Mitigating Unconscious Bias in Education
The project will run over 2020-2022 and develop and test the effectiveness of a training programme designed together with psychologists and educationalists to help teachers recognise UB and contrast it. The objectives are to increase teaching effectiveness and enjoyment by reducing the effects of UB in class, in teaching materials and methods and in students evaluation, and it aims to help increase inclusion, improving the educational and behavioural outcomes of all pupils. A pilot was run with Pearson Italy in 2017-18 and builds on an earlier intervention designed in the UK in 2015-16.
The Baby and the Bathwater: Professional Women and Delayed Fertility
The project seeks to examine the experience of infertility amongst professional women who due to prolonged pre-labour market qualifications start families at a stage of life in which natural fertility levels decline, the likelihood of (multiple) miscarriages increases and long periods of fertility treatments often ensue, with a particular view to understanding the career consequences of these events. The aims are to i) achieve a deep and wide-ranging study that can move forward to inform scholarly debate in the fields of politics, economics, law and public health; ii) create public awareness of (in)fertility as a legitimate and serious public/mental/health issue; and iii) identify best practices in the workplace to protect women in line with the 2010 Equality Act.
Lucky Girls! Women of the 1950s and the effects of pensions reforms
Women born in the 1950s in the UK have been subjected to successive pension changes (some of which unannounced) which have seriously impacted their welfare and that of their families and communities. The WASPI campaign explains the issues in detail. The project will comprise secondary data analysis, film and fieldwork with the key stakeholders and will assess the impact beyond affected women’s incomes (which has thus far been the sole focus of the government’s response) and into wider welfare effects to the women themselves (their mental and physical health) their families and communities (they are often carers of grandchildren, of partners and other family members as well as volunteers in their communities).
Algorithms to measure bias
This is a joint interdisciplinary project with Almudena Sevilla (UCL), linguist Sylvia Jaworska (Reading) and computational linguist Viviana Patti (UNITO) to devise deep learning algorithms to identify bias in written reports and appraisals.The majority of studies of discrimination focus on discrimination in observable, easy to quantify behavior, such as callback rates for job applications, offers for products and services, enforcement or housing. However, discrimination can also occur along dimensions that are harder to quantify, such as the language used when engaging with and evaluating members of a targeted group. Our project will: 1. develop an algorithm to measure of gender bias in appraisals, study the connection between bias in appraisals and women’s career outcomes, and 2. conduct a causal assessment of the existing bias and diversity training via Randomised Control Trial to establish effectiveness in BOTH reducing bias AND increasing diversity in the organisation. The algorithm will be developed on the basis of existing anonymised appraisals and evaluate along a number of dimensions including length and tone of text, specificity of comments, stereotyping, objectiveness, halo effects etc.Collaboration is underway with both La Caixa and EY to develop this project.