Volume 6, No. 1, Art. 44 – January 2005

Secondary Analysis in Exploring Family and Social Change: Addressing the Issue of Context

Val Gillies & Rosalind Edwards

Abstract: This paper explores the issue of context in conducting secondary analysis and draws out the particular epistemological, methodological, practical and ethical challenges associated with the re-use of historically specific, contextually bound archived datasets. Focusing on the topic of change and continuity in family life, it outlines our efforts to design a method of comparing contemporary data on parents with accounts collected in the 1960s. We discuss the contextual complexities and constraints we encountered in attempting to construct a viable approach to evaluating social change through comparative historical analysis, and we describe how we sought to address these issues.

Keywords: secondary analysis, context, qualitative research, social change, parenting, family

Table of Contents

1. Introduction

2. Secondary Analysis in Evaluating Social Change

3. The Context for Investigating Family Life: Demise, Transformation or Continuity?

4. Addressing Context: Substantive and Conceptual Questions

5. Context and the Research Relationship

6. Working Within the Constraints of Context

7. Conclusion






1. Introduction

While the secondary analysis of quantitative data is a well established practice, data sharing amongst qualitative researchers is far less common. In the UK efforts are being made to archive and encourage the re-use of qualitative data, but methodological concerns about the extent to which detailed, situated studies can be re-analysed have hampered the development of this approach. From a quantitative perspective, data exists independently from the researcher and can be re-used in the future to assess the reliability and validity of particular findings. This contrasts sharply with most qualitative approaches in which interactions between researchers and interviewees are viewed as crucial in shaping interpretations. The significance placed on context in facilitating qualitative understanding is often conveyed through reference to the intimate bond that the researcher inevitably develops with the data, particularly when they have designed the framework, immersed themselves in the field and drawn on personal grounded insights to make interpretations. Such methodological concerns are compounded when qualitative data re-use is proposed as part of a comparative historical analysis. Attempts to compare data sets across different time frames raise numerous questions about contextual commensurability. Nevertheless, raw data from early qualitative studies represents a valuable historical record, documenting the concerns and experiences both of interviewees and researchers at a particular point in time. As such, early source data has the potential to provide a crucial snapshot of social history, thereby enabling sociological understanding of change and continuity. [1]

In this paper we describe our attempts to think through the issue of context to generate a methodologically feasible proposal for re-using data from the 1960s. These attempts are themselves taking place in the contemporary context of preoccupation with the issue of social change. Having recently conducted a study of parenting resources1), we are keen to explore theories of change and transformation by comparing our contemporary qualitative interviews with source data collected in the past. For example, we are interested in questions such as whether and how the experience of parenting children has changed in the past half century, and whether and in what ways the resources and social support networks available to particular social groups of parents have changed? We are currently in the process of seeking funding to enable us to revisit and reanalyse some of the archived datasets from classic studies conducted in the 1960s that address the topics of family, community and class2), in order subsequently to counterpose our findings from this reanalysis with our findings from our contemporary study. In attempting to construct a research design for our secondary analysis of historically situated archived datasets, we faced a series of methodological, epistemological and ethical challenges. This paper draws on our preparatory work in order to explore these issues, rather than practical experience. Nonetheless, we believe that we raise matters that need to be considered by researchers undertaking such work. We begin the paper by outlining the potential role for secondary analysis in charting the evolving social landscape, and demonstrate how the context of narratives of change is particularly central to current understandings of our own proposed focus on parenting and family life. We then consider the specific contextual complexities we encountered in attempting to design a method for conducting an historical comparative analysis of parenting resources. We conclude with a discussion of how these constraints might be addressed to conduct a meaningful comparative secondary analysis. [2]

2. Secondary Analysis in Evaluating Social Change

A preoccupation with the nature and impact of social change lies at the heart of sociology. As Liz STANLEY (2002) points out, the discipline was founded in an attempt to theorise and understand the social consequences of emerging structural changes. The theme of transformation can be traced from a 19th century focus on industrial verses pre-industrial societies, to a more contemporary interest in the individualising effects of post-industrialisation. Viewed by some as a part of a "modernist project" to privilege the forces of change and adaptation (ABERCROMBIE & WARDE 1992; BERMAN 1983), this fixation on transformation can be criticised on a number of levels. As Nicholas ABERCROMBIE and Alan WARDE (1992) suggest, sociology has tended to emphasise the forces of social change at the expense of equally significant social continuities. A dominant focus on transition and difference risks obscuring, or even distorting enduring aspects of social life, particularly if social change is regarded as a necessary condition for sociological study. In addition interpretations of social change are never theoretically or ideologically neutral (ABERCROMBIE & WARDE 1992). Disputes rage about the definitions and significance of change and, as Graham CROW and Tony REES (1999) point out, discussions of social transformation are commonly conducted through the language of winners and losers. [3]

