Social Research Matters – January 2019 update

Issues

The Changing Shape of Research Council Funding  

Research funding is going through a period of rapid change and recent paper from the ESRC aims to provide information on these changes in order to help researchers to navigate the new UKRI (UK Research and Innovation) funding streams, some of which will not be led by the ESRC. 

The piece makes the point that social science issues and methods are key to many of the emerging topics, but stresses that agility will be required by institutions and researchers in order to respond to relevant opportunities, which are likely to emerge at short notice and require working across disciplinary boundaries. A recent paper from Jennifer Rubin gives a flavour of this.

University Leaders warn of dangers of no-deal Brexit

University leaders have written to politicians and government spelling out the risks for research, staff and students if the UK leaves the EU without a deal. Bodies, representing more than 150 higher education providers across the UK, say the impact of a no deal scenario could lead to “an academic, cultural and scientific setback from which it would take decades to recover.” Large numbers of EU staff and students are starting the new year facing significant uncertainty about their futures and vital research links will be compromised. 

2021 Census (E & W) White Paper published 

The 2021 Census White Paper was published just before Christmas and sets out the recommendations made by the UK Statistics Authority. Following user engagement, they are recommending the same content as the 2011 Census with the exceptions of adding new questions on sexual orientation, gender identity and past service in the armed forces, and dropping of questions on the year in which the respondent last worked, and the number of rooms in the household.  

The census will be predominantly online but with a high level of assistance available to those who need it and paper forms for people who cannot complete the form online. 

New ONS Policies on Data and Surveys

Improved access to datasets, including those held by government bodies, will allow ONS and others to provide the UK with better statistics.  However, new data- sharing powers mean a heightened responsibility to protect personal information and ONS has published updated and revised policies on the use, management and security of data.

In a separate blog from ONS, Laura Wilson discusses how they are tackling the current challenges of undertaking social surveys in order to preserve their value and relevance. Measures discussed include: designing ‘mobile first’ and using behavioural insights to design respondent materials.

GDPR

The British Psychological Society has published new guidance on handling research data following the implementation of the EU General Data Protection Regulation (GDPR). S

 

Research Findings

Brexit and Public Opinion 2019

The UK appears to be ever more polarised on Brexit with ‘Brexit identities’ becoming stronger than party identities – e.g. whilst only one in 16 people do not have a Brexit identity (e.g. remain or leave), more than one in five say they had no party identity. This is the headline message in a major new report drawing in contributions from 34 authors. This significant polarisation, across the general population, and amongst MP’s, is exacerbated by the finding that new information about Brexit is interpreted in ways that reinforce pre-existing views.

Measuring Trends in Crime

Measuring the level of crime is complex and a recent release demonstrates the value of looking at data from different sources alongside each other in order to gain a balanced picture.

The Crime Survey for England and Wales (CSEW) measures people’s experiences of crime and is widely regarded as the best source on overall trends in crime. The latest data release shows  that whereas recent decades have seen falls in overall levels of crime, in the last year this fall seems to have stabilised.

Police data on recorded incidents has a number of limitations and is used to complement data from the crime survey. This and NHS data suggest that, although there has been no change in commonly occurring types of violent crime, there does appear to have been a rise in lower-volume, but higher-harm, violence that the crime survey does not capture well  e.g. knife crime.

The recent release also, for the first time, includes data based on the new method of calculating repeat victimisation. This new methodology has little impact on the long-term picture of total crime, though the number of incidents of violent crime increases slightly.

Productivity from a Workers Perspective

A recent report based on analysis of first findings from the 2017 Skills and Employment survey, suggests that a significant number of workers can identify changes that would make them more productive and efficient. The ‘productivity drivers’ identified included listening to staff. However, in the assessment of the authors, these practices have become less prevalent since 2006, precisely at a time when UK productivity growth has been sluggish and the economy would, potentially, have benefited most. 

Research Methods

Response Latencies as indicators of survey data quality

Survey methodologists are increasingly interested in how long it takes respondents to answer questions – the so-called ‘response latencies’. In cognitive psychology shorter latencies are taken as indicative of more strongly held attitudes. But it has also been argued that short latencies represent the amount of cognitive effort a respondent has expended in answering a question, with shorter response times indicating less effort and a lower quality response. The characteristics of the questions themselves will also have an impact and NCRM has been undertaking research into the joint influences of respondents, questions and interviewers.

The ONS Longitudinal Study

The ONS Longitudinal Study is perhaps the least well known of the longitudinal studies despite its  potential for undertaking research across the life course. It started in 1971 and involves linking census data (from 1971, 81, 91, 01 and 2011) on individuals in the sample with routine event registrations (e.g. deaths, cancer registrations). 

The size of the LS sample means  it is possible to study relatively small groups and has minimal bias due to non-response or attrition. The key disadvantages are the limited set of questions asked, changes in definitions/questions for several variables over time, lack of behavioural data and the fact that census updates only happen every 10 years.

Realist Evaluation

A recent book on ‘realist evaluation’ aims to support researchers to deal with the complexities involved in measuring impact, and to foster critical thinking and flexibility in making informed choices during the research process. 

Linear Regression Models using R

The Pew Research Centre has produced a short guide on how to estimate and interpret linear regression models with survey data using R.

  

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s