There are several types of quantitative research designs, but the most popular ones according to extant literature include descriptive, correlational, cross-sectional, longitudinal studies, causal-comparative, quasi-experimental and experimental (Hoe & Hoare, 2012; Welford, Murphy & Casey, 2012). The present paper illuminates some findings on the last two quantitative designs – quasi-experimental and experimental.
Research designs characteristically vary in the context of the researcher’s interference, with some designs such as correlational field studies registering minimal interference while others such as experimental typified by interference, manipulation or control over the independent variable. In quasi-experimental research designs, also described as nonrandomized, pre-post intervention studies, the researcher has the opportunity to evaluate the effects of interventions by manipulating the independent variable; however, quasi-experimental studies do not occur in completely controlled environments as there are likely to be other changes occurring that are not being manipulated explicitly for the experiment, which may be causing the effect (Creswell, 2003). This therefore implies that although manipulation of the independent variable may occur in quasi-experiments, the level and scope of control is generally weaker than would be the case in a true experiment (Ellis & Levy, 2009).
As one example of a quasi-experimental study, an organization introduces a new employee training program on customer satisfaction and wishes to study the impact of this intervention on the number of customer-related complaints before and after the intervention. In this specific example, the researcher has leeway to control the subjects (employees) that undergo the training program but has no capacity to randomize or control customers who may feel either satisfied or unsatisfied with the level of service accorded by the employees.
In the above example, the independent variable is exposure to the training program and the dependent variables include (1) severity of customer complaints, (2) performance of employees before and after exposure to the training program, and (3) organizational competitiveness and performance before and after exposure to the training program. In such a scenario, it becomes more difficult for the researcher to, for instance, say that it is the independent variable (training program) that is responsible for changes in the dependent variable (severity of customer complaints), not only due to lack of control of the dependent variable but also due to other confounding variables that differ across conditions (Harris et al., 2006; Levy & Ellis, 2011).
There are several essential conditions that are required in quasi-experimental designs. Extant literature demonstrate that these designs are not only used when it is logically impractical or unethical to conduct a randomized controlled trial, but also in situations when the researcher want to demonstrate instances of causality between an intervention and an outcome (Harris et al., 2006). Additionally, quasi-experimental studies occur in instances where the level of control of the variables of interest is generally weaker than in a true environment, not mentioning that there is no random allocation of study subjects to groups (experimental vs. control) in purely quasi-experimental studies (Creswell, 2003).
The study highlighted in this paper can be changed into a true experiment by enhancing the level of control of all the variables so that the researcher is able to manipulate them to know with an adequate level of confidence that it is the independent variable that is causing changes in the dependent variables (Ellis & Levy, 2011). A true experimental study must have an experimental group that the researcher manipulates or controls, a control group that do not receive the experimental treatment, not mentioning that the researcher must have the capacity to randomly allocate participants to either experimental or control groups (Creswell, 2003). In our case example, therefore, the researcher may change the study from being quasi-experimental to a true experiment by recruiting a specific number of customers into the study and randomly allocating them to either group – control or experimental. Then the customers should be exposed to the independent variable (employees who have received/not received the training) to see if exposure to training indeed reduces the level of customer complaints.
Extant literature demonstrates several advantages of using a true experimental research design over a quasi-experimental design. In a true experimental design, for example, a researcher is able to say with an adequate level of confidence that it is the independent variable(s) that is responsible for changes in the dependent variables, but in quasi-experimental the outcome can be demonstrated to differ statistically with the intervention due to weak control of the variables (Harris et al., 2006). While researchers using true experimental designs are endowed with the capacity to control all important confounding variables due to randomization, the same cannot be achieved in quasi-experimental designs, implying that statistical tests done in quasi-experimental studies may be meaningless due to lack of proper randomization (Ellis & Levy, 2009). Another advantage of true experimental designs, which is intrinsically related to the previous one, is that the findings of true experimental studies are able to stand up to rigorous statistical scrutiny by virtue of the fact that the researcher is able to control other factors or influences that may have affected the results. This is not the case in quasi-experimental designs (Levy & Ellis, 2011).
References
Creswell, J.W. (2003). Research design – Qualitative, quantitative and mixed methods approaches (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc.
Ellis, T., & Levy, Y. (2009). Towards a guide for novice researchers on research methodology: Review and proposed methods. In E.B. Cohen (Eds.), Growing information: Issues in informing science and information technology, volume 6 (pp. 323-337). Santa Rosa, CA: Informing Science Press.
Harris, A.D., McGregor, J.C., Perencevich, E.N., Furuno, J.P., Zhu, J., Peterson, D.E., & Finkelstein, J. (2006). The use and interpretation of quasi-experimental studies in medical informatics. Journal of the American Medical Informatics Association, 13(1), 16-23. Web.
Hoe, J., & Hoare, Z. (2012). Understanding quantitative research: Part 1. Nursing Standard, 12(15-17), 52-57.
Levy, Y., & Ellis, T.J. (2011). A guide for novice researchers on experimental and quasi-experimental studies in information systems research. Interdisciplinary Journal of Information, Knowledge & Management, 6(1), 151-161.
Welford, C., Murphy, K., & Casey, D. (2012). Demystifying nursing research terminology: Part 2. Nurse Researcher, 19(2), 29-35.