A principal wants to evaluate whether a new math curriculum improves math achievement. Which evaluation method is most effective?

A principal wants to evaluate whether a new math curriculum improves math achievement. Which evaluation method is most effective?

• A. A comparison of test scores of students randomly selected to use the new curriculum with the scores of students who did not use the new curriculum
• B. Pretest-posttest within the same group
• C. Anecdotal teacher impressions
• D. Only comparing state test scores

Answer: (C)

To determine whether a new math curriculum actually boosts achievement, you want a design that makes groups comparable at the start and then measures outcomes after the instruction. The most robust approach is a randomized assignment where students are randomly placed into the new curriculum or a control condition (continuing with the current curriculum), with math achievement measured after the intervention (and ideally with a pretest to verify baseline similarity). This setup controls for selection bias and other confounding factors, so any observed difference in outcomes can be attributed to the curriculum itself. Using a single group with a pretest and posttest can show change over time, but it can't rule out other explanations like maturation, outside events, or teachers’ growing familiarity with the material. Relying on anecdotal teacher impressions is subjective and biased, and comparing state test scores alone often reflects many uncontrolled variables, such as prior achievement, test alignment, and instructional differences across schools. So the strongest evidence comes from a randomized controlled design (with appropriate pretest/posttest measures), because it best isolates the effect of the curriculum from other influences.

To determine whether a new math curriculum actually boosts achievement, you want a design that makes groups comparable at the start and then measures outcomes after the instruction. The most robust approach is a randomized assignment where students are randomly placed into the new curriculum or a control condition (continuing with the current curriculum), with math achievement measured after the intervention (and ideally with a pretest to verify baseline similarity). This setup controls for selection bias and other confounding factors, so any observed difference in outcomes can be attributed to the curriculum itself.

Using a single group with a pretest and posttest can show change over time, but it can't rule out other explanations like maturation, outside events, or teachers’ growing familiarity with the material. Relying on anecdotal teacher impressions is subjective and biased, and comparing state test scores alone often reflects many uncontrolled variables, such as prior achievement, test alignment, and instructional differences across schools.

So the strongest evidence comes from a randomized controlled design (with appropriate pretest/posttest measures), because it best isolates the effect of the curriculum from other influences.