UCL-IOE Q-Step Centre - Short Courses in Quantitative Methods

The Department of Quantitative Social Science at the Institute of Education (IOE), London, offered four short courses open to undergraduates interested in extending their knowledge of quantitative methods in the social sciences.

Each course was free to attend and encouragingly, the undergraduates who attended came from universitites that do not have a Q-Step Centre, thus sharing the exerptise and resources beyond the 15 Centres

The Courses:

  1. Survey design

    With several months to go before the referendum, how many Scots support independence from the UK? How much help do parents really give their children in the labour market? Addressing questions such as these requires good survey data – true of research on a vast range of issues in the social sciences. Even if you have no plans to conduct your own survey, to interpret existing data critically you need to understand how they are collected. This course will draw on the long experience we have of running three of the world-renowned British birth cohorts in our Centre for Longitudinal Studies(CLS). These studies follow large samples of people through their entire lives. You will learn more about different modes of data collection (face to face, phone, and web), the design of survey questions, response to surveys and how to try to maximize it, measurement error, and recall bias
  2. Stata computing skills for research projects

    You want to do some serious data analysis for a 3rd year dissertation? You would like to hone your computing skills to improve your position in the job market? There is a big difference between using a statistics package to work through exercises that support a lecture course and using it to do some real research. This course focuses on the latter. It will provide you with training in skills for work on your undergraduate dissertation, for further study, or for jobs after university – the skills covered are highly valued in the labour market. The course will focus on Stata, a leading computer package used in the social sciences. But the skills will be transferable to other packages too. The course will focus on (i) writing programmes (‘syntax’ files) of commands that can be edited/revised and that provide an ‘audit trail’ rather than using ‘point and click’, (ii) reading/handling datasets, (iii) some of our favourite simple Stata procedures that we find really useful in our own research
  3. Impact evaluation – measuring the effect of policy

    How many more people are in work because of a policy to help those on the margins of the labour market? Does including a free pen in a mail shot increase the donations that a charity receives?Correlation is not equal to causation – as any introductory quantitative methods course will tell you. The methods to identify the causal impact of policies of governments, firms, or NGOs, are a key element of quantitative methods training. The last 20 years have seen a revolution in the tools used by social scientists to identify causal impacts of policies. But this material is still rarely covered at the undergraduate level. A basic understanding of these tools is easily within your reach as the main techniques are not mathematically challenging. We will introduce you to pros and cons of randomized controlled trials – now rapidly gaining ground in the social sciences – and some of the ‘quasi experimental’ methods that are also heavily used
  4. Longitudinal analysis

    How much social mobility is there in Britain today, allowing people from disadvantaged backgrounds to climb up the socio-economic ladder? What is the impact of bullying in childhood on later life outcomes? Answering these sorts of questions requires analysis of ‘longitudinal data’ – repeated observation over time on the same individuals, families, or firms. Specifically, we will introduce you to using the birth cohorts run in CLS that contain a wealth of social science data. Longitudinal data are at the heart of many exciting developments in research in the social sciences as they have the potential to shed light on many issues of key policy interest. We will show you how to obtain the birth cohort data and how to start using them to address simple research questions. You will learn and practice methods to start realisingthe potential of longitudinal data, whether simple cross-tabulations with a time dimension (‘transition matrices’) through to regression models that exploit the fact that you have repeat observation over time in order to allow for unobservable factors.

See also

UCL-IOE Q-Step Centre Coordinator:

Dr Jennifer van Heerde-Hudson (UCL)

UCL-IOE Q-Step Centre Deputy Coordinators:

Professor John Micklewright (IOE)

Dr Slava Mikhaylov (UCL)