#paired-data

Question 1: Single choice

A teacher asks her students at the start and at the end of the semester whether more active learning in class would help them learn, and wants to know whether the proportion answering "yes" changed. Why is the two-proportion Z-test for independent samples not appropriate?

The same students answer twice, so the two samples are not independent.1
A proportion measured at two times cannot be compared with any hypothesis test.2
A two-proportion test requires the two groups to have different sample sizes.3
Yes/no answers are categorical, so a $t$ -test for means is required instead.4

Timed test