Wednesday, September 17, 2014

September 19th ESRC Workshop on Early Life Influences on Later Life Health and Economic Outcomes

ESRC Workshop 3: Early Life Influences on Later Life Health and Economic Outcomes (19/9/14)

This is the third Behavioural Science Workshop in a series of six that will take place in 2014/15. These workshops are funded by the Economic and Social Research Council. The venue is the Court Room on the 4th Floor of the Cottrell Building at Stirling University. There will be drinks and dinner after the days talks to which all attendees are welcome.

This workshop will address the ESRC’s priority objective of fostering research that capitalizes on the growing data resources available in the UK Data Archive and comparable international depositories. There is now an abundance of large databases, which assess detailed psychological, economic, and health measures in samples of tens of thousands of participants over several years or even decades.

The theme of the workshop is how the measurement of constructs such as childhood personality, well-being, intelligence and adverse conditions in early life can be used to understand the unfolding of economic, health and welfare outcomes throughout adulthood. The comparative benefits of contemporary measurement and retrospective accounts of early conditions will be addressed, studies utilizing this data presented, and limitations (e.g. recall, desirability biases) discussed.

Sign up to the workshop here


09:00-09:20: Registration

09:20-09:30: Welcome and workshop introduction

09:30-10:00: Fionnuala O'Reilly (Stirling University)
Associations between childhood self-regulation and adult socioeconomic status
Abstract:Uncovering the childhood determinants of socioeconomic status (SES) in adulthood is an important social goal. In this paper, we utilised the British Cohort Study (N = 6,700) to examine the association between childhood self-regulation and a set of socioeconomic factors measured in adulthood, adjusting for a range of important potential confounding variables including childhood cognitive ability and parental SES. Specifically, we tested the association between self-regulation at age 10 and the cohort members' income, social class, educational attainment, home ownership and self-ratings of their financial position at age 30 and 42.

We found that higher self-regulation at age 10 had a substantial and significant association with better SES outcomes at both age 30 and 42. On average a 1 SD increase in childhood self-regulation was associated with a 0.13 SD increase in adult SES; an effect size comparable to that of a 1 SD increase in childhood cognitive ability (0.17 SD). On average 30% of the relationship between childhood self-regulation and adult SES was explained by educational attainment. Finally, we found that childhood self-regulation acts as a medium through which individuals may attain higher social standing, both inter-generationally and over the course of their own lives.

10:00-10:30: Dr. Michael Daly (Stirling University)
Poor childhood self-discipline predicts physiological dysregulation in midlife (with Liam Delaney).
Childhood self-discipline emerges early, is malleable, and could contribute substantially to a healthy life. The present study examined associations between self-discipline at ages 7 and 11 and physiological dysregulation in middle age. Participants were 6,878 British men and women born in March 1958 who took part in the National Child Development Study. Self-discipline was gauged using a 13-item teacher-rated scale from the Bristol Social Adjustment Guide assessing concentration (e.g. ‘cannot attend or concentrate for long’), perseverance (e.g. ‘can never stick at anything long’), restlessness and impulsive behaviour (e.g. ‘constantly needs petty correction’). Blood plasma samples and anthropometric data were collected and analysed using standard procedures at age 45. An overall physiological dysregulation index was derived from a set of 12 biological variables: systolic and diastolic blood pressure, HDL cholesterol, triglycerides, body mass index, waist/hip ratio, C-reactive protein, fibrinogen, Von Willebrand factor, glycosylated haemoglobin, tissue plasminogen activator, and peak flow (Cronbach’s α = .76).

Higher levels of self-discipline were significantly associated with lower physiological dysregulation (B = -.073, SE = .013; β = -.073; t = -5.80, p < .001), after controlling for sex, intelligence at age 11, and socioeconomic status at birth. This association was relatively unaffected  by further adjustment for a large set of childhood controls (B = -.068, SE = .017; β = -.068; t = -5.30, p < .001) including parental characteristics (e.g. age, mother’s education), family difficulties (e.g. housing, financial), aspects of the home environment (e.g. region, crowding), conditions at birth (e.g. birth weight, breast feeding), physician assessed medical conditions (e.g. asthma, emotional maladjustment, diabetes) and relative weight at age 7. By adjusting for a broad set of important covariates in a large-scale representative cohort these analyses provide robust evidence that childhood self-discipline is associated with long-run health effects that cannot be attributed to other psychological factors like intelligence or emotional problems or to initial health or environmental conditions.

10:30-11:00: COFFEE

11:00-11:45: Professor Markus Jokela (University of Helsinki)
Adolescent verbal ability and health outcomes in the British Household Panel Survey

11:45-12:30: Dr. Iris Kesternich (Munich)
Early-life circumstances predict measures of trust attitudes among adults
(with Maximiiana Hörl, Jim Smith & Joachim Winter).
Trust in strangers plays a decisive role in economic interactions, and at the same time it shows substantial heterogeneity across individuals. The sources of this individual-level variation are largely unknown. This paper investigates whether a major shock experienced in childhood can permanently shape preferences for trust. We relate a measure of trust in strangers available for a nationally representative sample of the German population to exposure to a hunger episode after the Second World War. We collected data on caloric rations that vary by month and across regions to capture exposure to hunger. We find that trust is significantly diminished for those more affected by the hunger episode.

