The algorithms for the different research designs.
Sample Solution
Research Design and Sampling for Comparing Student Groups
Quantitative Algorithm:
According to Paul D. Leedy & Jeanne Ormrod, Practical Research: Planning and Design, Pearson Education, 10th Edition (2015) [invalid URL removed], one algorithm used to identify research designs considers the level of manipulation of the independent variable and the presence of a control group.
Research Design:
The appropriate research design for comparing two groups of students (school team vs. no school team) is a quasi-experimental design specifically a non-equivalent groups design.
Explanation:
- In a true experiment, the researcher randomly assigns participants to groups (intervention and control). Here, we cannot randomly assign students to be on a school team or not. This lack of random assignment removes the ability to establish causation between being on a school team and the outcome variable (e.g., academic performance).
- A non-equivalent groups design acknowledges the pre-existing groups and compares them as they are.
Sampling Method:
To achieve the highest level of research design possible within the constraints (non-equivalent groups), we should strive for a sampling method that minimizes selection bias.
Selected Method: Propensity Score Matching
Propensity score matching is a statistical technique used in non-experimental research to create comparable groups. Here's why it's a good fit:
- It attempts to statistically balance the two pre-existing groups (school team vs. no school team) on factors that might influence the outcome variable (e.g., prior academic performance, socioeconomic background).
- By statistically matching students on these characteristics, we can create groups that are more similar, allowing for a more fair comparison of the impact of being on a school team.
Conclusion:
While a true experiment with random assignment would be ideal, due to the nature of pre-existing student groups, a non-equivalent groups design with propensity score matching offers a more robust approach to minimize bias and strengthen the internal validity of the research.