The cat in the stat: how cats and dogs teach us everything we need to know about statistics

Aysha Bellamy, Department of Psychology

College Postgraduate Teaching Commendation 2020

As a statistics tutor for CeDAS, and demonstrator for an MSc Statistics course, I encounter students who struggle with statistics. These students may not have studied statistics for a long time, or ever; they may feel ‘overwhelmed’ by the multitude of equations in-lecture. Perhaps they feel that they will never be good at statistics, if they are not good at maths.

These students highlighted to me the need to explain statistics in a way that goes back to basics, by explaining the necessity and power of statistics with an example that avoids mathematics. This led me to devise the ‘cat in the stat’ example which, after initial positive feedback, I now employ when teaching in one-to-one sessions. The example is simple: I ask students whether dogs are bigger than cats. This organically leads to discussing research questions and hypotheses. I then draw a sample of cats and dogs, ensuring that the cats are uniform in height, but the dogs consist of chihuahuas and greyhounds. The students then identify that chihuahuas are smaller, but most dogs are bigger than cats. We then discuss the problems of trying to measure all cats and dogs, so a sample and sample mean are needed. With prompts, the students then identify that a mean for dogs would be biased by the very small and very large observations of some breeds. This leads to why standard deviations are vital to measure spread, where upon the students often identify that the standard deviation for dogs would be larger than cats. This leads to a discussion of homogeneity of variances (are our samples really comparable?) and finally, the necessity of a t-test to compare the cats and dogs’ height. This is a basic analogy but can be expanded upon as necessary to illustrate more advanced statistical concepts.

The ‘cat in the stat’ example is beneficial for a multitude of reasons. Firstly, colour and drawings can be used throughout to aid visualisation. Removing numbers from the example is beneficial for students with dyscalculia/dyslexia, whose needs may reduce their ability to follow equations. Moreover, the cat example is presented in discrete steps, which aids DDS conditions that may affect a student’s ability to parse into discrete components (e.g. dyspraxia).

As well as being inclusive for different learning styles and DDS students, this example is also beneficial for students from different backgrounds. Students have diverse educational experiences, and so baseline statistical reasoning is not uninform. Moreover, RHUL is multicultural and statistics may be taught differently in different countries. The cat example can establish baseline individual strengths and weaknesses in students from diverse backgrounds.

Lastly, the culture of university changes as RHUL becomes more popular. Lectures become larger with less individualisation. Tailored one-to-one support, and the cat in the stat example, shows the student how much they know about statistics already. This increased confidence allows students to continue with one-to-ones, self-study and email follow-ups, showing a quantifiable increase in engagement.


Reflections and Feedback 

Most students ‘agree’ or ‘strongly agree’ that these sessions are helpful. Students that do leave feedback praise the every-day English and use of visual examples to reinforce their understanding. Informal feedback provided during the sessions indicates the students are often presently surprised after hearing the ‘cat-in-the-stat’ framework by how much they already know about statistics. After seeing the example, most students can express why statistics are useful and when certain statistics are needed, more confidently. 

Another potential benefit of this analogy is that it can be tailored to almost any students’ circumstance. CeDAS is open to students from all years needing statistics; from courses as diverse as social science to biology and economics. The applicability of this analogy cannot be overstated. Some MSc students I teach come from an educational background without statistics, so this analogy helps catchthem up to speed. Moreover, the base example can be applied to whatever area a student feels uncomfortable with. 



See other case studies on our Seminars, workshops & labs page