A student researcher conducts an experiment that investigates the effects of a new mathematics website on student achievement. She randomly assigns students to two groups: one that receives supplemental instruction from the website and one that doesn't receive the supplemental instruction. Each sample consists of 20 students. Observations of student engagement are rated by the researcher, according to an engagement scale. How could this study be improved?

Select all that apply.

A. Have another researcher provide ratings, as well.

B. Administer a quiz or exam that gauges student understanding.

C. Increase the sample size.

D. Limit observations to students who are engaged.