12:30-13:30: LUNCH

13:30-14:15: Mark Egan (Stirling University)
Childhood psychological distress and youth unemployment: evidence from three cohort studies
(with Michael Daly & Liam Delaney)
The effect of childhood mental health on later unemployment has not yet been established. This presentation reviews recent work examining whether childhood psychological distress places young people at high risk of subsequent unemployment and whether the presence of economic recession strengthens this relationship. We investigate these relationships using three nationally-representative cohort studies - 19,377 individuals from the Longitudinal Study of Young People in England (LSYPE) and the National Child Development Study (NCDS) in Britain and 6,474 individuals from the National Longitudinal Study of Youth 1997 (NLSY97) in the United States - with a combined total of 3.8m observations. Distress was measured using the General Health Questionnaire at age 14 in the LSYPE, via a teacher-rated measure of depression at age 7 and 11 in the NCDS and with the Mental Health Inventory at age 16-20 in the NLSY97.

There are two main findings. Firstly, children with higher levels of distress went on to experience higher levels of youth unemployment in all cohorts examined. These effects were large, statistically significant and could not be accounted for by early environmental factors, intelligence, or personality characteristics. Secondly, analyses of the 1980 recession in the UK and the 2007 recession in the United States reveals that children with higher levels of distress were disproportionately more likely to become unemployed during the fallout of these economic downturns. These findings point to a previously neglected contribution of childhood mental health to youth unemployment which may be particularly pronounced during times of economic recession. Our findings also suggest a further economic benefit to enhancing the provision of mental health services early in life.

14:15-15:00: Dr. Jan-Emmanuel De Neve (UCL)
Estimating the influence of life satisfaction and positive affect on later income using sibling fixed-effects (with Andrew Oswald)
The question of whether there is a connection between income and psychological well-being is a long-studied issue across the social, psychological, and behavioral sciences. Much research has found that richer people tend to be happier. However, relatively little attention has been paid to whether happier individuals perform better financially in the first place. This possibility of reverse causality is arguably understudied. Using data from a large US representative panel, we show that adolescents and young adults who report higher life satisfaction or positive affect grow up to earn significantly higher levels of income later in life. We focus on earnings approximately one decade after the person’s well-being is measured; we exploit the availability of sibling clusters to introduce family fixed effects; we account for the human capacity to imagine later socioeconomic outcomes and to anticipate the resulting feelings in current well-being.

The study’s results are robust to the inclusion of controls such as education, intelligence quotient, physical health, height, self-esteem, and later happiness. We consider how psychological well-being may influence income. Sobel–Goodman mediation tests reveal direct and indirect effects that carry the influence from happiness to income. Significant mediating pathways include a higher probability of obtaining a college degree, getting hired and promoted, having higher degrees of optimism and extraversion, and less neuroticism.

15:00-15:30: COFFEE

15:30-16:15: Professor Alissa Goodman (Institute of Education)
Long-term effects of childhood mental and physical health conditions
In this presentation I assess and compare long-term adult socioeconomic status impacts from having experienced psychological and physical health problems in childhood. The research is based on prospective data from the British National Child Development Study, a longitudinal study of a cohort of 17,634 children born in Great Britain during a single week in March 1958. Large effects are found due to childhood psychological problems on the ability of affected children to work and earn as adults and on intergenerational and within generation social mobility. Effects of psychological health disorders during childhood are far more important over a lifetime than childhood physical health problems. There is a strong interrelationship between cognitive development and emotional and behavioural disorders in childhood, which in part explains the significance of childhood psychological problems in later life.

16:15-17:00: Panel discussion

Using Large Publicly Available Datasets for Psychological & Social Science Research

Would your research benefit from being able to analyse already-collected data on psychological measures from thousands of different individuals at multiple time-points? 

There are now many publicly available datasets within the UK (such as those hosted by the UK Data Service) and across the world. These data have been collected with the primary purpose of enabling researchers to better understand how people function within the world around them. Although these data-sets are free to access and are commonly used within economics and epidemiology, they remain under-utilized in many disciplines in the social sciences, particularly psychology. This is unfortunate given that many of these datasets contain measures and scales relevant to cutting-edge psychological research, such as personality, well-being, attitudes, behaviour, physical health and mental health. One barrier to unlocking these datasets' potential is having the necessary skills to manage and analyse them. We at the Behavioural Science Centre, Stirling Management School, in conjunction with the Economic Social and Research Council (ESRC), are offering 2 training workshops specifically built around these datasets to equip you with the necessary skills, which includes an introduction to the statistical package Stata, to handle them.

