Quasi-experimentation A Guide To Design And Analysis Pdf Apr 2026

"Lena, look," Hartley said, tapping his desk. "I installed it in Ms. Chen’s third-grade class. She’s our best teacher. The other third-grade class, Mr. Abel’s, is using the old curriculum. After three months, I’ll compare their test scores. Simple, right?"

Hartley laughed. "You quasi-people have a workaround for everything."

But to be rigorous, she added a and used Huber-White robust standard errors (because monthly scores from the same class aren’t independent — a key point from quasi-experimental guides). quasi-experimentation a guide to design and analysis pdf

Here’s a short, engaging story that captures the essence of (as in the spirit of Cook, Campbell, and Shadish’s work, often summarized in guides like Quasi-Experimentation: A Guide to Design and Analysis ). Title: The Principal’s Predicament Dr. Lena Torres, a research consultant, faced a familiar problem. The school principal, Mr. Hartley, had just spent $50,000 on a new "MindGrow" reading software. He needed to know if it worked.

Lena sighed. "That’s not simple. That’s a . Ms. Chen is a star teacher. Her kids were already scoring 15% higher before the software. If her class does better afterward, was it the software or just… Ms. Chen?" "Lena, look," Hartley said, tapping his desk

Lena smiled. "That’s the guide to design and analysis. No randomization? No problem. Just more thinking." Quasi-experimentation isn’t “second-best.” It’s a toolkit for causal inference when experiments are impossible. Master the threats (history, selection, maturation, regression), choose a design (ITS, DID, nonequivalent groups), and analyze with care — robust standard errors and pre-trend checks are your friends.

Hartley frowned. "So I should flip a coin? Randomly assign kids to software or no software?" She’s our best teacher

Hartley nodded. "So we keep the software, but we train Mr. Abel on it too."

"You can’t," Lena said. "Parents would riot if their kid got ‘no software.’ Plus, the software is tied to Ms. Chen’s classroom computers. You have a —real-world, no randomization. But that doesn’t mean it’s hopeless."

"Exactly," Lena said. "And next time, if you can’t randomize, use a — give half the classes the software in Phase 1, the other half in Phase 2. Compare each against itself over time."

Result: The +7 points was statistically significant (p < .01) and practically meaningful. Lena presented to Hartley: "The software works, but only by 7 points, not the 15-point jump you saw in the raw comparison. The raw difference was inflated by Ms. Chen’s prior excellence."