Quantitative data analysis has been an inevitable part of social science research. Like any other discipline, the theories are taught to students to give them an idea and generalisation of social facts and books, journals and other sources are used for the same. However, at times these sources are not adequate and research needs to be done in order to gain a deeper knowledge. That is when quantitative data analysis comes into picture.
There have been questions raised on the need to study quantitative data analysis with the emergence of importance of qualititative data (Bryman 1988a), but the former's importance has waned little. It helps that a larger proportion of empirical research that is conducted draws upon quantitative data.
The research design for any research study undertaken includes the method that needs to be applied for the purpose of collecting and analysing data.
Data collection can be done two ways- primary collection and secondary analysis of data. When the researcher collects data on their own for the sole purpose of the research that they are conducting, they are making use of primary data collection. The procedures used in these collection best suits the research problem at hand. The secondary analysis of data involves collecting data for a different research purpose but that is reused for the present research question.
Primary and Secondary Data
As explained before, data collected for the specific problem, primary data involves addition of new data to the existing store of knowledge surrounding the research area. When this material is used by other researchers, then this becomes secondary. Hox and Boeije (2005) maintain that the primary data can be used for:
- Description of contemporary and historical attribute,
- Comparative research or replication of original research,
- Reanalysis for the purpose of asking new questions which were not addressed originally,
- Research design and methodological advancement and lastly for 5. Teaching and learning
Secondary analysis utilises the existing data, collected for the purposes of a prior study, in order to pursue a research interest which is distinct from that of the original work (Heaton 2002).
Secondary data analysis is usually made of quantitative data where the information is made of researched objects whose characteristics have been coded in variables that can have a range of values. In fact, secondary analysis of quantitative data is common but the practice is not the same when it comes to qualitative data (Hinds, Vogel and Clarke-Steffen 1997).
Strengths and Weaknesses
Social science researchers undertaking research have a choice of opting to go for primary data- information that they need to collect by themselves or for secondary- searching for data that relates to the research problem in hand. There are distinct pros and cons of going for both. In this section, we will discuss what advantages or disadvantages the researcher faces when using any of these data collection methods.
One of the important advantages of going the way of primary data collection is that making use of the theoretical constructs, research design and the data collection strategy can be built with the research question in mind. This will ensure that the research study is coherent and the data collected is distinctly relevant to the problem at hand.
A disadvantage lies in the fact that primary data collection can be quite time consuming and expensive affair and considering the limitations of certain research study in terms of time and budget, primary data collection might not be a viable option for many researchers.
Another aspect of primary data collection is in its error inducing nature. Sampling errors made by inefficient field workers can skew up the research.
Some of the prominent data collection methods in primary data are: experiments, surveys like interviews, mail and web surveys. In the case of the experiment, the researcher is able to have a control on who participates in the research and the research situation being under the researcher's control means that there is strong control of design and procedure permitting causal interpretation of the results. Thus the ability to have some control may be a distinct advantage for primary data, however this can turn into a disadvantage too as one might say that the researcher's control has made the research 'artificial'. In an experimental laboratory, variables are easily manageable and there is no place for the 'circumstantial issues' that dominate in everyday life. While conducting surveys, the researcher is able to gather both subjective as well as objective characteristics of the population. If interview questions are carefully designed, evaluated and tested, surveys are a very method to obtain first-hand valid responses from respondents.
Effectively, this leads us to understanding of the advantage of secondary data analysis. Secondary data is far easier to collect and is less expensive and the access to relevant information is faster. The disadvantage lies in the fact that secondary data was collected originally for a different purpose and therefore might not be optimal for the research problem that is being considered.
Heaton points out another argument favouring the use of secondary data analysis stating that it can be used to generate new knowledge, new hypotheses supporting an existing theory and it also reduces the burden that is placed on respondents (for primary data collection) by removing the need to further recruit subjects thus allowing a wider use of data from rare and inaccessible respondents.
Not all social research problems can allow the usage of secondary analysis. It has been determined that it is more convenient for certain researchers, namely students and in some cases by researchers re-using their own data rather than by independent analysts. (Szabo and Strang 1997).
In their own right secondary data analysis is an effective tool in teaching as it helps in introducing students to a discipline and provides a supplement to the process of teaching (Sobal 1982).
But the cons behind using the same are numerous. The researcher will need to locate the source of data that is more relevant to the study and this can be time consuming as pointed out earlier. The researcher should also be able to retrieve the data, which at times can be difficult. Also, the data should be able to meet the quality requirements of the present research. Besides, the reliability of the secondary data is also a major function of the organisation that gathers, organises and publishes the data.
Another crippling factor that arises in the use of secondary analysis is that it differs from systematic reviews and the meta-analyses of qualitative studies that aim towards compiling and assessing the evidence relating to a common research concern or area of practice (Popay, Rogers and Williams 1998).
An issue that doesn't come in forefront when discussing secondary analysis is the principle of ethics behind using it. In using sensitive data, the researcher cannot assume informed consent. A professional judgement needs to be made about the usage of the secondary data and whether that violates any contract between the researchers and the original researcher (Hinds, Vogel and Clarke-Steffen 1997).
In conclusion, one may say that there are several advantages and disadvantages of secondary data analysis to collection of one's own primary data, and its usage is best suited to some research issues. But secondary data analysis is a valuable asset as they can act as a model for the collection of primary data. Suffice it to say that there might not be a need to choose between primary data and secondary data analysis at all, as the researcher can easily incorporate both in their research to gain a degree of balance between their strengths and weaknesses. What is most important is that both primary as well as secondary data should be accurate, reliable, appropriate, valid, precise and timely.
- Bryman, A. 1988), Quantity and Quality in Social Research, London: Routledge
- Heaton, L; Secondary analysis of qualitative data, 2003, in R. Miller and J. Brewer (eds.) The A-Z of Social Research, Sage, pp 285-288
- Hinds, P.S., Vogel, R.J., Clarke-Steffen, L. (1997) 'The possibilities and pitfalls of doing a secondary analysis of a qualitative data set', Qualitative Health Research, vol. 7(3): 408-24.
- Hox, J.J. and Boeije, H.R. (2005). Data collection, primary versus secondary. in K. Kempf-Leonard (Ed.). Encyclopedia of Social Measurement, pp. 593-599
- Popay, J., Rogers, A., Williams, G. (1998) 'Rationale and standards for the systematic review of qualitative literature in health services research', Qualitative Health Research, vol. 8 (3): pp. 329-40
- Sobal, J. 1982, The Role of Secondary Data Analysis in Teaching the Social Sciences, Library Trends, vol. 30, n3, p479-88.
- Szabo, V. and Strang, V.R. (1997) 'Secondary analysis of qualitative data', Advances in Nursing Science, vol. 20(2): 66-74.