Workshop 1: Introduction to data analysis of large publicly available datasets (2-3 December 2014) 
Participants will learn how to obtain and manage data, use the statistical program Stata, conduct basic analysis and interpret the results. This workshop requires that participants are comfortable with basic statistics prior to the course and will enable researchers to begin using untapped resources immediately. 

Workshop 2: Advanced techniques for large publicly available datasets (18-19 March 2015) 
Participants will learn advanced statistical methodology to enable them to get the most out of large publicly available datasets. This will include panel data techniques such as understanding and implementing fixed effect and difference-in-difference models, as well as how to implement instrumental variable estimations. This workshop will require that participants have a basic knowledge of handling large datasets and using the statistical program Stata. Attendance at workshop 1, although not a pre-requisite, would represent sufficient knowledge. 

Further details: Both courses will take place at Stirling Management School, University of Stirling. At our Behavioural Science Centre we have a number of researchers, including Prof Liam Delaney, Dr Michael Daly (early Career Award recipient, UK Society for Behavioural Medicine), and Prof Alex Wood and Dr Christopher Boyce (joint winners, best paper using GSOEP data resource 2012-2013), with substantial experience using and publishing with these types of datasets. Both workshops are aimed at PhD students but advanced Masters students and post-PhD researchers are welcome to apply. The University of Stirling is approximately 50 minutes by train from Edinburgh, 25 minutes from Glasgow and 5 hours from London. The course is funded by the ESRC and the cost to participants is £50 (in addition to accommodation and transport).

How to apply: Participants are welcome to apply to attend one workshop or both. Please send a completed application form, including (1) a curriculum vitae, (2) a statement as to why you wish to attend and how it will benefit you and your research (suggested maximum 2 pages), and (3) a supporting statement from a supervisor or senior colleague, to

Tuesday, September 16, 2014

List of resources for the Stata commands 'margins' and 'marginsplot'

This is an ongoing list of resources for understanding the 'margins' and 'marginsplot' commands in Stata. If you know of links I have omitted, please let me know. 

The 'margins' and 'marginsplot' commands were introduced in Stata 11. They are very useful as a way of estimating and graphing predicted probabilities and are potentially much more informative than regression tables for certain kinds of analysis (e.g. when examining interaction effects, probit or logit regressions or any model which returns non-intuitive coefficients, like a negative binomial).

Despite their advantages, these commands are still somewhat underutilized, at least in terms of the papers I read. It may be the case that since margins is only a few years young, many researchers simply don't know about them or don't fully understand how useful they can be at clarifying regression results. Hopefully one side-benefit of this post will be to generate discussion from people who are using margins in their own research and/or flush out people who have valid objections to it that I may not be aware of.

1. Margins overviews:
- From Overview of margins & marginsplot.
- From UCLA: A brief overview of what margins can do & how to use it to examine interactions.
- From SSCC: Exploring Regression Results using Margins.
- The help file for margins in Stata 13.
- The help file for marginsplot in Stata 13.

2. I highly recommend Richard Williams's slides on "Using Stata’s Margins Command to Estimate and Interpret Adjusted Predictions and Marginal Effects" (2011). If you want to cite Williams, use his Stata Journal article on the same topic (2012).

3. Bill Rising's slides on "How to get an edge with margins and marginsplot" (2012).

4. Ben Jann's slides on "Predictive Margins and Marginal Effects in Stata" (2013).

Friday, September 12, 2014

Carnegie Trust Collaborative Research Grants

Details here and below

Aims and background

The purpose of the Collaborative Research Grants Scheme (CRGs) is to support joint research projects that bring together researchers from more than one Scottish university to develop new lines of study or to advance significantly existing areas of expertise.

The principal criterion for the award of a CRG is that the planned research is of excellent quality and is likely to be of benefit to two or more of the universities of Scotland (research consortium).
Grants of up to a maximum of £50,000 may be awarded.

The joint project must involve the active collaboration of researchers in at least two of the Scottish universities. Each university should specify a Principal Investigator, one of whom should act as the Lead Applicant.

The duration of the project should normally be from 1 to 2 years, but can be up to a maximum of three years.

The Trust is not a research council and will not consider applications that are judged more appropriate for submission to a research council or similar body.

Wednesday, September 10, 2014

Leverhulme Research Fellowships

Note: Our research centre hosts research fellows funded from various funding bodies and we are happy to discuss opportunities such as this below with potential applicants. 

Research Fellowships

Offering up to £50,000 over three to twenty-four months for experienced researchers to conduct a programme of research in any discipline.

Research Fellowships are open to experienced researchers, particularly those who are or have been prevented by routine duties from completing a programme of original research. There are no restrictions on academic discipline, and awards are not limited to those holding appointments in higher education.

The maximum value of a Fellowship is £50,000. The awards provide research expenses over and above normal living costs and/or provide a contribution towards reasonable replacement costs.

Fellowships are tenable for between 3 and 24 months, and the current round of awards must commence between 1 June 2015 and 1 May 2016.

Full details can be found here<>. Deadline 6 November 2014.