Inductive and Deductive Data, Piloting - Essay Example

?Methodology Inductive and deductive data Saunders et al. (2007) write that inductive approaches explore the data to develop theories that are then tested empirically. Inductive data is used when researchers use their observations and empirical data to construct new theories (Cargan 2007). Inductive data leads to the development of grounded theory, but the limitations of inductive approaches in research cannot be ignored. First, inductive approaches often lead to overgeneralization (Cargan 2007). Another problem is oversimplification (Cargan 2007). Finally, inductive approaches do not always allow researchers to capture the complexity of various social phenomena (Cargan 2007). This is why, when researchers cannot overcome the major limitations of inductive data, they begin their study with a theory, followed by a hypothesis and research (Saunders et al., 2007). In other words, they apply for deductive research approaches. Deductive data is usually derived from the study of earlier experiments, empirical and theoretical literature. In other words, deductive approaches rely on the previous studies and lead to the creation of generalizations and assumptions from earlier formulations (Cargan 2007). The use of deductive data in research is very common. According to Saunders et al (2007), deductive data allows developing conceptual frameworks from theories considered in the literature. These frameworks are then tested using the data (Saunders et al. 2007). Researchers who use deductive data develop new theories, based on the earlier studies in the same field. In deductive approaches, researchers usually begin with the analysis of literature. At times, meta-analysis designs are used to systematize earlier research findings. All these steps are integral to deductive approaches in research. With deductive data, researchers find it easier to make generalizations and develop recommendations for future research. It should be noted, that deductive approaches differ greatly from the inductive ones. Inductive models do not enable researchers to evaluate earlier findings. Researchers using deductive approaches in research can use earlier theories and findings to add to the knowledge derived from their own study (Cargan 2007). Researchers can use earlier observations to explain their findings. Mixed methods research Mixed methods have already become a defining feature of contemporary research. Mixed studies are often valued for their strength and reliability. Basically, mixed methods designs involve the use of both quantitative and qualitative features (Denscombe 2010). The concept of mixed methods can be traced to the end of the 1950s (Creswell 2003). Mixed methods exemplify a logical and promising response to the limitations inherent in all research methods. With the help of mixed methods, researchers can “neutralize or cancel the biases of other methods” (Creswell 2003, p.15). Triangulation is a popular method of research, which is used to ensure better validity and reliability of the study findings (Creswell 2003). For example, the results obtained from one method can be used to inform another method of research, or one method can be nested within another research strategy to ensure effective analysis of the different levels of data (Creswell 2003). Researchers who use mixed methods in their studies can choose from three research procedures. First, in sequential mixed method approaches, researchers use the data obtained from one research method to improve the quality of the results obtained from another method (Creswell 2003). The data obtained from one method can also be used to improve the second method used a mixed study design. Second, researchers can also use concurrent procedures, when qualitative and quantitative data are used simultaneously to study one research question (Creswell 2003). Finally, there are also transformative procedures when a particular theoretical framework is used to interpret both quantitative and qualitative study findings (Creswell 2003). The choice of particular procedures and methods to be included in mixed designs depends on numerous factors, from the goal of the study and its research question, to the availability of resources needed to obtain qualitative and quantitative results. The benefits of mixed methods and triangulation, in particular, are obvious: Denscombe (2010) writes that triangulation enables researchers to view the problem from more than one perspective. As a result, through triangulation, the researcher can get a better view of the problem (Denscombe 2010). In most cases, triangulation involves a comparison of the qualitative data to quantitative findings. The instruments to be used in mixed research will depend on the method adopted for each particular study, including questionnaires, surveys, and interviews (Denscombe 2010). Piloting Piloting is an extremely popular instrument of doing research. “Piloting, is the process whereby you try out the research techniques and methods which you have in mind, see how well they work in practice, and, if necessary, modify your plans accordingly” (Blaxter, Hughes & Tight 2006, p.137). Simply stated, piloting is a trial stage of the research process, when the researcher tests the methods and ideas that are to be used in the actual study. Piloting offers considerable methodological and research advantages. While planning a new research project, researchers should consider carefully research questions, target sample, sampling procedures, variables and covariates, as well as possible circumstances and limitations. Piloting, or pilot studies, is the most useful whenever researchers need to clarify certain aspects of the proposed study (Light, Singer & Willett 1990). If at least one facet of the actual study is unclear, piloting can help to improve the quality and validity of the study design. However, piloting should not be used while trying specific hypotheses and study questions: piloting is not intended to explore primary data. Piloting by itself relies on inductive logic, and it is neither confirmatory nor deductive (Light, Singer & Willett 1990). The primary goal of piloting is to check the appropriateness of the methods and designs, as well as the way they suit the research questions and context in which the study takes place. Piloting does not allow for generalizing results and can only serve as a good way to check the quality of various research methods. The central advantage of a pilot study is in that it allows the researcher to expand the scope of the actual study, making it deeper and more informative (Light, Singer & Willett 1990). Piloting allows collective both qualitative and quantitative data and provides a unique insight into the nature and potential limitations of the actual study. With a pilot study, researchers can evaluate the main benefits and limitations of the proposed research design and use the information they get from a pilot study to improve their methods and research frameworks (Light, Singer & Willett 1990). One of the main benefits of piloting is that is it not easy to implement. In piloting, no big samples are needed (Light, Singer & Willett 1990). However, even with small samples and simple research procedures, researchers can predict the problems and avoid limitations awaiting them during the actual study. Another benefit is in that the instrument the researcher uses during piloting should not be as huge and exhaustive as the one to be used during the actual study (Light, Singer & Willett 1990). Simply put, if the researcher has a questionnaire containing 150 questions, it is enough to choose the most important questions and use them in a pilot questionnaire. In any pilot study, not the size of the sample but its representatives truly matters. As a result, researchers can anticipate and avoid serious problems that can easily emerge during the actual study. Piloting is beneficial for researchers, especially when new research instruments need to be developed. Piloting is a unique opportunity provided to the researchers who seek to test new methods of data collection and analysis. With the help of pilot studies, researchers can avoid the risks of ambiguity while using their research instruments (Light, Singer & Willett 1990). If response scales are used, the only way to test their validity is during a pilot study (Light, Singer & Willett 1990). Unfortunately, even the most professional pilot study cannot eliminate all questions, which inevitably result from the researchers’ striving to achieve the best quality of the study findings. References Blaxter, L, Hughes, C & Tight, M 2006, How to research, McGraw-Hill. Cargan, L 2007, Doing social research, Rowman & Littlefield. Creswell, JW 2003, Research design: Qualitative, quantitative, and mixed method approaches, SAGE. Denscombe, M (2010) The good research guide : for small-scale social research projects. 4th ed. Maidenhead, McGraw-Hill Open University Press. Light, RJ, Singer, JD & Willett, JB 1990, By design: Planning research on higher education, Harvard University Press. Saunders. M, Thornhil.A, Lewis.P (2009) Research Methods for Business Students, 5th Ed. UK. Financial Times Press. Read More