Primary data collection

Above, we considered data arising from secondary sources and secondary analysis. Often, such data will be used in conjunction with other data collected first hand in order to achieve triangulation and increased validity. Primary research and analysis can be conducted in several ways by criminologists and victimologists. Here, we note a number of common methodological issues that relate to obtaining data from sub- jects first hand. We focus on surveys and samples, interviews, observations and ethnography.

Surveys and samples

One important method of collecting data from subjects is the social survey. Social surveys have been used extensively in criminological research, and crime/victim sur- veys typically use structured questions as a means of collecting data from individual respondents first hand. This can be done by interviewing them or by requesting that respondents fill in a self-completion questionnaire. The CSEW mentioned above is one such survey. Survey research lends itself to the collection of both quantitative and qualitative data. The cases surveyed in criminological research can include a wide range of units of analysis, including interactions or documents. Individuals and cat- egories of individuals are popular as primary and supplementary sources of data for criminological researchers. Whilst questions posed are typically structured, allowing the researcher to present the same stimuli and thereby collect the same kinds of data from a large number of people quickly, cheaply and with comparability of response, such questions in surveys run the risk of being too structured. Clearly, as you can already see, survey research involves complex issues (see Jo Deakin and Jon Spencer, Chapter 9).

It is very rare to collect data from the whole of the population in which a researcher

is interested: this is a very costly and time-consuming exercise. For this reason, social surveys are usually sample surveys. A sample survey is a form of research design which involves collecting data from, or about, a subset of the population with a view to mak- ing inferences from, and drawing conclusions about, that population (the term ‘census’ is generally used when all members of a population are included in a study). There are skills in selecting a sample which is representative of the wider population and several chapters in this volume refer to sampling issues, some in the context of gathering data from respondents first hand (see for example Chapter 2 by Peter Francis).

Several contributions in the chapters that follow discuss doing criminological research that involves gathering data first hand from respondents. Whether their

work has involved surveying sample populations, interviewing individual respond- ents and/or conducting focus groups, all have sought to avoid being overly structured, impersonal, and inflexible in their approach to gathering data first hand in order to demonstrate how these techniques can be used to produce valid, ethical, effective, rigorous and comprehensive data. Victim-oriented surveys inevitably deal with sensi- tive and emotive issues. Respondents are asked to reflect on personal and intimately harmful topics and experiences that they may not have disclosed previously, with strangers or via keying in data, albeit in a confidential manner. Hannah Bows in Chapter 4 and Vicky Heap and Jaime Waters in Chapter 5 illustrate how surveys can be employed when mixing methods. A persistent criticism of traditional crime sur- veys is that they tend to be confined to restricted ages of the population. Those under the age of 16 and those living in institutions as well as the homeless – all of whom might be deemed ‘vulnerable populations’ – are often excluded from national and supra-national/international surveys. From 2009 the CSEW has included a separate survey to record the experiences of young people aged 10–15. This interview is shorter than the adult one. Young people are selected to take part from the same households selected to take part in the adult survey. Permission from a parent or guardian is always obtained before an interview is conducted with anyone aged 10–15 (Francis, 2007). Both the sensitive topic and age-restriction critique are addressed head on in Chapter 9 by Jo Deakin and Jon Spencer, who discuss tackling difficult subjects and gathering sensitive data with vulnerable populations through large-scale national surveys.

Interviews

Interviews can be defined as a method of data collection, information or opinion gathering that specifically involves asking a series of questions. Typically, interviews represent a formal meeting or dialogue between people where personal and social interaction occurs (Davies, 2006). They are typically associated with qualitative social research and are often used alongside other methods. They can vary enor- mously in terms of the context or setting in which they are carried out, the purpose they serve as well as how they are structured and conducted. This means they are a flexible and adaptable tool and there are many different types of interview. Most commonly, interviews are conducted on a face-to-face basis and they can include one or more interviewers who are normally in control of the questions that are put to one or more interviewees or respondents. However, interviews can be informal, unstructured, naturalistic, or in-depth discussions in which the shape of the interview is largely determined by the individual respondent, through to very structured discus- sions according to a format with answers offered from a prescribed list in a questionnaire or ideal standardized interview schedule. An example of an interview with little interaction between the researcher and the researched is Computer Assisted Personal Interviewing (CAPI) where interviewers enter responses into a laptop computer, self-keying, to answer questions themselves. Since 1994 this mode

of interviewing has been used in the BCS/CSEW for more sensitive topics. The type, nature and range of interviews used within criminology are explored by various contributors to this book, including Pamela Davies in Chapter 12, Elizabeth Stanley in Chapter 14, Marie Segrave and Sanja Milivojevic in Chapter 15, Ross McGarry and Zoe Alker in Chapter 16 and Majid Yar in Chapter 18.

Sometimes interviews may be conducted by telephone, Skype or by way of elec- tronic communication such as e-mail. Interviews of this nature are popular for reasons of cost-effectiveness and the speed of data collection. Telephone interviews are routinely used for the conducting of opinion polls by market researchers. Political opinion polls are some of the most well-known types of interview conducted by this method.

As a means of collecting data first hand, interviews can be an invaluable source of information that generate valid, representative and reliable data. They enable you to follow up and probe responses, motives and feelings and in many of their forms, non-verbal communications, facial expressions and gestures, for example, can enrich the qualitative aspects of the data. However, assuming the use of the interview as the obvious method of choice for qualitative research can generate inappropriate or unmanageable data unfit for specific contexts and for specific purposes. In addition to this, there are skills to the practice of interviewing itself. Every aspect of the inter- view process can invite critique, for example over whether they are generating valid, sound and reliable data, and whether there is bias (including unconscious bias) surrounding the interviewer–respondent relationship.

Alternative types of interviews are associated with distinct advantages and disad- vantages. Unstructured interviews, where the respondent talks freely around a topic, can produce rich grounded data but can be very time-consuming to analyse and the potential for bias on behalf of the interviewer might be increased. The more guided or focused the interview, generally speaking, the less time-consuming and less prob- lematic is the analysis due to the more standardized nature of the responses. In opting for the latter form of interview, there is generally an increased likelihood that the researcher might not be asking the most significant questions.

Observations and ethnography

Participant observations and ethnography are among the most common methodo- logical traditions in criminological research (see Dolman and Francis, 2010). Observational research can probably best be described as the ‘hanging out’ school of research. Observations can be used in various criminal justice settings, including the prison, and might well be used in conjunction with other methods such as interviews. In the pilot stages of research it is desirable and often necessary to spend some time among the research populations and/or in the institutional setting, particularly if this happens to be a prison, before embarking on interviews or survey work (Martin, 2000). The ethnographic researcher will enter the field as soon as possible and is likely to undertake other tasks such as a literature review and conceptualization during and

on completion of fieldwork (Silverman, 2007). Participant observation and ethnographic methods, conversational and discourse analysis, documentary analysis, film and photog- raphy and life histories, can attract criticism. Often, they are seen as producing ‘soft’ data rather than ‘hard’ factual data (Hollands, 2000) and certainly ‘thick’, ‘rich’ and ‘intense’ are three strong words to describe the data produced from ethnographic research. Steve Hall and Simon Winlow (Winlow and Hall, 2009), the former a contributor to this cur- rent book, have long been involved in ethnographic research and readers are encouraged to see for themselves the intense meanings and understandings that are derived from such inquiry. See Chapter 17 by Steve Hall and Chapter 18 by Majid Yar.