Step 1
Scientists apply three traditional study designs, namely experiment, quasi-experiment, and non-experimental study. Random assignment and control groups are the distinctive framework elements and tools that distinguish the first category from the others. According to Rogers and Reverz (2019), it is a procedure in which “participants are randomly placed into the control and the experimental condition following a chance procedure” (p. 134). The presence of a control group is necessary for the experiment’s success and objectivity. The control group is used to observe and analyze critical variables and causal relations and gather desired data during pre- and post-tests (Rogers & Reverz, 2019). The format and nature of the two elements discussed differ in a quasi-experiment.
When doing a quasi-experimental study, researchers do not perform a random assignment. They use techniques and approaches for assigning subjects non-random by design (Rogers & Reverz, 2019). The framework may not present the control group and associated procedures. It is not a strict requirement and can be replaced by a comparison group (Rogers & Reverz, 2019). Rogers and Reverz (2019) argue that “a comparison group is an additional experimental group that receives a different experimental treatment” (p. 134). The non-experimental stage differs from the previous two in that the experimenter cannot manipulate the independent variable. Another unique feature is that the presence of both a control group and a comparative one is not required.
Step 2
Each research design is a broad category that includes different academic investigation types. There is a pretest-posttest control group design in an experimental approach that involves randomly assigned participants and two measurement procedures of the desired results before and after the intervention (Rogers & Reverz, 2019). Its benefits are that it allows one to study in detail the nature of the effect and its scale. Its disadvantages are an overly complicated selection of subjects and low reproducibility of outcomes (Choueiry, n.d.a). The Solomon four-group design is another experiment with four teams of participants divided according to whether they passed the treatment and subsequent measurement (Choueiry, n.d.b). It provides one with quality threat monitoring and pretest sensitization detection, but this type of scientific exploitation is financially and materially costly.
There is some variability in quasi-experiments and non-experimental methodologies too. For a quasi-experiment, these are randomized block designs with allocation based on the specific traits of the subjects and their random treatment and factorial one when effects are measured and their scaling. The former allows one to exclude unwanted variables and avoid the error terms (Choueiry, n.d.c). This design is not flexible, and its sampling model poses many difficulties (Choueiry, n.d.c). A factorial experiment gives the researcher all combinations of effects to explore, but there is a risk of variables being influenced by unwanted factors. A one-shot case study where a single team of subjects is analyzed and single-variable research aimed at studying one specific topic are standard non-experiments. Both are financially and materially cheap for experimenters. Their shared shortcomings are the inability to study causality and the weak reproducibility of inferences.
Step 3
Validity of measuring tools and chosen metrics is the key to truthful quantitative results and conclusions. Unverified and unconfirmed, it undermines the experimenters’ efforts, their hypotheses, and theories. Scholars distinguish seven types of validity, with face one as those associated with the researcher’s primary perception of intervention results (Sürücü et al., 2020). Internal, external, and statistical conclusion validities are linked to the final data’s reproducibility and cause-and-effect presence. Content, construct, and criterion-related ones are needed to determine the validity of the research topic, methods, and measures.
Step 4
Two effective ways to control participant characteristics in a quantitative study are to group subjects into randomized blocks and to measure the two compared groups during pre-tests and post-tests. An example of the former is the Yarnold and Linden (2019) intervention with blood samples. The COVID-19 study on distancing measures by Siedner et al. (2020) represents the second measure. System complexity and variance are interconnected internal validity threats. Since the system’s complexity overloads the data’s analysis and the unwanted variance creates too many variables, they must be considered.
References
Choueiry, G. (n.d.a). Pretest-posttest control group design: An introduction. Quantifying Health. Web.
Choueiry, G. (n.d.b). Randomized block design: An introduction. Quantifying Health. Web.
Choueiry, G. (n.d.b). Solomon four-group design: An introduction. Quantifying Health. Web.
Rogers, J., & Revesz, A. (2019). Experimental and quasi-experimental designs. In H. Rose & J. McKinley (Eds.), The Routledge Handbook of research methods in applied linguistics (pp. 133-143). Routledge.
Siedner, M. J., Harling, G., Reynolds, Z., Gilbert, R. F., Haneuse, S., Venkataramani, A. S., & Tsai, A. C. (2020). Social distancing to slow the US COVID-19 epidemic: Longitudinal pretest–post-test comparison group study. PLOS Medicine, 17(8), 1-12.
Sürücü, L., & Maslakçi, A. (2020). Validity and reliability in quantitative research. Business & Management Studies: An International Journal, 8(3), 2694-2726.
Yarnold, P. R., & Linden, A. (2019). Randomized blocks designs: Omnibus vs. pairwise comparison, fixed vs. relative optimal discriminant threshold, and raw vs. ipsative z-score measures. Optimal Data Analysis, 8, 122–129. Web.