Describing the Four Major Methods of Research - Essay Example

Describe Four Major Methods of Research Introduction: Several methods are available to conduct a research study, which are categorized by the strategies and techniques employed in the methods. Based on these strategies and techniques each of these research methods has their own merits and demerits. Hence for a researcher there arises the need to carefully evaluate the research method that is best suited to the subject of research and the questions that the research should answer, to ensure that the right research method is employed in the study. This paper attempts to evaluate four research methods consisting of the observational research method, the true experimental method, the quasi experimental method, and the correlation research method. The Observational Research Method: The Observational research method is an ethnographic research method that has had a long association with social research, as it provides the means to study individuals or groups of individuals in their normal environment and understand how they view things. The notion that observational research is a simple research method that merely requires looking, listening, learning, experiencing and noting it down may arise from the nomenclature of the stuffy method. However, this is far from true and the observational research method is one of the most personally demanding of all research methods and analytically difficult too. Based on the aims of the research, the observational research method may require the spending of a large amount of time in an unfamiliar environment, interacting and maintaining relationships with individuals with whom there is minimal affinity, taking notes of everyday ordinary events, and at the end of the field work spending a great deal if time analyzing the data collected. The observational method then looks like an onerous one and not attractive for research purposes. The fact however is that observational research is one of the most rewarding methods of research that provides the means to a fascinating insight into the social life and relationships of individuals and groups of individuals, with the added benefit of bridging gaps that may exist between the people’s understanding of alternative life styles and the prejudices, which are often the result of differences an diversity. (May, 1997, p. 138-155). The observational method may be use a group of participant observers or a single researcher playing the participant observer role. Through this process it is possible to observe the reflexive rationalization of conduct, which is the continual interpretation and application of new knowledge by individuals in their social environment. The ethnographic researcher becomes the instrument for data collection through their entry into the social environment of the individuals. This is accomplished through different means one of which is gathering information while moving closely among the individuals, where very often it literally means living among the people and observing their everyday lives. Alternatively it may also take the forms of study like in-depth interviews and life-histories, which provide the means to an understanding that is of greater depth than mere observation. It is quite likely that these two methods are often combined in the observational research method (May, 1997, p. 138-155). The True Experimental Method: The true experimental method is one of the experimental designs for the conduct of research. The essential purpose behind the use of an experimental design in research is to study causal links. In its simplest form an experimental design asses the impact of a given factor X on another factor Y. Moving into greater depth the experimental design might assess the magnitude or importance of the causal effect and its relation to other factors that give rise to a similar effect. An experimental research design from a social research perspective is restricted in the information that it provides through its narrow focus, but its main advantage is that it provides more definite answers to questions about causal links over other research methods. (Hakim, 2000, pp. 127-144). Sample selection bias is one of the major problems with an experimental research method and the true experiment method attempts to overcome this problem by randomization and controlled application of the given factor X that is being studied. Experimental control consists of deciding which groups of individuals or organizations are to be included in the study. Such a rationale arises from the grounds that those who voluntarily agree to participate are identified as different from those who do not wish to participate and as such the impact of the experimental factor will be confounded with the impact of the selection effects. Randomization consists of allocating the volunteer subjects or other units under investigation to an experimental group, which is exposed to the information, experience or event that is under study) or a control group, which is not exposed to such an experience. The entire allocation process is on a random basis. The rationale behind randomization is that the two groups thus selected would be exactly similar to all aspects of the study. The two groups are then evaluated before and after the experimental treatment at the same periods of time and comparisons of the data collected in this manner made to provide the results (Hakim, 2000, pp. 127-144). From the design of the method, it becomes clear that true experimental methods are unsuited for many topics addressed by social research. Individuals cannot be randomly be assigned to undergo experiences like unemployment, imprisonment, marriage, war, and migration to study these topics of social research interest. Furthermore professional and ethical reasons prevent exposing individuals to stressful treatment just for the purpose of research, as that would tantamount to treating them as animals used in such research designs. Volunteering makes up the basis for research subjects in a true research method and it is not always easy to find the required number and type of volunteers, making true experimental research to be conducted within these constraints and limiting the generalization of its findings. However, true experimental research has been found to be advantageous in studies relating to beneficial effects like increased income or improved quality of education. Examples of true experimental research carried out in real life settings are not easy to find and are mostly found in policy research, where large samples are employed and extrapolated to provide a national perspective. In the 1970s the governments of Canada and the U.S. funded guaranteed annual maintenance experiments. The largest of these experiments consisted of the Seattle/Denver Income Maintenance Experiment (SIME and DIME), where 4.800 families were involved and cost approximately $70 million. It is this huge setting and cost that is an impediment to true experimental research in spite of its capacity to provide more definitive answers to causal links than other research methods (Hakim, 2000, pp. 127-144). However, the relative