A study is referred to as experimental if it contains random allocation of participants to experimental conditions or treatments in which a variable of interest is manipulated. Such experiments can allow claims that the manipulation has caused changes in outcomes.
For example, if we wanted to study the influence of rewards during class on students’ biology exam scores in an experimental study, we would randomly assign students to two conditions: In condition 1, students would receive candy bars for active participation in class, whereas in condition 2, students would not receive any candy bars.
Then we would observe the exam scores for each group of students, to judge if our candy-bar treatment improved the scores compared to the no-candy-bar control condition. We could then conclude if rewards cause better exam scores in this context.