Combining Primary and Secondary Data: Unlocking the Potential of Secondary Data

Intrepid Researcher Seminar at Leicester University:


"Abstract

Over the course of the last decade the amount of secondary data available to researchers has increased considerably and at the same time has become much easier to access. While some groups of researchers have taken advantage of these resources and used them to great effect, it is often the case that this work has been concentrated among 'specialists' who work exclusively with data of this kind. In contrast, large numbers of researchers make very little use of secondary data in their work and often collect primary data in areas where comparable - but higher quality - secondary data already exist. This imbalance in the use of secondary data means that some data sets are being under-used - or being used only for very specific purposes - and that many researchers are missing opportunities to incorporate secondary data into their research.

 

This session provides an overview of the data sources available to academics and demonstrates how different types of data can be incorporated into research regardless of its substantive focus or research design. Particular attention is paid to how secondary data can be combined with primary data in order to produce high quality research outputs.

Speaker

Dr Emma Smith (School of Education, The University of Birmingham)

Emma Smith is Reader in the School of Education, University of Birmingham. Her research interests are concerned with equity issues in the field of education and in the role that educational policy can play in reducing inequalities and closing achievement gaps. So far her research has included boys' underachievement, barriers to participation in Higher Education, pupils' views of fairness and equity in school, formative assessment, teacher education and the impact of No Child Left Behind in the USA. One of her current projects is an ESRC funded study into patterns of participation in post-compulsory science programmes. She is the author of Using Secondary Data in Educational and Social research (Open University Press, 2008)."


....My notes from the lecture:

Secondary data is not used as much as it could be  - some research Elizabeth did on publications showed very low use. Sometimes this poses a problem for doctoral students that secondary data is considered part of the literature review, but actually it can be used effectively to add value as part of the research in contextualising the study. NB you need t be clear at the start what sort of analysis you want to  do and decide if the data is robust enough to warrant use.

Good to raise awareness of the richness of data that’s already out there!

Sources: large datasets:

-         Very useful for combined approaches, to contextualise your work. (Examples of research with excluded young people, & PhD researching home educated - to give context, and add questions to ONS survey.

CASWEB

NOMIS (UK labour force)

ONS (neighbourhood statistics – useful for small samples)

OECD (society at a glance publications, quite up to date and cover a period of time – can help aggregate and access to data)

ONS (social trends e.g. transport, holidays, TV viewing)

ESDS (Nesstar facility will analyse data for you. Lots of national and international datasets)

New developments:

Data.gov.uk includes blog, wiki, move to open access of government data

Wikiprogress.org (OECD) looks beyond gdp as a measure of wellbeing – different dimensions of wellbeing and aims to provide an index of wellbeing factors.

Gapminder.org for visualising data. Can run as an animation over time and change axes, also download software to map your own datasets

Numeric secondary data: individual level data –

e.g.  UK census SARS (need permission)

labour force survey

national pupil database

1970 birth cohort study

world values survey (online access tools including happiness, citizenship, etc)

pisa2009 (international student assessment including questionnaires for schools, parents, students – e.g. includes ICT)

Benefits of secondary data:

Contextualising study

High quality (of course still potential for errors)

Help understand who is and isn’t participating in a study in your area  and therefore address bias in research and inform you how to develop your sample and research questions.

Can answer the ‘what’ ie. Patterns but not necessarily the ‘why’.

Issues: ethical issues of using data for a different function to the original survey purpose.

Getting specific questions answered:

ONS opinions survey happens quickly as sweeps of 1800 households at random. Can pay to put a question in. Lots of data form the survey can combine and enhance a small scale study.

UK data archive is a good starting point – search key terms.

Question bank (part of the research methods hub at Southampton / national research methods council) – good to search for research question examples

UK research data service – Leicester Uni. is part of pilot for registering small scale data sets.

ESRC now requests all funded data to be added to archives.