For the most part, sociologists derive evidence of social change from large scale quantitative social surveys (for example, in the UK, the General Household Survey or Census). Major sociological theorists such as Ulrich BECK and Anthony GIDDENS use such survey data to evidence transformation theories standing independently from empirical research. Consequently, a common emphasis on macro, demographic change is rarely accompanied by a research led exploration of the impact on lives as they are lived—a vital method of assessing the significance of such changes. A reliance on abstract and quantitative data may overstate the meaning or effect of observable change (RIBBENS McCARTHY et al. 2003, chapter 6), and could lead to a mistaking of cyclical patterns for linear transformation (STANLEY 1992). Nevertheless, transformation is a commonly cited premise in contemporary literature on personal and social relationships, driving theories, models and prescriptions, and forms an overarching context—and indeed rationale—for revisiting previous social studies. [4]

In this respect, in recent years, researchers have begun to adopt a more in-depth empirical approach to social change, conducting comparative re-studies based on classic works from the 1950s by Clive ROSSER and Chris HARRIS (AULL DAVIES & CHARLES 2002) and Michael YOUNG and Peter WILLMOTT (PHILLIPSON et al. 2001). While such re-studies provide a valuable insight into the nature and meaning of change, they stop short of re-analysing the original data. Yet, evolving conceptual sociological frameworks make distinct social transformations harder to detect (AULL DAVIES & CHARLES 2002) without reanalysis. Also there is a possibility that the specific geographical location of such studies reveal changes that are place-specific rather than more broadly applicable. [5]

An alternative approach to the study of social change involves a return to qualitative data collected at a particular point in time in order to conduct a comparative re-analysis from a contemporary perspective. In the UK, the ESDS Qualidata has preserved and archived a number of classic studies conducted in the 1950s, '60s and '70s. Re-use of these early primary qualitative datasets is relatively low and they remain comparatively untapped resources in the field of qualitative research (CORTI & THOMPSON 2004; HEATON 2000, 2004). Yet secondary analysis of historically situated qualitative data has the potential to offer crucial new perspectives on contemporary issues. As Martyn HAMMERSLEY (2004, p.25) notes "[t]he past can set us new problems, or cast old problems in a new light". [6]

3. The Context for Investigating Family Life: Demise, Transformation or Continuity?

The concept of social change is particularly central in contemporary social theorising on family and community life. It forms the contemporary context that stimulated our own interest in secondary analysis of historical data addressing (however implicitly) parenting. Debates in this field, in the main, are structured around the premise that social and economic changes have profoundly influenced the way people relate to one another in family and intimate life (GILLIES 2003). [7]

In one perspective, post-industrialisation is viewed as giving rise to a de-traditionalisation and individualisation of social life, a transformation associated with rising rates of divorce, cohabitation and births outside of marriage. Such "breakdown" of established social ties is seen as leading to the disintegration of moral frameworks; family relationships are characterised by a fracturing of traditional support systems and a decline in values of duty and responsibility. This is said to place great strain on the institution of the family, drastically undermining the practice of good parenting and thereby damaging social cohesion more generally (COLEMAN 1990; DAVIES 1993; DENNIS & ERDOS 1992; ETZONI 1993; FEVRE 2000; MURRAY 1994). Concern over a perceived demise in community relations and trust have generated an interest in social capital as a framework for theorising and promoting social resources. Robert PUTNAM's (1993, 1995, 1996) and James COLEMAN's (1988) work in the field has been particularly influential, focusing on norms and networks, within a "social capital lost" mould (EDWARDS 2004). Both identify diminishing levels of social capital, linking this to perceived changes in parenting and family life. [8]

More specifically, recent years have seen an explicit focus on parenting as a designated area of policy intervention in the UK. This intervention is premised on assumption of the breakdown of social relationships and a loss of collective social norms. From this perspective family comes to exemplify the concept of dutiful community through the practice of parenthood (e.g. Home Office 1998, 2003; and commentaries by DRIVER & MARTELL 2002; WASOFF & HILL 2002). This view has underpinned a number of recent social policy initiatives designed to tackle effects of family change by regulating childrearing practices (MACLEOD 2003). The raft of initiatives include: the National Family and Parenting Institute, Parentline Plus, the Sure Start programme, and the Parenting Fund. Underlying these policies are assumptions about the deteriorating nature of change in contemporary family relationships and support systems. The preoccupation with transformation is also reflected in the attention given to the concept of social capital in policy, focusing on bolstering community and the impact on the resources and support that are generated from social networks (e.g. ONS 2001; PIU 2002; and commentaries in BARON et al. 2000; GAMARNIKOW & GREEN 1999). [9]

