You Conducted a Study That Investigated Whether People's Moral Beliefs

You conducted a study that investigated whether people's moral beliefs made them more likely to engage in helping behavior when a person is in need; therefore, you measured behavioral intent (i.e., how likely would you be to help a person in this situation). In providing informed consent, participants were informed that they would be completing a survey that asked about their general beliefs and how they think they would respond in certain situations. After completing those items, participants were told that the study was finished and they were thanked for their time. After the ostensible completion of the study, as participants left the laboratory building they all encountered a confederate in need of help. Participants' behavioral response (helping or not, and in what way) to the confederate in need was recorded as the participants' final data point.
Can you use this behavioral measure in your analyses (the way the scenario is currently described) - why or why not? Since the participants' behavior occurred in a public place (outside the laboratory building), does the participant need to provide informed consent, or is it considered "in the public domain?" Does the informed consent participants initially provided include permission to use their behavior as data in your research? If your answers to any of these questions is "no," what could you do to make that behavioral data useable?