Ethics in Statistics
The following statement was prepared by the NZSA Education committee after a meeting of the NZSA at NZAMT 2013.
Ethics – an opportunity for discussion with our students.
Ethics, and addressing ethical aspects of practice, is a very important part of all research that involves people and animals. Some of our current standards give us a wonderful opportunity to discuss this issue with our students.
Experiments Standards 91265 (2.10) and 91583 (3.11)
There are many very interesting historical psychology experiments that, nowadays, would never be permitted due to ethical considerations. See http://mentalfloss.com/article/52787/10-famous-psychological-experiments-could-never-happen-today for examples that include the Harlow’s Rhesus Monkey Experiments, the Milgram Obedience Experiment and The Standford Prison Experiment.
Examining some of these classic experiments allows students to look at the ethics from our current society’s perspective, including aspects of informed consent. This is an opportunity to naturally incorporate some of our New Zealand Curriculum values into a lesson.
Ideally, when students are designing and running their own experiments on other students, they should include a statement around consent. A standard phrase could simply be: “By completing this experiment you are indicating your consent to participate in this research.”
Questionnaire Design Standard 91263 (2.8)
Informed consent can be easily incorporated into the questionnaire design process with a simple phase:“By completing this questionnaire you are indicating your consent to participate in this research.” Discussion of what information questionnaire participants may want before consenting can occur for example: Will I remain anonymous? What will the information I give be used for? Who will see my responses? These considerations are especially important when questionnaires involve sensitive subjects.
Bivariate Data Standard 91581 (3.9)
Discussions of causality and how to prove a causal relationship take students back to designing experiments. An example of this is the claim that smoking causes lung cancer – how would you prove that nowadays? Note that a causal claim cannot be made from one observational study.
If you have any resources you use to teach ethical considerations involving statistical experiments or causality and would like to share these please contact email@example.com
Correspondence regarding this post may be sent to:
Add a teaching resource
Have you created a resource which is suitable for CensusAtSchool and you would like to share with other teachers?