Key Variables in Social Investigation

Robert Burgess, Robert Burgess

• Publisher: Routledge
• Year: 2018
• ISBN:
  9781351169981
• Edition: 1
• Topic: Ciencias sociales
• Subtopic: Sociología

Key Variables in Social Investigation encourages sociologists and other social scientists to think about the conceptual and empirical problems of using and evaluating key variables in social research. The book contains reviews of ten major variables: age; gender; race and ethnicity; health and illness; education; social class and occupation; work, employment and unemployment and unemployment; leisure; politics; and voluntary ways in which concepts can be specified and translated into variables and indicators.

1

Introduction

Robert G. Burgess

In the preface to an earlier discussion on the use of variables in social research T.H. Marshall (1969) remarked that ‘one essential condition for the development of sociological theory and of sociology as a scientific discipline is the comparability of data assembled by those engaged in research’ (p. viii). Most sociologists and social researchers would be in broad agreement with this statement which highlights the need for sociologists to be aware of the ways in which comparability can be achieved in first hand empirical study, the secondary analysis of existing data sets and the use of published research. In each case researchers are engaged in considering basic questions relating to the definition and use of concepts, the identification of variables and their measurement where this becomes appropriate.

While these aims and objectives can be easily identified their use is much more problematic. Indeed, critiques of social research seem to suggest that many social researchers have yet to come to terms with the main questions and difficulties involved in the use of concepts, variables and indicators in social research for Smith (1975) has identified the main criticisms arising from current research as being characterized by:

1. lack of conceptual clarity and theoretical precision;
2. the failure to clearly identify variables and hypotheses statements;
3. misappropriate selection of indicators used to measure variables;
4. lack of sampling controls;
5. inattention to construction of measures and scales;
6. inattention to problems of reliability and validity;
7. the failure to consider alternative interpretations of data;
8. disproportionate attention to psychological variables at the expense of social structural variables; and
9. problems defined on the basis of untested assumptions and selected for political reasons.

(Smith, 1975, pp. xvii-xviii)

This volume takes up some of these issues as the essays focus on the ways in which researchers who are engaged in empirical research can define and collect data about key social variables. It seeks to bridge the gulf which divides the theoretical and empirical realms, particularly in sociology, by analysing key concepts and discussing how they can be studied empirically. There are still a number of gaps in the social science research literature, as despite developments in substantive sub-fields of sociology, researchers have still not bridged the gap between theory and research, and between theoretical debates and empirical practice. As a consequence, basic questions concerning the place of sociological concepts and the rise of key variables in sociological research still need to be answered as do major problems involved in the relationship between data collection and data analysis.

The objective of this volume is to follow, complement and update the essays in Stacey (1969) and Gittus (1972). However, over ten years have elapsed since the last volume was published and this book therefore seeks to provide some guidance in the light of advances in empirical research during the 1970s and early 1980s; especially in Britain. The rationale for the book lies in the perennial problem of bridging the gap between theory and empirical evidence, in trying to relate the one to the other meaningfully, so that data are not treated merely as a formless mass nor are a priori theoretical categories merely imposed on the data.

This book, therefore, contains papers that address three broad themes. First, how to strengthen the links between social science theory and empirical research. Second, how to use general concepts in empirical research. Third, how to encourage comparability between different pieces of empirical research into similar phenomena. While each of the contributors has covered these themes to a greater or lesser degree the whole volume sets out to provide a critical analysis of key variables in social investigation by:

1. reviewing key basic concepts and indicators used in social and sociological research;
2. examining the relationships between data collection and data analysis bringing out their crucial interdependence;
3. providing some guidance on the kinds of questions that can be posed in a field of study with a view to generating comparable data.

There is no list of variables that all sociologists would agree constitutes the best descriptive schema. However, a fairly well standardized list has been identified by Zetterberg (1965) as the schema that pollsters used in describing a person. Many of these items are referred to as face sheet variables and classificatory variables that are selected from the following list:

