Case Study – Post 16 – Exeter Maths School

In the first pilot study, two groups of mathematically able Y12 students (16-17 year old) explored how they could detect errors in a very large weather dataset, provided by the UK Met Office. This was a very open task and the students had to decide and explore by themselves what problems they would study, what statistical techniques could be used and what conclusions they could make.

For example, one of the groups focused on temperature. They attempted to separate data from different weather stations, and fix erroneous data points in the spreadsheets. In order to separate multiple weather stations that had been compiled into a single data file, they used the Latitude and Longitude coordinates. A Python script was written to make this process automatic.

Using mean and standard deviations, they created a boundary to judge whether a data point would be rejected or not. This process was very challenging even for very mathematically able students; for example the size of the dataset meant that even simple manipulation of the data was extremely challenging and their computers often crashed.

Click to enlarge

As well as the data wrangling and analysis techniques, the students also learned many softer skills, including planning a long-term project, and how to work in a team efficiently. They also wrote a report and presented their findings to a group of professionals at the MET Office.

University of Exeter logo Erasmus logo Erasmus logo