ease with which the findings the experimental research method can be reported and supported through the use of statistical analysis, which lead directly on to the conclusion are factors that weigh heavily in favor of the use of the experimental research method (Long, pp. 97, 1980). Quasi-Experimental Research Method: The limitations of the true experimental method in its use of social experiments have necessitated the use other types of research like longitudinal studies, case studies, surveys and research analysis to study causal relationships. These types of research for the study of causal links are called quasi-experimental research methods. These methods are less demanding than the true experimental research method, but suffer from limitations in terms of the absence of control groups and assignment to exposure of the factors being studied on the basis of self-selection or administrator selection. In a bid to overcome the lack of randomization a quasi-experimental method uses the comparison of pretest results between the experimental and control groups to indicate the equivalency between the two groups In addition the quasi-experimental design employs additional tests like the strict application of rigorous logic to ascertain the validity of the results and to remove other possible explanations for the results obtained. (Leedy, 1997, pp. 232-233). One of the strongest quasi-experimental research methods is the regression discontinuity design, which can be employed when treatments are assigned on the basis of a quantitative measure, normally a measure of need or merit. The example of freshmen entering some universities with a verbal scholastic aptitude test of a specified value of say 400 being assigned to take a remedial English course, while other students with higher scores are assigned directly to regular English class helps to explain this concept. The outcome of interest here is the performance of the students on a test of writing skill. The essential feature of this example is that the participants are assigned to the control conditions purely on the basis of whether they exceed a particular cut off point or not on the quantitative assignment variable and that an outcome hypothesized to be affected by the treatment is evaluated subsequent to the treatment. Such a design removes the disadvantage of randomization, through the availability of potentially beneficial treatments to the neediest participants. (West, Biesanz & Pitts, 2000, pp. 54-62). Such a research method provides an excellent approach when participants are assigned to conditions based on a quantitative measure. The regression discontinuity design provides high quality of causal inference, by removing most of the threats to internal validity and has a good measure of external validity, when its assumptions are met. However, on the negative side it has a much lower power than a true experiment method (West, Biesanz & Pitts, 2000, pp. 54-62). Correlation Research Method: The correlation research model attempts to evaluate the cause-effect relationship between two sets of data. In essence it is the comparison of two sets of variables. For example the variables of gender, age or employment status can each be considered as variables that have the capacity to influence or affect the dependent variable of attitude. There is no logic in saying that the causal relationship is the other way round in that the attitude of an individual has an influence or affects the gender, age or employment status of the individual. Such a research method cannot in the true sense demonstrate causal link in that gender, age or employment status causes attitude in an individual. The only influence or affect attitude of the individual. Thus there is an interrelation between the two variables, which also could be affected by other factors. In addition there is the possibility of the two variables affecting each other to a certain extent. The correlation research method studies this interrelation or correlation rather than causation. Thus the correlation research method assists in making sense of different sets of data that have a relationship. However, it still requires experiments with rigorous control over the variables that enable proper conclusions to be drawn about causality (Hall & Hall, 1996, pp. 149-152). Multiple Research Methods: The use of multiple research methods in research studies emanates from the concept that qualitative research and qualitative research are complementary rather than competitive. Employing multiple research methods is also called triangulation, which may be considered as combining research methodologies to study the same phenomenon. Triangulation is grounded in geometry, where multiple reference points are used to provide the exact location of an object. In a similar manner triangulation in research collects different kinds of data on the same phenomenon using different methodologies to improve the accuracy of the conclusions arrived at. Triangulation provides the means to cross validation, when two or more distinct research methods are found to be congruent and produce comparable data. For example in organizational research it may be required to find the effectiveness of a leader. The effectiveness of a leader may be studied through interviewing the leader, observing the behavior of the leader or evaluating the performance of the leader from records. These are multiple and independent measures of the effectiveness of the leader, which could be combined to provide a more definitive portrayal of the leadership phenomenon. Thus the advantage of multiple research methods or triangulation lies in it being more accurate than any single research method. The disadvantage is that there is no definite prescription for how triangulation is to be actually performed. Study of research methods provide the strengths of benefits of the different research methods or what may be more appropriate or preferred for a particular study, but not the means combine the different research methodologies effectively to provide an accurate conclusion (Jick, 1979, pp. 602-611). Literary References Hakim, C. (2000). RESEARCH DESIGN: successful designs for social and economic research. Second Edition. Routledge: London. Hall, D & Hall, I. Practical Social Research. MACMILLAN PRESS LTD., Hampshire. Jick, D. Todd. (1979). Mixing Qualitative and Quantitative Methods: Triangulation in Action. Administrative Science Quarterly, 24 (4), 602-611. Leedy, P.D. (1997). Practical Research: Planning and design. Sixth Edition. Prentice-Hall: New Jersey, Long, B. H. (1980). Experimental Research. In CHANGING APPROACHES TO STUDYING ADULT EDUCATION. Huey B. Long, Roger Hiemstra and Associates. Jossey-Bass: California, pp. 78-99. May, Tim. (1997). SOCIAL RESEARCH METHODS: ISSUES, METHODS & PROCESS. Second Edition. Open University Press: Bristol. West, G. S., Biesanz, C. J., & Pitts, C. S. (2000). “Causal Inference and Generalization in Field Settings: Experimental and Quasi-Experimental Design”. In Handbook of Research: Methods in Social and Personality Psychology. (Eds.) Harry T. Reis & Charles M. Judd. Cambridge University Press. Cambridge, pp. 49-74. Read More