In my PSY 220 (Behavioral Statistics) class, students really struggle with both the concepts and the calculations. My exams are generally 30-40% multiple choice (conceptual) with the rest being calculations. I always advise students to make flashcards because I feel they are very effective. They may be low level learning, but stats requires a lot of low level learning. I also feel that BMCC students struggle with low level learning for various reasons. It’s easy to just say engage them in some high level discussion, but this is not always realistic. I do ask the students to find real world examples of misleading statistics, but this is more geared towards critical thinking rather than learning about statistics itself.

Concept maps might work with some of the conceptual stuff, but linking the different statistical tests is much more advanced. For example, regression and ANOVA are both part of the General Linear Model where ANOVA is a specialized case of regression. To have my students draw these links conceptually I think is asking for too much. I am trying to keep it simple and not overload the students. Unfortunately, a lot of the information for my course has to be memorized. For instance, the number of degrees of freedom for the independent groups t-test compared to the correlated groups t-test. One can either memorize this information, or alternatively, try to conceptualize the number of population parameters to be estimated. In this case, one would also have to know that a correlated groups t-test uses difference scores and we are estimating the parameters of the distribution of differences scores (therefore df= N-1). As opposed to two separate distributions with the independent groups t-test (therefore df= n1+n2 -2). I do talk about these theoretical distributions in class, but it is a tough ask for our students to conceptually grasp this. Teaching about a theoretical distribution of all possible scores (or more commonly sample means derived from an unmeasurable population) is a big ask for students who barely grasp basic mathematics and struggle with algebra. Whenever I talk about sampling distributions and the central limit theorem, I am met with quizzical looks. There are some curious students who are interested, but I am afraid to expand on things since the rest of the class would get buried in extraneous detail. This comes back to the heterogeneity of the classes I teach (10 % of the students are ready for more complex material, 90% are not).

I love BMCC students and the diversity that they represent. We had a great deal of diversity at Rutgers, but the students tended to be spoiled and defiant. I need to do a better job encouraging students and reinforcing their ability to succeed. I have to find a way to diminish their math-related anxiety and increase their growth mindset. I had previously mentioned using meditation at the start of each class. Another intervention that I would try is using expressive writing to decrease math anxiety. It would be a daily journal for them to express their anxiety about the class and/or the topics covered. This is similar to the Pennebaker (1986) expressive writing paradigm that has been successfully implemented in a variety of contexts. Obviously, math anxiety is only one part of why students struggle with statistics, but it is an important part that can be addressed.