Compare the different approaches and data collection instruments used in social science research to explore issues in relation to children and young people.
Social science involves different specialist fields that use a variety of research approaches that are not exclusive to each area of study. An important decision for the researcher is to select the appropriate and effective research methodology to explore their area of interest. All adopt an empirical approach to creating knowledge, which may provide measurable benefits for children and young people. Empirical research should be a systematic investigation of experience??¦sceptical and ethical (Robson, 2002, cited in Fraser, 2011, p17. It may expand knowledge but settles few fundamental arguments (Fraser 2011).
Through examples of methodology, it is possible to identify issues that can cause contention and possibly confound results. Although mainly theoretical this essay will draw on research examples, from Bowlby, Takei (2001), Sutton et al (1999), Punch (2001) and Aldgate & Bradley (1999) and examine arguments from Punch (2011), Frazer (2011), Glaser & Strauss (1967), Creswell, (1994) and Fontana & Frey, (1994).
Different paradigms emerged in relation to research with children and young people. Firstly the scientific paradigm, encompassing positivism and verification; this was based on natural/noble science methodology, which led to issues regarding the subjective nature of human subjects. The structuralist paradigm favoured by Durkheim and Piaget suggests an underlying objective structure, which can be discovered through empirical research. This still involved experimental method, which could be tested for reliability and validity, through repetition. The Interactionist paradigm moved away from scientific ???facts??™ and tried to establish attribution of meanings by individuals. Advocates of this include Mead and Goffman. The Social constructionist/post-structuralist paradigm proposes that understanding of the world is mediated and constructed through discourse. These paradigms do not travel a chronological continuum, but overlap and persist; however there will often be a prevailing paradigm at any given time.
Research is undertaken as theory verification or theory generation. Wolcott (1992, cited in Punch, 2011) describes this as theory first or theory after. Verification, starts with a theory from which a hypothesis is derived; a design through which this hypothesis may be tested is produced. Often quantitative analysis is used for here. Generation aims to deduced a theory from data collected during research. Qualitative research is often used. Grounded theory is an explicit theory generation strategy; it was developed as a reaction to the repeated use of theory verification (Punch, 2011). Continued use of verification theory led to lack of investigation of new problems (Glaser & Strauss, 1967, cited in Punch, 2011) and challenges to prevailing paradigms Whether a research design is for generation or verification will impact on all other design components.
Before the 1960s quantitative methods were dominant in empirical research, however this was challenged by the growth in qualitative method. Recently there has been less of a division between the two schools with many researches adopting a pragmatic approach combined methods of design and analysis. The emergence of the pragmatic approach focuses on what the research is trying to establish, fitting the method to the question; although functioning within the prevailing paradigm. Questions can only be asked and answered that are within the possibility of that particular time. In this way all research is contextual. For example Bowlby??™s Attachment Theory has been criticised by feminist, arguing that it is sexist, and in essence ???forced??™ mothers to stay at home with infants. This research was carried out within the context of post war Britain, where stay at home mothers had been the norm before 1939, and not only did this research stress the importance of mothers but was undertaken at a time when the country as a whole was arguably trying to return to pre-war values.
Quantitative strategies may use correlation and include comparison; this strategy can be experimental, quasi-experimental or non-experimental/correlational. Qualitative strategies include, case study, ethnography and/or grounded theory and are generally more naturalistic. Multi-methods can be applied within a research design, e.g. employing qualitative and quantitative methods/analysis. These strategies imply the amount of manipulation or intervention by the researcher moving from high levels of manipulation in experimental strategy to minimal manipulation in naturalistic strategy. Qualitative studies are usually non-interventionist (Punch, 2011). Research design encompasses planning and executing of a research project, and interpretation of data/results to establish reliability and validity. This involves constraining to the greatest extent possible confounding variables, which could lead to alternative interpretation of results.
The initial process in research design is to establish an area of interest, followed by question generation, but the fit between question and method needs to be as good as possible (Punch, 2011). The method used should enable the researcher to collect adequate data to analyse, satisfying the objective of the study. The what may predict the how or as discussed by Punch (2011) content precedes method (p.20). Consideration of goodness of fit goes beyond method and question, it is informed by paradigm, conceptual framework, study design, and perspective. Perspective will influence the focus of the research question or hypothesis. The wording of a research question can be indicative of the methods and theoretical framework. Phrases such as factors that effect tend to imply a quantitative approach. The qualitative approach is more concerned with discover or explore (Punch, 2011). Grounded theory may also seek to discover and question the understanding of social process; ethnography may seek to understand, focusing on symbolic aspects of behaviour; case study tends to explore and phenomenology looks to describe experiences (Creswell, 1994, cited in Punch, 2011).
Another important feature of research is the decision of longitudinal design, which allows for variation over time to be explored or cross-sectional design allowing for transitions to be seen. The choice of study design may be constrained by resources. This aspect may be involved in sample selection and size, which again involves many decisions, some which may be beyond the researchers control. In the case of Takei??™s (2001, cited in Lewis et al, 2011) research into attainment of ???first signs??™; the number of Japanese infants born to deaf parents limited his sample. A small sample may lead to results not being societally representative. However larger samples can lead to multiple covariates. There are methods employable for reduction in the effect of covariates dependent on research design. Informed consent must also be gained from various parties, dependant on design. This may include, schools, parents and children and here there is an issue with a child??™s ability to provide informed consent.
Interview is the main source of data collection for many researchers. The three main types of interview are; structured, semi-structured and unstructured (Fontana & Frey, 1994), rather than being 3 separate entities they may be considered as a continuum. Structured interviews are tightly focused with a pre-designed questionnaire and possibly simplistic responses, which are categorized, there is little manoeuvrability and the interviewers role is often neutral. They stress rational rather than emotional responses (Fontana & Frey, 1994, cited in Punch, 2011). Pickett (2002, cited in Lewis et al 2011) utilized this method gathering data regarding risk taking and injury in young people. This type of data allows for statistical analysis and can be manipulated to alleviate confounding variables. Raw data may be gathered through interview notes or self-reporting as in the case of Sutton et al (1999, cited in Lewis et al 2011); however self reporting can lead to subjects attempting to promote a certain image in their answers.