Strengths and Limitations of Quantitative Research Designs - Term Paper Example

Strengths and Limitations of Quantitative Research Designs Experimental research proposes a control of any factor which can determine the events analyzed and then quantifies the effects when the factors are manipulated (Walliman, 2011). The experimental design depends on the event analyzed, the variables involved, and the level of accuracy and reliability aimed for each specific research. This method requires manipulation of an independent variable and then quantifies the effect on a dependent variable, namely cause and effect. True experimental design is considered the most reliable of the quantitative research designs because it is the most closest to explain the effect of the independent variables on the dependent variable. Due to randomization and assignment and use of a control (“placebo”) group, this design has no internal validity threats. Another advantage would be that the results obtained can be relied on to make generalizations and be used in other studies. As a limitation would be possible threats to external validity because of the interactive effects of the treatment with selection, instrumentation, history, or maturation. Finally, it can be said that the random assignment gives the degree of success of this kind of experiment. The quasi-experimental design is less reliable than the true experimental design. A limitation of this design will be the lack of randomization, which is the assignment of treatments to subjects in order to reduce bias (Valiela, 2001). This inconvenience can be corrected by matching the control and the experimental group, reducing the possible initial difference between the two groups. Compared with a true experimental design, it has inferior internal validity. However, it has the benefit that can be more easily implemented than the true experimental design. A cross-sectional design is a non-experimental design which is characterized by the fact that the data is gathered at one time. In this situation, the researcher cannot figure out with certainty if the cause precedes the effect. This limitation can be corrected in those studies in which alternative interpretations are plausible, and validly measured. These difficulties with cross-sectional design bring some doubt over their potential to support strong causal inferences (Beins & McCarthy, 2012). Design for the Research Paper The topic of the research is the impact of foreign direct investment (FDI) on the economic growth of developing countries. The framework of the study includes an independent variable i.e. FDI and a dependent variable i.e. economic growth. The economic growth would be measured through cash inflow, employment generation, technological diffusion and management and labor skills. The intervening variables which can directly influence the relationship between FDI and economic growth are government policies and the infrastructure development Research designs can be affected by bias and design threats, determining a lower internal validity and external validity of study results as was mentioned at the beginning of this paper. Considering this, it is very important to choose the most appropriate design for this research. Starting from previous studies related to this topic, but also considering the fact that this research should bring something new in this area of study, the following ideas were taken into account when considering the layout of this specific research. Casi and Resmini (2011) applied a quasi-experimental design and used a propensity score matching (PSM). This method uses an index (i.e. propensity score) composed of pre-sample characteristics which makes the matching of the two groups feasible and reduces the bias behind sample selection (Becker & Ichino, 2002). Another study made by Veloso (2008) considered also a quasi-experimental design with a few differences. This research used two approaches i.e. static group and pretest-posttest in order to accept or reject the null hypothesis, which was South Africa’s FDI inflows are the same as other emerging economies with the same level of development, but without ownership equity policies. Mello (1999) used time series and panel data to estimate the effect of FDI on total capital, and output and total factor productivity on economic growth. In this study, there were constructed a cross-country and growth equations in order to test the hypothesis. For time-series growth equations, Jones (1995) admitted the uncertainty of predicting permanent changes in factor accumulation which can lead to permanent changes in output, assuming a linearity hypothesis. For this research it will be implemented a quasi-experimental design for the following considerations. Firstly, previous research considered a similar approach i.e. a quasi-experimental design. Secondly, quasi-experimental design provides the possibility to make inference on a variety of matters related to the connection between FDI and economic growth, compared with more traditional approaches (Zhang, 2001). Moreover, it could be a selection bias because in order to estimate the impact of FDI on developing countries growth rates, the difference between countries’ growth rates with and without FDI should be available (Islam, 2003). It is not feasible to compute the growth rates for the same country at the same time and also to take the average growth rate of countries without FDI as an approximation would not be correct due to differences between countries. The design for this topic will be similar with the approach of Casi and Resmini (2011), which applied the propensity score matching to reduce the bias from the selection of the two groups, the experimental and the control group. The hypotheses of this research study will be analyzed for both experimental and control group: HA: FDI results in improving the economic growth of the developing countries. HO: FDI does not directly influence the economic growth of the developing countries. This design will require manipulation of the independent variable – FDI and then quantifies the effects on a dependent variable – economic growth. Although, the results of this research would not be generalized for other studies by choosing a quasi-experimental design, it would be another step for the argumentation of bringing FDI inflows in developing countries. Why not other Research Designs? A true experimental design would have been difficult to implement for this topic and probably not the most appropriate design. In order to apply such a design, it should have data for all developing countries under analysis, and there are differences between these countries, although there was considered developing or emerging. Also, in a true experimental design, the impact of intervening variables (in this study – the government policies and the infrastructure development) can make the randomization not such reliable. Moreover, in the field of finance, when studying macroeconomic variables are almost never used this type of design, but it is preferred a more easy and efficient design such as a quasi-experimental design. A traditional quasi - experimental design pretest-posttest would not have been appropriate for reasons specified above i.e. it is not feasible to compute the growth rates for the same country at the same time. Also, a simple posttest design would not provide the possibility to compare the data so it would not be the most reliable research. Finally, the results of a cross-sectional design would not be reliable because it does not have the certainty that FDI precedes the economic growth and it cannot be corrected by considering alternative interpretations, validly measured. Also, this design does not offer the chance to reduce the effect of intervening variables and it is not appropriate for the hypotheses, this study proposes to test. References Becker, S.O., & Ichino, A. (2002). Estimation of average treatment effects based on propensity scores. The Stata Journal, 2 (4), 358-377. Beins, B., & McCarthy, M. (2012). Research Methods and Statistics. Boston: Pearson Education, Inc. Casi, L., & Resmini, L. (2011). Foreign Direct Investment and Growth: a Regional European Perspective. (Unpublished paper). University of Valle d’Aosta, Grand Chemin. Islam, N. (2003). What Have we Learnt from the Convergence Debate?. Journal of Economic Surveys, 17, 309-362. Jones, C.I. (1995). Time Series Tests of Endogenous Growth Models. Quarterly Journal of Economics, 1110, 1127-1170. Mello, L. (1999). Foreign direct investment – led growth: evidence from time series and panel data. Oxford Economic Papers, 51, 133-151. Valiela, I. (2001). Doing Science. New York: Oxford University Press, Inc. Veloso, V. (2008). The Impact of Broad-Based Black Economic Empowerment on Foreign Direct Investment Inflows into South Africa. (Unpublished dissertation). University of Pretoria, Pretoria. Walliman, N. (2011). Research Methods: The Basics. New York: Taylor & Francis Group. Zhang, K. (2001). Does Foreign Direct Investment Promote Economic Growth? Evidence from East-Asia and Latin America. Contemporary Economic Policy, 19, 175-185. Read More