I was talking with a statistics professor from a state university who, like the state universities in my state is experiencing budget cuts as states seek to find ways to curb their spending. Indeed, many of us are living in the midst of real challenges. Some interesting research demonstrates that the challenges in which we live, may actually make us stronger http://www.psychologicalscience.org/index.php/news/releases/traumatic-experiences-may-make-you-tough.html This is the kind of research study that could be used as an interesting example for students.
As we attempt to get across to students the fundamental concepts of applying statistics to answer questions, nothing seems to help do that better than showing students how others are doing just that. Of course, in this world of complex theories and multivariate statistics/mathematical modeling, much of the contemporary research may simply take up too much class time to be useful. So, today, I thought about sharing a few articles, like the one above, that could have use for us as current concrete examples in applied statistics.
The “Traumatic Experiences May Make You Tough” (link above) example from the Current Direction in Psychological Science is a good example for talking about differing hypotheses (less stress is good/stress is good), qualification hypothesis testing (some stress is good, but not too much, and not too little), non-linear relationships, correlation and regression.
As you explain the differences between an independent variable and a subject variable, the study on how religious people feel better about themselves and their lives is an easy to understand example. http://www.psychologicalscience.org/index.php/news/religious-people-feel-better-than-non-believers-but-only-in-devout-societies-study.html
If your students are like mine, you may have already heard about this study, chronicled on line at CNN as it really seemed to catch their attention … after all, haven’t we all had “psychopaths” for a boss??? http://edition.cnn.com/2012/01/19/business/psychopath-boss/?hpt=ibu_t3. This study actually helps characterize descriptive statistics very well. It seems 1% of the general population are psychopaths, while 4% of “management” exhibit such behavior. This is also a great example for talking about sampling error, chance differences and real differences.
Of course, all of these examples can be used to remind students association and causality are not the same thing.
By keeping your examples in class current, and from websites that are easy for students to access and read students can begin a pattern of reading about applied statistics on a regular basis. Some web sites include: CNN, MSNBC, FOXNEWS on line all have Science Sections that regularly have research in areas of psychology, education, business, and social sciences. APA.org and Psychologicalscience.org also provide students with summaries of research for which they can easily understand.
One exercise I have students several times throughout a semester is to find hypotheses they find interesting. Some times I ask them to generate those hypotheses, other times I ask them to seek out an article that is testing an hypothesis and bring it to class. Many times, students find the hypothesis from one of the websites listed above. They have to keep the hypotheses in their note book, and after a while, they do add up. I ask them to present a hypothesis that fits a criteria … e.g., it has an independent variable, it has an interaction. This kind of activity not only gets students thinking, but it gets them reading and critiquing what they read. Though most of the students we teach in applied statistics will probably never become statisticians themselves, we can be certain they will all become consumers of statistics … and getting them to read about research that uses applied statistics to draw its conclusions is a great way to start them on the path of being critical consumers.