 |
Box 9. 2 hard-to-reach groups. Survey research. Criticisms of surveys
|
|
|
|
BOX 9. 2 HARD-TO-REACH GROUPS
A hard-to-reach group can be any group which is difficult to access for any reason, such as:
· physical inaccessibility (e. g. older frail people)
· language (e. g. first-generation immigrants to the UK)
· cultural perceptions and traditions (e. g. disadvantaged young people)
· social expectations (e. g. children and young people who are often not considered an appropriate consultee and whom often also do not consider themselves as likely to be taken seriously). (Home Office crime reduction website)
Hard-to-reach groups could include:
· homeless people
· children and young people
· drug users
· gay men, lesbian women, transsexual and transgendered people
· minority ethnic communities (particular sections of)
· victims of domestic abuse
· older people (especially older frail people, and isolated older people)
· travelers
· asylum seekers
· people with learning difficulties
· people with disabilities
· people with mental health problems
· faith communities
· people who travel or commute into the area
· small businesses
· rural communities
· tourists.
Source: www. crimereduction. homeoffice. gov. uk/toolkits/p03110701. htm referenced in Davies et al. (2007)
SURVEY RESEARCH
Survey research involves the systematic collection of large quantities of data, by means of questionnaires and/or interviews, from a broad sample of a target population. A range of complex issues are involved, including sampling methods,
developing meaningful research instruments ( questionnaires, interviews), admin- istering the survey, maximizing response rates, and a variety of ethical considerations (protection of respondents, confidentiality, etc. ) (see, for example, Bachman and Schutt, 2008; Fowler, 2009, 2014; Nardi, 2014; Seale, 2004). We cannot cover all of these issues in this chapter (suggestions for further reading are offered at the end of this chapter); instead, we provide an overview of survey research, signpost key areas for further reading, and focus more closely on the use of surveys in the area of sensitive topics and with hard-to-reach groups (see Box 9. 2).
Survey questionnaires are an important tool of the social scientist in collecting, categorizing and quantifying data. They allow the researcher to frame a research problem using a structured set of questions. Surveys are most commonly used to answer questions about the extent of beliefs or experience and can be used with a whole population or a sample of a population to provide data that can be general- ized across a larger group. For this purpose, the sample or cross-section must accurately reflect the complexity and heterogeneity of the population to be studied. A cross-section of the prison population, for example, may be surveyed to gain information that can be applied to the whole of the prison population: a prison- based sample must reflect the characteristics of the whole prison population as closely as possible, including gender, black and minority ethnic populations, those serving different sentence lengths for different types of crime, and so on. A prison- based survey must also take account of differing levels of literacy (a self-completion questionnaire would not be suitable for those who cannot read and write) and should be translated where necessary. An example of such a survey is Measuring the Quality of Prison Life (MQPL) which surveys a sample of prisoners in each prison in relation to a number of key elements of prison life, including trust, respect, fair- ness, order and safety (see Ross et al., 2008).
|
|
|
The aim of survey research may be to generate statistics to describe the character-
istics of a population – descriptive statistics (e. g. the extent of physical violence suffered by children); or to explore the correlation between two variables – analytic statistics (e. g. the relationship between the extent of physical violence suffered and the level of fear of violence experienced by children). The statistics generated are often used to inform policy and decision making.
CRITICISMS OF SURVEYS
The disadvantages of a quantitative survey methodology largely centre around five claims:
1. That the complexities of social data cannot reasonably be measured or recorded using an intrinsically positivistic method (Cicourel, 1964)
2. That the survey method assumes respondents understand and interpret the world around them in the same way, as if one were measuring a natural, unthinking phenomenon (Schutz, 1962)
3. That the meanings and definitions people assign to their experiences are ignored through the use of a structured method (Cicourel, 1982; Oakley, 1981), creating a ‘static’ image of social experience (Blumer, 1956)
4. That they present an obstacle to open discussion and prevent flexibility and spon- taneity (Fox Keller, 1980; Oakley, 1981; Reinharz, 1979)
5. That where surveys involve interviews, the influence of the interviewer and the location of the interview significantly distort responses (Phillips, 1971).
Marsh (1982) counters these claims by arguing that, when surveys are well designed and properly administered, they make a key contribution to our understanding of the social world. She suggests that it is a particular style of survey research, known as smash ‘n’ grab, which has resulted in the bad press associated with this method. For Marsh, smash ‘n’ grab refers to the process whereby researchers design research instruments, undertake fieldwork and analyse data within a very short time frame, literally ‘smashing’ into the field, ‘grabbing’ data and exiting quickly. The result is a hastily developed, carelessly designed and poorly implemented survey that is well deserving of criticism.
The questionnaire is the most heavily criticized element of survey research. Questionnaires conducted in interview form are accused of failing to acknowledge the respondents’ or the researchers’ experiences. It is argued that such an approach objectifies the respondent as data, while their subjective experiences and personal meanings become lost in the process (Oakley, 1981). Self-completion questionnaires that do not require direct contact between the researcher and the respondents are viewed as generating detached, stilted data and preventing ‘an interactive process’ (Kennedy Bergen, 1993: 203).
|
|
|
However, problems with questionnaires, and the survey method more generally, are not insurmountable as Kelly (1990) demonstrates with the use of question- naires in her research on child sexual assault. With a background in qualitative research, Kelly’s transition towards quantitative research came as a result of experimenting with different approaches suitable for answering particularly sensi- tive research questions. She recognized the significance of collecting quantitative data from large groups of respondents to answer questions about the extent of personal experience, and began to explore the survey method. However, for a survey to deal with sensitive issues ethically, it must allow for subjective defini- tions and experience, and create space for views to be expressed (Kelly, 1990). Since most surveys work on pre-defined categories and tick-box style answers, Kelly needed to rethink the traditional survey style to fit her sensitive topic. Kelly’s questionnaire, in addressing these problems, allowed respondents to define their own experiences and provided support for participants as required, thereby minimizing distress.
Perhaps most importantly, the research team spent several days, prior to adminis- tering the questionnaire, working in the participating institutions to facilitate links and develop relationships. This type of participatory approach to research allows researchers to begin to understand the research environment and become known to potential respondents. Through participation, researchers can gain the respondents’ trust, provide appropriate support and have an active role in facilitating change. Even when self-completion questionnaires are the chosen method, the process doesn’t have to be detached and stilted. Engaging with the participants in the research environment before and after collecting the data can encourage an interac- tive process that values the meanings people assign to their experiences and provides support.
The value of the survey method lies in its ability to gather large amounts of data to reveal the extent of a phenomenon. In so doing, it does not have to adhere to the stilted, bullish image it has so often been associated with. For the most part, the charges levelled at the survey are failings of the smash ‘n’ grab method; they are not failings of all surveys. Perhaps questionnaires will always have elements of inflexibil- ity in their stable agenda and pre-defined categories, but, in sensitive research, these elements have probably been derived from a qualitative pilot study conducted to explore the area.
|
|
|