Other theorists take a more optimistic view of social change, suggesting that a greater diversity and plurality of lifestyles generates new opportunities for more democratic family relations and the resources that parents can draw on for support. People are seen as now seeking more fulfilling family and community relationships based on egalitarian values of respect and negotiation, as opposed to duty and obligation, hailing a "new golden age" of social capital. Trends towards cohabitation, separation and re-partnership are viewed as indicators of a shift in family relations from a "community of need" to "elective affinities", with parents building their own social networks, seeking out and accessing support and knowledge for themselves. New "families of choice" are said to be emerging from a context of diverse social interactions, marking the generation of alternative social capital networks and resources for supporting parenting (BECK & BECK-GERNSHEIM 1995, 2002; BECK-GERNSHEIM 1998, 2002; GIDDENS 1991; STACEY 1996; WEEKS 1995; WEEKS et al. 2001). [10]

In contrast to arguments of demise or regeneration in family and community life, there is also a "plus ca change" perspective, which questions the extent of social change. The continued importance that individuals place on family relationships and obligations in contemporary studies is highlighted (see review by JAMIESON 1998; and GILLIES et al. 2001; RIBBENS McCARTHY et al. 2003). It is suggested that while analysis of current trends in family forms and household composition can emphasise increased diversity in living arrangements (more divorce, lone parents, step-families), such figures also reveal an enduring continuity of traditional ties, with the majority of families still composed of heterosexual couples and their biological children. In addition, there are claims that diversity and plurality always have been a feature of family relationships (CROW 2002; STANLEY 1992; VASCOVICS 1991 in BECK-GERNSHEIM 1998). Similarly, the emphasis placed on change in social capital is challenged by Pierre BOURDIEU's (1986, 1990) view of social capital as inextricably linked to a number of other central resources, or capitals, which determine an individual's standing as well as their likely trajectory. From this perspective contemporary society is witnessing neither the erosion nor the transformation of social capital, but rather its consistent deployment in the reproduction of privilege and inequality. [11]

The challenge of empirically identifying and theorising change within the context of this contemporary debate deserves a more concerted effort given the significance it is accorded within sociology and to theories of family and parenting in particular (ABERCROMBIE & WARDE 1992; STANLEY 1992; CROW 2000). Many doubt the basis upon which assumptions of social change are made, with writers questioning the way a fixed "othered" past is defined and differentiated from an ephemeral present (ADAM 1996). For example, Graham CROW (2002) claims there was considerably more fluidity and diversity in past family relationships than was previously recognised, while Liz STANLEY (1992) draws on her own qualitative research to show how official, abstract definitions of family structure and employment status can conceal considerable ambiguity and complexity. [12]

Qualitative, empirical research tends to expose the contradictory, tangled complexity of real life experience, which often stands in stark contrast to neatly packaged theoretical accounts of social change. According to Graham CROW (2000), community studies were (and remain) effective empirical tools for evaluating abstract theorising on the nature of social change. He suggests they represent a grounded analysis of social relationships revealing "the local expression of macro-social forces and their impact on ordinary people's everyday activities as they are lived out in the locality" (paragraph 1.3). Although some criticise such studies for their descriptive, atheoretical style, the original data informing them exist as crucial temporal evidence in the face of evolving conceptual and methodological expectations. More specifically, in our view, the existence of archived in-depth community and family studies from the 1960s represent valuable source material for the study of family and social change. This period is often identified in both "social capital lost" and "new golden age" perspectives as the historical point after which either decline or renewal began to take hold. Re-analysis of this early data is an opportunity to capitalise on the situated and specific details of everyday life, while engaging with theories of social transformation in a more critical and systematic way. For us, they offer the ability to compare parenting resources and social networks across time, and assess claims of social change. Liz STANLEY (1992) makes clear that any attempt to pinpoint social change and understand its meaning will encounter the intricate complexities associated with trying to understand the past from the viewpoint of the present. However, identifying and attempting to work within these constraints may bring vital new insights in terms of sociological theory. [13]

4. Addressing Context: Substantive and Conceptual Questions

Exploring social change through secondary analysis requires far more than a simple historical comparison of data. The focus of sociological studies shifts over time limiting the contextual commensurability between different historically and culturally specific data sets and generating numerous conceptual and methodological questions. In attempting a major comparative re-study based on ROSSER and HARRIS's (1965) classic work "The Family and Social Change", Charlotte AULL DAVIES and Nickie CHARLES (2002) were aware of inevitable discontinuities between their source data from the 1960s and the contemporary material they would subsequently collect, requiring them to deviate slightly from the original research design. For AULL DAVIES and CHARLES' restudy, evolving understandings of ethnographic research required consideration and compromise in terms of methods and methodology. [14]

