A brief teaching note on quasi-experiment as a research method: for MBA and Housing Studies students

A brief teaching note on quasi-experiment as a research method: for MBA and Housing Studies students (information from perplexity.ai)

My question 1: "Briefly highlight the main ideas of quasi-experiment as a research method in social science with reference to the following topics: (1) favored research philosophy, (2) favored research method design techniques, (3) favored research method design quality criteria, (4) favored data analysis techniques, and (5) favored researcher's impression management approach in conducting the quasi-experiment."

The reply: "Quasi-experiment as a social science research method can be summarized under the given topics as follows:

(1) Favored research philosophy: Quasi-experimental research aligns with a realist and pragmatic philosophy. It aims to establish causality in naturalistic settings where true randomization is impractical or unethical, balancing realism (high ecological validity) with some level of control to infer cause-effect relationships.

(2) Favored research method design techniques: Quasi-experiments typically use non-equivalent groups, pretest-posttest designs, interrupted time series, regression discontinuity, and natural experiments. They rely on selecting comparison groups without random assignment and use designs that approximate experimental conditions despite practical constraints.

(3) Favored research method design quality criteria: Key criteria focus on internal validity (managing confounds since random assignment is absent) and external validity (real-world applicability). Strategies include matching groups on key covariates, statistical controls (e.g., regression), use of longitudinal data, and robustness against threats to validity to strengthen causal inference.

(4) Favored data analysis techniques: Analysis often involves propensity score matching, interrupted time series analysis, regression analysis, t-tests, and correlational methods to address confounding and approximate causal effects. Methods aim to control for bias due to lack of randomization and isolate treatment effects statistically.

(5) Favored researcher's impression management approach: In conducting quasi-experiments, researchers adopt rigorous design transparency and statistical control to manage impressions about validity and credibility. They emphasize matching, controlling for confounders, and openly addressing limitations to sustain trustworthiness and reduce skepticism about causal claims.

In summary, quasi-experimentation in social science pragmatically balances realistic natural settings with methodological rigor to infer causality, employing non-randomized designs, sophisticated statistical controls, and transparent reporting to maintain research quality and credibility."

My question 2: "Please provide 1 illustrative example for each of the following two research themes:

1. MBA related: "to assess the effectiveness of using quality circle to improve the customer service of the Hong Kong jewelry retail outlets of ABC Ltd in 2025."

2. Housing Studies related: "to assess the effectiveness of a new home-ownership incentive scheme based on generous government subsidy for gen z in Hong Kong in 2025"."

The reply: "Here are illustrative quasi-experimental research examples for the two themes:

1. MBA Related Example (Quality Circle in Jewelry Retail):

• Research Theme: To assess the effectiveness of using quality circles to improve customer service of Hong Kong jewelry retail outlets of ABC Ltd in 2025.

• Illustrative Quasi-Experiment: ABC Ltd implements quality circles in selected jewelry retail outlets starting January 2025, with other outlets continuing usual customer service operations. Customer satisfaction scores and service quality metrics are collected monthly for all outlets before and after introduction. Customer service improvements in outlets with quality circles are compared to control outlets without intervention to infer effectiveness under real-world operational conditions without random assignment.

2. Housing Studies Related Example (Home-Ownership Incentive for Gen Z):

• Research Theme: To assess the effectiveness of a new home-ownership incentive scheme based on generous government subsidies for Gen Z in Hong Kong in 2025.

• Illustrative Quasi-Experiment: The government launches the subsidy scheme in mid-2025 targeting Gen Z home buyers. Housing purchase rates, affordability measures, and satisfaction with housing conditions are tracked for eligible Gen Z applicants compared to a similar demographic group just outside the eligibility criteria (e.g., slightly older millennials). Changes pre- and post-scheme introduction and between groups are analyzed to estimate the scheme's impact on home ownership uptake and housing stability.

Both examples use quasi-experimental designs by comparing treatment and comparison groups without randomization, employing real-world policy or organizational implementations in 2025 Hong Kong contexts for MBA and housing studies respectively. This approach aligns with the method's focus on practical causality assessment in naturalistic settings."

A collection of blog notes on using chatgpt for research purpose.