On the 11 December 2018, the Royal Statistical Society (RSS) Merseyside Local Group held an event titled 'Hidden Statisticians', which showcased the use of statistics in unexpected places. The event was attended by 25 individuals, approximately half of whom were RSS members. Attendees were from a range of institutes including United Utilities, various departments of the University of Liverpool (including Biostatistics, Psychology, Institute for Financial and Actuarial Mathematics and Chemistry), Liverpool John Moore’s University, Liverpool School of Tropical Medicine and the Royal Statistical Society.
The event began with a talk titled 'Homelessness by Numbers' from Rob Farnos of the Whitechapel Centre, which works to prevent rough sleeping and homelessness in the Merseyside area. He discussed how use of statistics is necessary when securing funding, communicating the work undertaken by the centre to funders and the public, and to inform service development. Examples were given, including statistics indicating that of the 285 people recorded as rough sleeping in October 2018, 22% were new to rough sleeping in the last 12 months, whilst 38% were returning to homelessness.
The second talk was delivered by Dr Jennifer Sigafoos and Dr Samantha Currie of the University of Liverpool School of Law, titled 'Led astray? Statistics and the law'. The talk covered several examples where statistics are vital to law work, including tracking Home Office asylum decisions for victims of trafficking, and data from the Citizen’s Advice Bureau. Various matters were highlighted, such as the issue of access to data. An example of this was that database changes for the Citizen’s Advice Bureau made it impossible to access the old database, meaning that the original raw data was no longer available, leaving only summary statistics. Questions included how to improve use of statistics in law, with the discussion including the role of universities to teach statistics to undergraduate law students, and the tool kit 'Statistics and Law' produced jointly by the RSS and the Law Society.
The third talk was delivered by Dr Padraic Walsh of First Data, and discussed fraud prevention methods. In the past, companies have used historic data to detect fraud using rule-based algorithms. However, due to the increase in remote purchase fraud over recent years (partly as a result of the increase in purchases from mobile devices) rule-based fraud detection can be out of date by the time it is implemented. In the future, machine learning techniques such as deep learning neural networks could be used to predict fraud, however this requires analysis of large, complex, high dimensional datasets.
The final talk was given by Dr Lisa Holmes and Dr Andy Moss of Chester Zoo. Dr Holmes discussed use of statistics to model networks of relationships between the Asian elephant population in Chester Zoo. This is important for the conservation breeding programme, where elephants may be moved between zoos. In the future, such networks could be used to ensure that removal and introduction of elephants from a zoo population does not severely damage the social structure already in place. An example of using statistics to establish the efficacy of a new way of measuring the internal temperature of komodo dragons was also presented. Dr Moss then discussed data on visitors to the zoo – for example content analysis of work done by visiting school children before and after being taught about biodiversity demonstrated increased awareness of the subject at the end of their visit. The importance of use of statistics when making decisions in the zoo was highlighted.
Events in 2019 held by the RSS Merseyside local group will be announced over the coming months.