A tutoring program improves test scores for most participants. The argument: expand the program.

• A. The program is inexpensive to implement.
• B. The program has been shared by other districts.
• C. A randomized controlled trial across several schools showed similar score improvements.
• D. The program requires extensive trainer certification.

Answer: (C)

Expanding a program should be based on evidence that its benefits will generalize beyond the original group. A randomized controlled trial conducted across several schools that shows similar score improvements provides strong proof that the gains come from the program itself, not from peculiarities of one site or lucky circumstances. Random assignment minimizes bias, and seeing consistent improvements across diverse settings suggests the effect is robust and likely to occur in new schools too.

Other points may touch on cost, spread to other districts, or implementation requirements, but they don’t by themselves confirm that the program will work in different contexts. They don’t address generalizability or causal impact the way a multi-site randomized trial does.