Social Science Research as a Means of Scientific Enquiry

Social science is defined as the study of society or a population suggesting its inquiry refers to any systematic attempt to study, analyze and interpret data on a population. While it is often impractical to do research that targets all the members of a population, a simpler more desirable approach is to limit inquiry to a specific subset of the population. The subset, or sample as it is better known, is taken to be indicative of the general characteristics of the whole population.  The presumption is however only limited to representative samples, and on this basis, social science research offers a great tool for study and interrogation of scientific fact.

Social science research seeks to answer questions on a varied range of topics. From exploratory investigation, to descriptive and the explanatory, investigation is used to arrive at and prove general claims. Research generally follows one of two ways of reasoning. Inductive reasoning depends on experience, deriving theoretical implications to explain the knowledge. As a result, it is better suited for qualitative design analysis and for the most part makes use of secondary data sources. Deductive reasoning on the other hand involves the conceptualization of an idea, gathering empirical proof for the idea, and testing it to prove its validity. Deductive reasoning follows quantitative design analysis and employs primary data sources (Leedy & Ormrod, 2005). While the distinction between qualitative and quantitative design analysis methods does exist, it is worth noting that most inquiry involves at least some use of both.

Research methods employed in social science takes the form of inferential statistics. A representative sample is taken, and tests are carried out on it to prove a hypothesis. If sampling is done correctly, then the results yielded on its study can be taken to be those of the whole population, with a respectable deviation margin. Inferential statistics will therefore serve to provide an estimation of the degree of deviation as well as a test of the hypothesis (Leedy & Ormrod, 2005). Hypothesis testing primarily involves meta-analysis and the use of statistical analysis packages like SPSS and SAS. Meta-analysis will typically involve conducting a fairly intensive search for relevant material on the topic of enquiry, subsequent identification of what texts or studies to analyze and their eventual compilation into a common statistical index. Alternatively, statistical analysis packages may be used. The packages come with a handy list of features that can quickly do most statistical operations, and since they are computer based, computation and analysis is usually very fast.

After analysis and testing, the next logical step is the presentation of data. Presentation involves a compilation of the evidence used in the investigation. Since some projects may be complex use of figures and tables often proves convenient. The data is codified, arranged and separated to coincide with the problem and its constituent sub problems. Mapping problems to the data makes it easy to solve the research question. Once data has been presented, it is subject to interpretation. In this step, the findings are related to the inquiry or literary material already available on the research topic. Thereafter, a probe on whether the findings have any statistical or practical applications is done. A hallmark of a good investigation entails the identification of the limitations of the study (Bhattacherjee, 2012). It is therefore important that limitations of the study be included.

It is near impossible to carry out an investigation that is perfect. In social research and in particular when testing the hypothesis, it is possible to make errors. Since errors affect the validity of results, it becomes necessary to minimize them as much as possible. Two key ways that errors are prevented during hypothesis testing involves ensuring the validity and reliability of measures used in the study and the use of parametric statistics (Creswell, 2003). Some measures are more valid and reliable than others. It therefore follows that these should be preferred when making any measurements. Further, parametric statistics require relatively smaller samples sizes to deliver accurate results compared to non-parametric statistics. With this in mind, it becomes necessary to employ parametric statistical method whenever representative samples will be used to lessen the error margin.

To conclude, social science research can be a valid way to test for the validity of hypotheses and in experimental design. Research in this context is well thought out and will always give accurate results on the truth of a general claim. Social science research is identifiably on of the key ways to carry out such inquiries and its benefits in scientic inquiry cannot be overemphasized.

References

Bhattacherjee, A. (2012). SOCIAL SCIENCE RESEARCH: PRINCIPLES, METHODS, AND PRACTICES (2nd ed.). Tampa, Florida: USP Press.

Creswell, J. (2003). Research design. Thousand Oaks, Calif.: Sage Publications.

Leedy, P., & Ormrod, J. (2005). Practical research. Upper Saddle River, N.J.: Prentice Hall.