational systems of different countries with the aim of learning from a different system are not a new idea; policy-borrowing has a long history. Phillips (2000), for example, points out that comparative research has frequently formed the basis for implementing some aspects of another country’s educational policy or system. However, it can be risky merely to pick and choose isolated elements of another system since the various elements act in conjunction with others. They have also developed in the context of a particular society and culture, which means that a single policy or aspect of a school system cannot always be imported easily. Politicians also occasionally make reference to another country’s system by way of justification of their own preferred policies, but this is not always based on sound empirical evidence (Morris, 2012; Phillips, 2000). My findings compile the sort of empirical evidence that would be suitable to inform policy decisions regarding the selectivity or otherwise of secondary schools, and might have something to offer for teacher education and professionals working in careers advice services. I explain why in Chapter 9. Policy evaluation and policy recommendations frequently make use of randomised controlled trials (RCTs) and other forms of experimental research. I briefly discuss these methods in Chapter 2.

1.5 Structure of the book

In the remainder of this chapter, I summarise each subsequent chapter briefly. In addition to giving an overview of the structure of the book, this is intended particularly for the benefit of those readers who may not wish to read the whole book, but who prefer to focus on particular aspects. For those particularly interested in the methodological approach, I suggest reading Chapters 2, 3 and 9, and possibly one of the substantive Chapters 5 to 7 for an illustration. For those who are mainly interested in the substantive, sociological findings, it is possible to skip Chapters 2 and 3, especially if they have some prior knowledge of QCA. Chapter 4 introduces the theoretical background to Chapters 5 to 8, so should be read regardless of the particular substantive focus of interest. Chapters 5 to 8 then each have their own particular substantive focus and it is possible to read just one or all of them. I suggest reading Chapter 9 for a summarising overview, conclusions and discussion of policy implications, regardless of the particular substantive point of interest.

Chapter 2 outlines my view of the relationships between description, causal explanation and theory, building on the assumption that causal mechanisms generate observed empirical regularities. Good description can be important and useful in its own right, but it is also required for successful explanation, the aim of much of social science. This sounds obvious, as Merton (1987) notes: ‘In the abstract, it need hardly be said that before one proceeds to explain or to interpret a phenomenon, it is advisable to establish that the phenomenon actually exists, that it is enough of a regularity to require and to allow explanation.’ In this chapter, I suggest that statistical models, for example those based on regression analysis, can describe such regularities, or ‘establish the phenomenon’ (see also Goldthorpe, 2001), but that other forms of sophisticated description can be provided by configurational methods such as QCA. Pawson (2008) suggests that both regression-based and configurational models fall short of providing knowledge of generative or causal mechanisms. While I agree with this view, I also suggest that QCA as a configurational method is better suited to describing the complex cross-case regularities that characterise social life than regression-based analyses. I then show, drawing on other studies as examples, how various methods can and should be used together to analyse generative mechanisms.

Drawing on some key publications (Cooper & Glaesser, 2011; Ragin, 1987, 2000, 2006b, 2008), Chapter 3 provides an introduction to the central features of QCA. QCA is the core method employed throughout the book for the cross-case analyses and to construct the typologies of cases from which interviewees are chosen for in-depth study. This typological approach allows me to integrate quantitative and qualitative approaches.

QCA draws on set theory, Boolean algebra and the concepts of necessary and sufficient conditions. Sufficiency and necessity involve sub-sethood relations. If a condition is sufficient for an outcome to occur, the set of cases with the condition is a subset of the set of cases with the outcome. Conversely, if a condition is necessary for some outcome, the set of cases with the outcome is a subset of the set of cases with the condition. Ragin (2006b) has developed measures of quasi-sufficiency and quasi-necessity, in other words, an index of how close a relationship between condition(s) and outcome is to perfect sufficiency/necessity. This is important given the fact that social relationships are not usually perfect, that is, without exceptions to general patterns, so the notions of quasi-sufficiency and quasi-necessity are useful. Likewise, conditions often do not act in isolation, but in conjunction with other conditions. QCA is able to describe the complex regularities that arise from such conjunctural causal relationships. Boolean notation is used, where capital letters stand for the presence of a condition and lower case letters for its absence, or logical NOT. The * symbol represents logical AND, the + logical OR. An illustrative example of a QCA model from my Chapter 6 is ALEVEL_1P*SC_1P + male*SC_1P => FT_EDU17. This may be read as indicating that the configuration ‘having at least one parent with A-levels’ AND ‘having at least one parent in the service class’ OR ‘being female [=NOT MALE]’ AND ‘having at least one parent in the service class’ is sufficient for ‘being in full-time education at the age of 17’. Such a model, or ‘solution’, to use the proper term, is identified on the basis of a so-called truth table which maps out all the possible combinations of conditions and the relationship of these configurations with the outcome. All these concepts and technical details are explained in Chapter 3.

Chapter 4 gives an overview of the design of the study on which this book is based. It st...

Table of contents

• Cover
• Title
• Copyright
• Contents
• List of Figures, Tables and Boxes
• Acknowledgements
• 1 Introduction
• 2 Description and Explanation: Methodological Rationale
• 3 Qualitative Comparative Analysis
• 4 Overview of the Project
• 5 Which Young People in Germany Are in a Gymnasium at the Age of 17? A Typological Analysis
• 6 Secondary Schooling Careers in England
• 7 What Types of Young People Are Bound for Higher Education at the Age of 17?
• 8 Using In-Depth Case Studies to Explore Causal Mechanisms in Detail
• 9 Summary and Conclusions
• Notes
• References
• Index