1. Past Contextual Variables
   1. Place of birth (native or foreign); sometimes also parents’ place of birth
   2. Type and size of community in which most of childhood was spent (rural, small town, city, metropolis)
2. Present Contextual Variables
   1. Type and size of community in which the respondent lives
   2. Geographical region of country
3. Contemporary Statuses: Ascribed
   1. Sex
   2. Age
   3. Ethnic background
   4. Religious affiliation
4. Contemporary Statuses: Achieved
   1. Occupation; sometimes also husband’s or wife’s occupation
      1. classified according to occupation rank (upper, middle, lower)
      2. classified according to work situation (salaried, self-employed)
      3. classified according to institutional realm (business or industry, civil service or politics, education or science, religion, art, welfare, institutions, private household)
   2. Family Statuses
      1. Marital (single, married, widowed, divorced)
      2. Parental (no children, children living at home, children living away from home)
   3. Memberships in voluntary associations (including business associations and unions); political party affiliation or preference
5. Past Statuses
   1. Father’s occupation
   2. Type of schools attended
   3. Military Service
   4. Past full-time occupations
6. Stratification
   1. Riches
      1. Family income
      2. Family property
         1. Residence (owns, rents, boards)
         2. Consumer goods (e.g. auto, TV)
   2. Knowledge or competence
      1. Years of schooling; sometimes also husband’s or wife’s years of schooling and children’s education
   3. Power
      1. Executive position
      2. Political office
      3. Office in voluntary associations
         (Zetterberg, 1965, pp. 58-60)

In addition a British Sociological Association working party convened by Margaret Stacey in the mid 1960s found that the most common variables used in social research included: age, sex, marital status, occupation, family and household size and composition, education, income, place of birth, housing, leisure activities, social class, religion and politics. It was these lists of variables from Zetterberg and from Stacey’s group that helped in the identification of variables that were to be reviewed here. In particular, attention has been given to variables that had not been examined in the earlier volumes (Stacey, 1969; Gittus, 1972) and those such as education that needed updating, given the administrative changes that have occurred in the intervening period.

Common to all the chapters is a language of concepts, variables, and indicators all of which are treated in a colloquial way within the social sciences. Accordingly some brief definitions are required here as a general guide to the reader (cf. O’Muircheartaigh and Francis, 1981; Mann, 1983). The main terms (which are also discussed in the final chapter in this volume by Bulmer and Burgess)are as follows:

Variable: The representation of a social characteristic or social factor in empirical research. Variables are constructed by defining a concept and developing an indicator or indicators for a concept.

Indicators: A means of representing or measuring sociological concepts using items for which empirical data may be collected.

The key issue involves the way in which concepts can be operationalized and the gap bridged between concepts and indicators in order that theory and research can be linked. Among sociologists, it is Paul Lazarsfeld who has suggested that the main questions to be addressed are: How are broad conceptual ideas converted into instruments of empirical research to provide evidence on a topic of enquiry? How can the ‘variables’ so developed be manipulated to lead to broader generalizations? Lazarsfeld (1958) suggests that the typical process involved in establishing ‘variables’ for measuring complex social objects involves four steps:

1. Breaking the concept down into dimensions that are essential to translate it into any kind of operation or measurement.
2. Specifying the dimensions of the concepts.
3. Selecting observable indicators.
4. Combining indicators into indices.

As subsequent contributions to this volume indicate this is an ideal sequence that provides a framework within which the researcher can operate.

Although the following chapters cover a range of different areas there are a number of issues which are common to several variables:

1 Variables involved in the research process

Recent discussions of research methodology (Hammersley and Atkinson, 1983; Burgess, 1984) have indicated how certain variables: age, sex and gender, race and ethnicity and social class are not only variables in the substantive area of investigation but also how these characteristics of the researcher may influence the relationship with the researched. In this volume, the essay by Morgan on gender draws specific attention to the ways in which this variable influences the research process. Clearly, age, race and ethnicity, social class and educational background are also social characteristics that can have some bearing on the relationship between researcher and researched. Although not all the authors have considered this element it is as well for readers to recall that background variables assist both the researcher and the researched to place each other within the social structure.

2 The interaction of variables

While each of the contributors has been invited to review a particular field of study several have indicated the importance of interconnections between variables. For example, Macintyre highlights the importance of the relationship between social class and health, while Burgess reviews some of the ways in which social class and education relate to each other. Meanwhile, both Parker and Pickvance consider from different perspectives the interconnections between work and leisure. Further discussions of the relationship between different variables are contained in many of the chapters; a situation that leaves the reader to consider how variables may be used alongside each other.

3 Variables in quantitative and qualitative research

It has become commonly assumed that there is a sharp dividing line between quantitative and qualitative research (cf. Filstead, 1970; Halfpenny, 1979) with the result that some commentators have assumed that the language of variables and indicators has no part to play in qualitative styles of investigation (cf. Bogdan and Biklen, 1982). Nevertheless, recent studies (Silverman, 1984) have indicated that this is too sharp a division and that there are distinct advantages to be gained from the juxtaposition and integration of these two styles of researc...