Such considerations would also apply in the context of conducting a meaningful secondary analysis of historically situated data to explore social change, such as our own intentions in the area of resources in parenting. Comparisons across different timeframes will inevitably encounter substantive and conceptual gaps. These gaps are immediately apparent when an attempt is made to match contemporary research data with past studies. Having recently conducted in-depth interviews with mothers and fathers as part of our study of parenting resources, we are able to draw on this as a contemporary marker for conducting a comparative secondary analysis. However, while our contemporary research shares many of the same themes of earlier community studies, it diverges sharply in terms of focus and theory. For example, as discussed above, parenting is a current concern in the UK and reflects a policy-driven preoccupation with parenting support (EDWARDS & GILLIES 2004; GILLIES 2005). Although there is a long history of evaluative, quantitative research on parenting practice, few qualitative studies based on parents' own accounts were conducted in the past. Those that were carried out predominantly concerned themselves with aspects of development and parent-child interaction. As a result, finding suitable historical comparative data to contextualise claims about the decline of social capital in families is not just a matter of returning to similar, earlier research. Equivalent studies simply do not exist. Instead, relevant themes and accounts are likely to be embedded within a range of topic areas that were previously ascendant. For example, the themes of community and class dominated the sociological agenda in the 1950s and 60s. [15]

These studies should provide useful evidence for comparison, but further complications arise from their characteristic focus on specific populations. The classic community studies inevitably centre on social relations in particular geographical areas. Many early studies of social class are similarly place-based in order to explore demographic shifts such as the establishment of new towns or the influx of the middle class into working class areas. In contrast, our contemporary data on parenting resources is geographically dispersed, precluding any area-specific comparison. However, it could be argued that a simple place-based analysis would risk confounding short term and/or area-specific demographic change with more general concepts of social transformation. For example, Bethnal Green in East London, the site of the classic YOUNG and WILMOT study (1957), has been transformed by new patterns of immigration. From this perspective, the range of environments from which our contemporary accounts stem could be viewed as constituting a more rigorous starting point to explore change and continuity. [16]

In line with their place-based characteristics, early sociological studies were, in the main, confined to specific social categories such as class, gender or family structure. For example, some studies focused exclusively on working class communities, on wives/mothers, or on single mothers. Thus, achieving a meaningful historical comparison would require a "disembedding" of appropriate data from a range of sources. However, further complexity is introduced when considering the changing analytical contexts governing research agendas. AULL DAVIES and CHARLES (2002) cite changes in the way class, gender and ethnicity are now conceptualised, referring in particular to the distinct, historically specific framework governing early interview schedules. For example class definitions were derived from cultural markers with limited contemporary relevance, and were based solely on male "head of households". Furthermore, certain questions were deemed to be gender specific at the time of the original study and so were only asked of men or of women. Understandings of ethnicity were similarly time specific, reflecting the predominantly White make up of the original studies. [17]

Issues around original social class definitions are less significant in conducting a secondary analysis, given that a new analytical interpretation of class can be overlain on the original study as long as a range of family circumstances are included. It is more problematic if the secondary analysis reveals that different questions were asked of working and middle class families in the 1950s and 60s. In the case of gender, this divergent focus is to be expected in early studies of mothers and fathers. Meanings of fatherhood have changed dramatically over the last 50 years (for example, from ascribed to achieved—FURSTENBERG 1988), and inevitably this shapes the type of data collected at different points in time. However, to pursue any analysis of social change, these evolving conceptual frameworks require critical evaluation as points of reference in themselves. There is a tendency to view current approaches to social categories as enlightened in comparison with the "politically incorrect" assumptions made in the past. Yet contemporary analytic contexts may be viewed as similarly containing, projecting and promoting an ethos of equality in the face of continuing difference and disparity. For example, in our contemporary research on parenting, mothers and fathers were asked exactly the same questions, but this reflects an ideological shift rather than a practical change in gender roles, as the interviews themselves revealed the primary role mothers still play in childrearing. [18]

In discussing their re-study, AULL DAVIES and CHARLES (2002) also note how the social and practical circumstances of their present day sample differed from their predecessors, most particularly in terms of ethnic status and employment opportunities. Such changes, however, are commonly place-specific rather than universal. Although the minority ethnic make up of the population as a whole has grown substantially, many areas have remained predominantly White while others have seen particular ethnic populations settle. In the case of employment, levels vary from area to area, with high levels of unemployment in locations affected by demise of manufacturing industry. In terms of conducting a secondary analysis to evaluate family change, the most pervasive and widespread employment trend concerns the mass movement of women into the labour market. Other more concrete changes include developments in, and availability of technology like computers, mobile phones, cars, washing machines and other domestic products. In this context an historical comparative analysis of parenting would provide a nuanced insight into how such changes are lived. [19]

