On June 18 2019, the RSS Glasgow Local Group hosted a Glasgow Science Festival lecture by the William Guy lecturer Dr Lee Fawcett from the University of Newcastle. As Guy Lecturer, Lee has a volunteer role in RSS to provide school outreach activities. The lecture was attended by around 66 people from three different schools in the Glasgow area and was an interactive talk with handouts and the students answering questions and being encouraged to ask questions themselves.
The main takeaway from the talk was that statistics is not just calculating the mean, medians or modes but rather applicable to every other scientific field. The main aim of the talk was to inspire the students that statistics is an exciting field of study.
Before starting the talk, Lee checked the statistical knowledge of the students. None of the students knew about the normal distribution as most of the audience were geography students and their teachers which explained the lack of statistical knowledge.
Lee had prepared two versions of the handouts, depending on the knowledge of the students on the normal distribution. There were plenty of handouts to go around and the audience started to complete the tasks. None of the work required the use of calculators and there was plenty of time to fill the gaps. Furthermore, it was made very clear when and what to copy from the slides, as the colour scheme of the slides changed from blue to green. Besides calculations and information, the students even got the opportunity to fill out missing points in a graph based on their calculations.
The talk was split into four parts. The first part was an introduction to extreme statistics and provided an overview of famous storms in the last one hundred years. Specific attention was given to the Katrina storm which hit New Orleans in 2004. Lee wanted to address the fact that we often hear on the news when a storm hits a location but that we do not hear about the aftermath and the amount of time it takes to recover from such a storm.
The second part of the talk introduced the concept of relative frequency – a fraction of the times a certain event has occurred. Lee showed the students how to calculate the probability that the annual maximum wave height (AMWH) would be about 8.75 ft. The students then had to repeat the work for exceedances above 11.25ft and 14 ft. The value of 14ft was the most crucial as the Katherina waves were 14.4ft tall. The task introduced the idea of using historical data. Another focus was sometimes modelling the average is not informative enough, but you might be interested in modelling the extreme values, in this case the maximum.
The third part of the talk introduced the idea of (simple) probability models. The students were introduced to Extreme Value Theory and Gumbel’s model for exceedance probabilities. I have a very basic understanding for both, but Lee delivered a very simple and clear explanation on both them which made me go home and actually research more on the topic. Similarly, the students were even interested in finding out more about other models and Lee recommended Fréchet and Weibull. Additionally, Lee recommended 'Statistics of Extremes' by Emil Julius Gumbel as extra reading for anyone who wants to find out more about extreme theory. In this part the students got to estimate more relative frequencies and add those points to a plot.
The fourth and final part of the talk required the students to analyse the graph and draw conclusions about the height of the protective walls around New Orleans. The students felt really engaged and gave a lot of answers when asked to analyse the results from the graph. The teachers were also getting very involved and were asking questions about the estimations.
Lastly, Lee recommended a R Shiny app he has developed where the students and the teachers could explore different extreme events.
A lot of students went to speak to Lee after the talk about further reading. Many of the students took from the promotional badges and pens before leaving. Before leaving, all the teachers filled out feedback forms where they had rated the event as excellent and very good and were willing to participate in future events like this. The talk was very successful in promoting statistics as a career path to school students.
Lee was presented with a certificate from the RSS for his work as a William Guy lecturer.
Yoana Napier is a Statistics PhD student at the University of Glasgow. She works on high resolution modelling of air pollution in Scotland using Gaussian Process emulators. Previously, she had worked in the field of financial statistics.