Prof Paul Curzon

Research

Computer Science Education, Public Engagement in Computer Science and STEM, Interaction Design in Healthcare, Interaction Design and Human Error

Interests

My research combines the areas of Computer Science Education, and interaction design for healthcare. I also have a strong interest in the public engagement with computer science and STEM more generally. In the past I worked primarily in the areas of automated reasoning and formal verification.

Computer Science Education

I have done a variety of work in the area of computer science education such as around making learning fun, teaching of computational thinking, teaching of programming (including program design). My current focus is on the application of the sociology theory Legitimation Code Theory (LCT) to better understand what makes good computer science teaching and so learning. One focus has been on use of LCT Semantics to explore good explanation. Another is the use of LCT Autonomy to understand teaching multiple topics together (eg computer science and the underlying maths).

Public Engagement in Science

A major aspect of my work is in the public engagement in computer science (and science, maths and engineering more generally). I aim to generate excitement not just about the School's research but about interdisciplinary research in the subject more generally. The main way of achieving this is through the internationally renowned public engagement project cs4fn (cs4fn.blog) that I created with Peter McOwan, and its sister project for teachers Teaching London Computing (teachinglondoncomputing.org)

Interaction Design in Healthcare

My current research in the area of interaction design applied to Healthcare is focussed on improving life with Rheumatoid Arthritis (RA). I am working with rheumatologist, Amy McBrayne, on the AtTRA project which started as part of the PAMBAYESIAN project. We are looking at the needs of those with RA and so opportunities and barriers to provide intelligent support for people with chronic diseases and specifically RA. I am working on understanding the real needs of both patients and clinicians. We developed a series of personas of patients to illustrate what life with the disease is like to feed in to the design process and interviewed and surveyed clinicians and patients about their needs, the opportunities of technology and barriers to its uptake.

I have also worked on interaction design and human error. In the CHI+MED project we applied earlier work on human error to the area of healthcare and medical device design, exploring ways to ensure design prevented or reduced human error. My work on human error was selected for presentation at the Royal Society Summer Exhibition, 2009.

Human Error and Cognitive Models

This originally extended my early research on the verification of hardware/software systems to human-computer systems. The idea is to consider the human operators of such interactive systems as part of the system under verification, so bringing systematic human error, not just software and hardware error, within the scope of the approach. I am, in particular, exploring the use of formal models of human behaviour based on results from cognitive psychology in the design of interactive systems. This work was in collaboration with UCL and Swansea University. Questions we are exploring include: 'How can formal models of human behaviour form the basis of verification methods that can detect design flaws that lead to systematic human error?'; 'How can empirical investigations inform the development of formal models of human behaviour used for verification, and vice versa?' and 'How can formally-based usability evaluation methods best support the analyst?'

Verification of Verification Systems

My early work on the design and verification of hybrid verification systems was in collaboration with Concordia University. We developed a verification system that combined the power of the MDG and HOL tools. It harnesses the abstraction techniques of the automated MDG multiway decision diagram (which is superior to boolean decision diagrams) system combined with theorem proving power of HOL to manage the process. In related work we developed a novel methodology that justifies importing results into a theorem prover using verified linkage theorems. It is based on a combination of compiler verification techniques.

Social Aspects of Interaction Design

I am also working on several projects investigating social aspects of interaction design, for example related to design for all. Questions of interest include: 'How can systems be designed so as to build on our cognitive strengths, especially as we age?'.

Publications of specific relevance to the Centre for Human-Centred Computing

2023

2022

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

Grants