Change is to some extent inevitable, but the proliferation of different household and family forms should not be taken as a self evident marker of social transformation. A rigid focus on the structure and definitions of family can belie personal meanings, experiences and practices, and as such normative typologies may well conceal more than they illuminate. However, any secondary analysis of historically located data is likely to be severely constrained by the narrow definitions that were at one time (and sometimes still are) applied to families. Previous studies were confined in the main to married, two-parent, heterosexual couples, although it might be possible to access wider material from oral histories in order to problematise simplified accounts of cohesive traditional versus fragmented contemporary family forms. It is feasible that in the wealth of archived material that exists there might be evidence for more variability in the actual practice of family than the demographic statistics suggest. [20]

Clearly contemporary research findings are just as context-bound. Some barriers to historical comparison (such as the expansion in family typologies) might be read as self evident of change in themselves, but their impact on people's lives has only been assessed from a perspective in which demographic change is conflated with personal experience. Thus it is assumed that family life is different because statistics suggest change has occurred. A question remains as to whether these structural changes obscure enduring continuities in the way people actually live and interact with each other. In short, the issue of context represents an intractable obstacle to the simple measurement of historically situated data sets. Nevertheless, re-analysis of elderly data from a contemporary frame of reference has the potential to change our understanding of the present by generating new perspectives on the past. [21]

5. Context and the Research Relationship

A major methodological issue for us and others attempting to conduct a historical comparison centres on the relationship between the analyst and the data. Many researchers have emphasised the significance of the contextual knowledge that can only be derived from involvement in the research at the time of its collection. For example, Martyn HAMMERSLEY (1997) describes the "cultural habitus" that is acquired through direct involvement in fieldwork, suggesting that the key role of this intuitive knowledge and experience limits the usability of other people's data. This view is reflected in what Paul THOMPSON (2000) calls the "strange silence close to the heart of the qualitative research community" (paragraph 1), manifested in a general reluctance to draw on material created by other research teams. Louise CORTI (2000) argues that the notion of data not existing independently from the researcher has proved to be a serious barrier to the development of qualitative archives for the purposes of secondary analysis. However, as she points out, "re-use" of data is in fact common practice where researchers are employed solely to conduct interviews, or where research teams share material. [22]

Re-analysis of data by researchers responsible for the primary study is a more established method (HEATON 2000), but as Natasha MAUTHNER and colleagues (1998) note, memories fade and personal perspectives change over time, drastically altering the researchers' relationship to the original data. They argue that concerns over the contextual substance of qualitative research masks more fundamental epistemological questions associated with secondary analysis. While it is commonly acknowledged that research findings are social constructions, with constituent interactions and interpretations mediated through culturally and historically specific frameworks, the attention given to the context of primary research suggests they exist as a discrete and somehow authentic entity. MAUTHNER and colleagues argue that "naïve realism" ensues unless reflexivity encompasses the personal, intellectual and theoretical filters through which (primary and secondary) data is viewed. As Janet HEATON (2000, p.2) states "all analysts whether or not they were "there" at data collection, produce (primary and secondary) analyses which are socially contingent". [23]

6. Working Within the Constraints of Context

Having explored the conceptual and epistemological complexities associated with the comparative re-analysis of context-bound data, we sought to design an approach to our own reanalysis that would enable us to work within the limitations and affordances we have outlined. This required us to make various decisions concerning the status we accord to the data, sampling rationales and analytical frameworks. Defining our research framework to take account of the demands of context was further complicated by the practical restrictions of dealing with elderly data. The advent of word processing and availability of compact, efficient tape recorders have transformed the way qualitative data is collected. In the 1960s, transcripts and notes were predominantly hand written, although some of these were subsequently manually typed. Tape recordings, where they were made, have rarely survived. While ESDS Qualidata are currently attempting to categorise and digitalise this original data, lack of funding ensures this is a long and slow process. Thus while we are able to identify numerous potential sources for historical comparison of parenting resources, the most relevant archived studies for our proposed research are all paper-based, consisting of large, typed and sometimes barely legible handwritten, data sets, with no details provided about interviewee characteristics. Consequently, identifying and selecting appropriate transcripts is a labour-intensive, time-consuming process corresponding to the time and effort involved in primary data collection. [24]

