Answer:
B. Select a random sample of students who go to the work laboratory
Explanation:
The best way to get acurate results is by selecting a random sample of students. If the students administers a survey to the students that are seated at the front of the laboratory he could get biased results.
The students that are seated in fron of the laboratory could obey to a certain characteristic (ie. They could be very applied students), which could definitely provide us with a different result.
The survey is just fine. We don't need any instrument to test statistical analysis understanding. Also, using a larger sample size of students seated in front of the laboratory won't make much difference. Finally, Administering instruments to a group who does not go to the work laboratory makes no sense. You can not measure effectiveness on people that don't assist to class.