Department seminar speaker

Events

Departmental and Data Science seminar series

Termly events held by the Department of Methodology and the Social and Economic Data Science (SEDS) research unit.

 

Leading social scientists consider cutting-edge quantitative and qualitative methodologies, analyse the logic underpinning an array of approaches to empirical enquiry, and discuss the practicalities of carrying out research in a variety of different contexts.

Department of Methodology Seminar Series 

Department seminars take place in COL 8.13 from 12:30 - 14:00 unless otherwise stated. A sandwich lunch will be provided from 12:15.


JackBlumenau

Does online activism affect legislative behaviour?
Date: 
30 November 2017
Speaker: Jack Blumenau, UCL.
Abstract: Are legislators responsive to the issue priorities of their constituents? Voters in the UK are able to sign government e-petitions which provide information regarding the salience of different issues across constituencies. E-petitions constitute an explicit mechanism to try to gauge public interest but we know little about their effects on the parliamentary behaviour of MPs. In this paper, I provide quantitative evidence of the effectiveness of the UK government e-petition system. I examine whether varying levels of support for different e-petitions in a given constituency affect an MP's propensity to participate in parliamentary debates relating to those petitions, and their propensity to speak in favour of the proposals contained in the petitions. The results suggest that while local support for an e-petition can affect an MP's behaviour in parliament, these effects are conditional. MPs are more likely to turn up to debate a petition, and speak in favour of it in the debate, when the petition is not strongly linked to the main dimension of partisan competition. When the petition addresses a pre-exiting and politically-salient issue, signature rates have no effect on MP behaviour. Finally, these effects are stronger for MPs in electorally competitive constituencies, and weaker for MPs who hold government positions.

 

Social and Economic Data Science Seminars

Data Science seminars returns in Michaelmas Term 2017 and are run in conjunction with the Departments of Mathematics and Statistics. These seminars take place from 16:15 - 17:45 in COL 8.13 (unless otherwise stated) with drinks and snacks at the end.

AIseminar

Controlling Bias in Artificial Intelligence with Nisheeth Vishoi and Elisa Celis
Date: Wednesday 29 November 2017
Time: 3:30 - 5pm **not the usual timeslot**
Speakers: Nisheeth Vishoi and Elisa Celis
Abstract: Bias is an increasingly observed phenomenon in the world of artificial intelligence (AI) and machine learning: From gender bias in online search to racial bias in court bail pleas to biases in worldviews depicted in personalized newsfeeds. How are societal biases creeping into the seemingly “objective’’ world of computers and programs? At the core, what is powering today’s AI are algorithms for fundamental computational problems such as classification, data summarization, and online learning. Such algorithms have traditionally been designed with the goal of maximizing some notion of “utility” and identifying or controlling bias in their output has not been a consideration. In this talk, Nisheeth and Elisa will  explain the emergence of bias in algorithmic decision making and present the first steps towards developing a systematic framework to control biases in several of the aforementioned problems. This leads to new algorithms that have the ability to control and alleviate bias, often without a significant compromise to the utility that the current algorithms obtain.

jeff-gill-300x200

A Measure of Survey Mode Differences
Date: 
Tuesday 5 December 2017
Speaker: Jeff Gill, Department of Government, American University, Washington DC
Abstract: Jeff will evaluate the effects of different survey modes on respondents' patterns of answers using an entropy measure of variability. While \emph{measures of centrality} show little differences between face-to-face and Internet surveys, he will find strong patterns of \emph{distributional differences} between these modes where Internet responses tend towards more diffuse positions due to lack of personal contact during the process and the social forces provided by that format. The results provide clear evidence that mode matters in modern survey research, and he will make recommendations for interpreting results from different modes.