While we do not suggest that any objective measurement of social change is possible, we do believe that re-reflecting on markers of the past and the present through a common contemporary lens can broaden understandings, even if the result is to foreground the complexity of distinguishing then from now. It is from this critical perspective that we are seeking to identify data sets for comparison. As we have previously stated, early data on parenting resources and family is likely to be embedded in a range of themes and topic areas including class, social mobility, community and social relations. In fact it could be argued that orthodox theories of social capital have merely reframed these concerns, leading to a differently-termed focus on social exclusion / inclusion, parenting values and support (as Harry GOULBOURNE has argued specifically in relation to minority ethnic families, 2005). As Mildred BLAXTER (2004) notes, historically specific understandings of social problems define research frameworks and, as such, studies are embedded in social trends that may themselves be representative of social change. [25]

In considering a potential sample framework we realised that efforts to exactly match contemporary and historical sample characteristics may be impossible or even counterproductive. For example, the sharp rise in numbers of full time employed mothers suggests they are unlikely to be a well represented category in early studies. Consequently, a contextualisation of the data is required before a sampling frame can be constructed. This amounts to an initial level of analysis to thoroughly assess the relevant archived collections by categorising their contents and logging demographic details of interviewees. In terms of then generating an appropriate sample to act as a historical comparison, it is important to remain sensitive to the possibility that certain characteristics may constitute a marker of social change in themselves. [26]

Once a sample for the historical comparison has been identified, the issue of how to deal with the socially embedded nature of particular research accounts must be addressed. We would argue that meaning is made rather than found. From this perspective any historical comparison has to include an analysis of the original material (alongside the contemporary study) as a socially produced, situated construction. This involves careful analysis of the original researchers' questions, fieldnotes, letters, memos, reports, publications and any other related sources of information. Consultation with original researchers (where possible) would also generate crucial background information. This attention to context is not about filling "gaps" in the data, but rather illuminating the very particular perspectives knowledge was (and is) created from. [27]

Finally, issues around context present the secondary analyst with a number of ethical decisions that require careful thought. Confidentiality and consent are of particular concern when dealing with qualitative accounts of family life. ESDS Qualidata presides over a number of procedures in order to enhance the protection of confidentiality for interviewees who have contributed to their collections, including the anonymisation of datasets and the issuing of user undertakings to guard against the dissemination of identifying information. However, while it is relatively simple to avoid actual personal names and geographical places, other aspects of an individual's life may be conspicuous such as an unusual job or experience. In addition, certain identifying details such as the names of towns or particular employers can sometimes provide a crucial context when presenting an analysis. The use of interviews from the 1960s neutralises the issue to a certain extent as data this old often bears little connection to people's current lives (CORTI, DAY & BLACKHOUSE 2000). In practice though, we would argue that there is a need to maintain a balance between including rich informative detail and disclosing information that might break original promises of privacy and anonymity. [28]

Questions are also raised as to whether renewed consent should be sought from research participants when conducting secondary analysis. While early sociological researchers would have obtained permission from participants to conduct and disseminate the original research they were unlikely to have gained explicit consent for future re-use. However, it would be extremely time-consuming, and in many cases impossible, to trace those who were interviewed in the 1960s in order to seek their explicit informed consent. Further, this practice could be viewed as unethical in itself, given that interviewees were often told that no further contact would be made. Again, it might be argued that the passing of several decades renders the information contained in research material progressively less relevant to participants themselves (CORTI, DAY & BLACKHOUSE 2000). Reflection on these issues underlines the extent to which such ethical dilemmas are situation specific and not amenable to universal prescriptions (EDWARDS & MAUTHNER 2002). [29]

7. Conclusion

In this paper we have sought to consider the role of context in containing and shaping the practice of secondary analysis. We have outlined our own attempts to grapple with epistemological and methodological problems associated with the re-use of contextually embedded qualitative data in planning towards an historical comparative analysis of resources in parenting. We have focused in particular on the feasibility of drawing insights on family change from a comparative secondary analysis. Although we have deliberated on these issues to construct a research design that attempts to work within the constraints of context, we have yet to put it in to practice. We are aware that many more practical and conceptual problems are likely to emerge in the process of actually conducting this comparative secondary analysis. Re-use of early primary qualitative datasets is relatively low, and as such they represent a largely untapped resource in the field of qualitative research (CORTI & THOMPSON 2004; HEATON 2000, 2004). The complex challenges associated with secondary analysis, combined with the shortage of empirical models exploring methods and practices, may partly explain this under use. However, if the issue of context can be addressed, secondary analysis of early data has the potential to generate crucial new perspectives to feed into wider sociological and theoretical debates. [30]


1) The "Resources in parenting: access to capitals" project is part of the Families & Social Capital ESRC Research Group programme of work, funded by the ESRC under award no. M570225001. Details about the Group's remit and the specific project can be found at http://www.lsbu.ac.uk/families/. <back>

2) Examples archived at ESDS Qualidata include John GOLDTHORPE and colleagues' "Affluent Worker" collection, and Dennis MARSDEN's "Salford Slum Re-housing", "Parents and Education", and "Mothers Alone: Poverty and Fatherless Families" studies. <back>


Abercrombie, Nicholas & Warde, Alan (Eds.) (2002). Social change in contemporary Britain. Cambridge: Polity.

