Research experiments test theories, establish behavioral norms or testbed new market designs or policies. They are conducted in carefully controlled environments using with financial incentives and procedures that are intended to allow for scientific replication. They are often computerized, require a lot of advance planning and often take more than an hour to complete. Subjects are not told the purpose of the experiment to prevent "experimenter effects" where subjects do what they think the experimenter wants. In addition, subjects are not allowed to communicate during the experiment unless communication is part of the environment to be studied.
Unlike research experiments, classroom experiments do not require that students be paid in order for the experiment to be successful. Often the desire to "do well" in class is sufficient motivation for students. Sometimes instructors use performance in the experiment as part of the grade for the exercise or introduce other incentives to increase student interest. Learn More about Incentives
Sometimes people are concerned about whether they need approval from their University's Institutional Review Board or Human Subjects Committee in order to conduct classroom experiments. Ultimately this depends on what you plan to do with the data and what your university's policies are - it is best to check to be certain. If you are using the experiments purely for teaching and do not plan to publish the results then classroom experiments are usually treated like any other classroom activity. It is a different matter if you hope to publish the data you collect, either as an experiment on successful teaching or as a teaching exercise for other instructors to use. In either case you need to work with the appropriate office at your university well in advance of the start of the semester to make sure that your activities are compliant with your university's policies.