Mark decides to conduct a variation of the original study in which he measures participants' ability to solve verbal analogies as the cognitive task instead of measuring their ability to solve math problems. He finds that people who drink caffeine beforehand solve more verbal analogies than those who drink water beforehand, suggesting caffeine consumption causes improved cognitive performance. This study is known as a:

1. Conceptual replication
2. Secondary replication
3. Replication-plus-extension
4. Statistical replication