Adam, Barbara (1996). Detraditionalization and the certainty of uncertain futures. In Paul Heelas, Scott Lash, & Paul Morris (Eds.), Detraditionalization: critical reflections on authority and identity (pp.134-148). Oxford: Blackwell.

Aull Davies, Charlotte & Charles, Nickie (2002). Piano in the parlour: methodological issues in the conduct of a restudy. Sociological Research Online, 7(2), http://www.socresonline.org.uk/7/2/davies.html.

Baron, Stephen; Field, John & Schuller, Tom (Eds.) (2000). Social capital: critical perspectives. Oxford: Oxford University Press.

Beck, Ulrich & Beck-Gernsheim, Elisabeth (1995). The normal chaos of love. Cambridge: Polity Press.

Beck, Ulrich & Beck-Gernsheim, Elisabeth (2002). Individualization. London: Sage.

Beck-Gernsheim, Elisabeth (1998). On the way to a post-familiar family: from a community of need to elective affinities. Theory, Culture and Society, 15(3-4), 53-70.

Beck-Gernsheim, Elisabeth (2002). Re-inventing the family: in search of new lifestyles. Cambridge: Polity.

Berman, Marshal (1983). All that is solid melts into air. London: Verso.

Blaxter, Mildred (2004). Understanding health inequalities: from transmitted deprivation to social capital. International Journal of Social Research Methodology: Theory and Practice, 7(1), 55-59.

Bourdieu, Pierre (1986). The forms of capital. In J.E. Richardson (Ed.), Handbook of theory for research in the sociology of education (pp.241-258). Westport. C.T: Greenwood Press.

Bourdieu, Pierre (1990). The logic of practice. Cambridge: Polity.

Coleman, James S. (1990). Foundations of social theory. London: Harvard University Press.

Corti, Louise & Thompson, Paul (2004). Secondary analysis of archive data. In Clive Searle, Giampietro Gobo, Jaber F. Gubrium, & David Silverman (Eds.), Qualitative research practice (pp.327-343). London: Sage.

Corti, Louise (2000). Progress and problems of preserving and providing access to qualitative data for social research—the international picture of an emerging culture, Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 1(3), Art. 2, http://www.qualitative-research.net/fqs-texte/3-00/3-00corti-e.htm.

Crow, Graham & Rees, Tony (1999). Winners and losers in social transformations. Sociological Research Online, 4(1) http://www.socresonline.org.uk/socresonline/4/1/crow_rees.html.

Crow, Graham (2000). Community studies: fifty years of theorisation. Sociological Research Online, 7(3), http://www.socresonline.org.uk/socresonline/7/3/crow.html.

Crow, Graham (2002). Social solidarities: theories, identities and social change. Buckingham: Open University Press.

Davies, Jon (1993). The family,is it just another lifestyle choice? London: Institute For Economic Affairs.

Dennis, Norman & Erdos, George (1992). Families Without Fatherhood. London: Institute For Economic Affairs.

Driver, Stephen & Martell, Luke (2002). New labour, work and the family. Social Policy and Administration, 36(1), 46-61.

Edwards, Rosalind (2004). Present and absent in troubling ways: families and social capital debates. The Sociological Review, 52(1), 1-21.

Edwards, Rosalind & Gillies, Val (2004). Support in parenting: values and consensus concerning who to turn to. Journal of Social Policy, 33(4), 623-643.

Edwards, Rosalind & Mauthner, Melanie (2002). Ethics and feminist research: theory and practice. In Melanie Mauthner, Maxine Birch, Julie Jessop, and Tina Miller (Eds.), Ethics in qualitative research (pp.14-31). London: Sage.

Furstenberg, Frank (1988). Good dads—bad dads: two faces of fatherhood. In Andrew J. Cherlin (Ed.), The changing american family and public policy (pp.192-217). Washington D.C.: Urban Institute Press.

Etzioni, Amitai (1993). The parenting deficit. London: Demos.

Fevre, Ralph (2000). The demoralisation of western culture. London: Continuum.

Gamarnikow, Eva & Green, Anthony (1999). The third way and social capital: education action zones and a new agenda for education, parents and community?.International Studies in Sociology of Education, 9(1), 3-22.

Giddens, Anthony (1991). Modernity and self-identity: self and society in the late modern age. Cambridge: Polity Press.

Gillies, Val (2005 forthcoming). Meeting parents needs? Discourses of "support" and "inclusion" in family policy. Critical Social Policy, 25(1).

