Dr Tropez Conducts a Study Examining the Relationship Between Exposure

Dr. Tropez conducts a study examining the relationship between exposure to religion in popular media and religiosity (defined as "the belief in and worship of a superhuman controlling power, especially a personal God or gods"). In his study, he randomly samples a group of 113 participants from his local community college. He generates a list of movies with religious themes (Year One, Dogma, Passion of the Christ) and asks participants to indicate how many they have seen. He then measures religiosity by asking participants how many times they have attended a house of worship (e.g., church, temple, synagogue, mosque) in the past year. He finds that having seen a higher number of religious-themed films is positively correlated with religious attendance.
A colleague of Dr. Tropez questions how generalizable his study is to other participants by highlighting that he only studied 113 people. Dr. Tropez responds that generalizability comes not from the "how many" of the sample but the "how" of the sample. What does Dr. Tropez mean, and how generalizable is his study?