Gillies, Val (2003). Family and Intimate Relations: a review of the sociological research, families, & social capital. ESRC Research Group Working Paper No. 2, London, South Bank University.

Gillies, Val; Ribbens McCarthy, Jane, & Holland, Janet (2001). Pulling together, pulling apart: the family lives of young people. London: Family Policy Studies Centre/Joseph Rowntree Foundation.

Goulbourne, Harry (2005). Families, communities and social capital. Paper presented at the "Families, minority ethnic communities and social capital" seminar, London, South Bank University.

Hammersley, Martyn (1997). Qualitative data archiving: some reflections of its prospects and problems. Sociology, 31(1), 131-142.

Hammersley, Martyn (2004). Towards a usable past for qualitative research. International Journal of Social Research Methodology: Theory and Practice, 7(1), 19-29.

Heaton, Janet (2000). Secondary analysis of qualitative data: A review of the literature, ESRC research report R000222918.

Heaton, Janet (2004). Reworking qualitative data. London: Sage.

Home Office (1998). Supporting Families: A consultation document. London: HMSO.

Home Office (2003). Respect and responsibility: Taking a stand against anti-social behaviour. London: HMSO.

Jamieson, Lynn (1998). Intimacy. Cambridge: Polity Press.

MacLeod, Mary (2003). Drawing the boundaries between the state and family. The Edge, 13, 25-27.

Mauthner, Natasha; Parry, Odette & Backett-Milburn, Kathryn (1998). The data are out there, or are they? Implications for archiving and revisiting qualitative data. Sociology, 32(4), 733-45.

Murray, Charles (1994). Underclass: The crisis deepens. London: Institute for Economic Affairs.

Phillipson, Chris (2001). The family and community life of older people: social networks and social support in three urban areas. Ageing and Society, 18(3), 259-289.

Putnam, Robert (1993). Making democracy work: civic traditions in modern Italy. Princeton: Princeton University.

Putnam, Robert (1995). Bowling alone: Americas declining social capital. Journal of Democracy, 6(1), 65-78.

Putnam, Robert (1996). The strange disappearance of civic America. Policy, Autumn, 3-15.

Ribbens McCarthy, Jane; Edwards, Rosalind & Gillies, Val (2003). Making families: moral tales of parenting and step-parenting. York: Sociology Press.

Stacey, Judith (1996). In the name of the family: rethinking family values in the postmodern age. Boston: Beacon Press.

Stanley, Liz (1992). Changing households, changing work. In Nicholas Abercrombie, & Alan Warde (Eds.), Social change in contemporary Britain (pp.85-102). Cambridge: Polity.

Thompson, Paul (2000). Re-using qualitative research data: a personal account [48 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research [Online Journal], 1(3), Art. 27, http://www.qualitative-research.net/fqs-texte/3-00/3-00thompson-e.htm.

Thompson, Paul & Corti, Louise (Eds.) (2004). Special issue: celebrating classic sociology: pioneers of contemporary British qualitative research. International Journal of Social Research Methodology: Theory and Practice, 7(1).

Wasoff, Fran & Hill, Malcolm (2002). Family policy in Scotland. Social Policy and Society, 1(3), 171-182.

Weeks, Jeffery (1995). Inventing moralities: sexual values in an age of uncertainty. New York: Columbia University Press.

Weeks, Jeffery; Donovan, Catherine & Heaphy, Brian (2001). Same sex intimacies: families of choice and other life experiments. London: Routledge.

Young, Michael & Willmott, Peter (1957). Family and kinship in East London. London: Routledge.


Val GILLIES is a Senior Research Fellow in the Families & Social Capital ESRC Research Group at London South Bank University. Her research interests include qualitative research methods, family and social class.


Val Gillies

Families & Social Capital ESRC Research Group
London South Bank University
London SE1 OAA, UK

E-mail: gilliev@lsbu.ac.uk
URL: http://www.lsbu.ac.uk/families/


Rosalind EDWARDS is Professor in Social Policy and Director of the Families & Social Capital ESRC Research Group at London South Bank University. She has published widely on issues of family policy, family lives and social capital.


Rosalind Edwards

Families & Social Capital ESRC Research Group
London South Bank University
London SE1 OAA, UK

E-mail: edwardra@lsbu.ac.uk
URL: http://www.lsbu.ac.uk/families/


Gillies, Val & Edwards, Rosalind (2005). Secondary Analysis in Exploring Family and Social Change: Addressing the Issue of Context [30 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 6(1), Art. 44, http://nbn-resolving.de/urn:nbn:de:0114-fqs0501444.

Revised 6/2008

Copyright (c) 2005 Val Gillies, Rosalind Edwards

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.