Methods and machine learning

MiSoC’s methodology work covers the following areas:

• Machine learning techniques from computer science for the analysis of social science data
• Genomic, epigenetic and proteomic data for understanding the relationship between education and mental health
• Genomic databases for understanding the causes of historical demographic trends

Research area leader: Paul Clarke

We are currently working on

• Machine learning in the social sciences
• Measuring subjective expectations
• Covid-19 health risks and ambiguity

Click on the links above to find out more.

Recent Publications

• Double Machine Learning with High-dimensional Interactive Fixed Effects – Binzhi Chen, Annalivia Polselli, Paul S. Clarke
• Double Machine Learning for Static Panel Data with Instrumental Variables: New Method and Applications – Binzhi Chen, Annalivia Polselli, Paul S. Clarke

• Double machine learning with high-dimensional interactive fixed effects -
  1. Binzhi Chen
  2. Annalivia Polselli
  3. Paul S. Clarke
• Analysis of household effects on longitudinal health outcomes using a joint mean-correlation multilevel model with grouped random effects -
  1. Fiona Steele
  2. Siliang Zhang
  3. Paul S. Clarke
• Double machine learning for static panel data with instrumental variables: new method and applications -
  1. Anna Baiardi
  2. Paul S. Clarke
  3. Andrea A. Naghi
  4. Annalivia Polselli
• Double machine learning for static panel models with fixed effects -
  1. Paul S. Clarke
  2. Annalivia Polselli
• Gender segregation: analysis across sectoral dominance in the UK labour market -
  1. Riccardo Leoncini
  2. Mariele Macaluso
  3. Annalivia Polselli
• Robustness of algorithms for causal structure learning to hyperparameter choice -
  1. Damian Machlanski
  2. Spyridon Samothrakis
  3. Paul S. Clarke
• Adaptive social protection in Indonesia: stress-testing the effect of a natural disaster on poverty and vulnerability -
  1. Katrin Gasior
  2. Gemma Wright
  3. Helen Barnes
  4. Michael Noble
• Undersmoothing causal estimators with generative trees -
  1. Damian Machlanski
  2. Spyridon Samothrakis
  3. Paul S. Clarke
• Policy uncertainty and information flows: evidence from pension reform expectations -
  1. Emanuele Ciani
  2. Adeline Delavande
  3. Ben Etheridge
  4. Marco Francesconi
• Academic and non-academic investments at university: the role of expectations, preferences and constraints -
  1. Adeline Delavande
  2. Emilia Del Bono
  3. Angus Holford

More Publications

Explainers

• Automation will continue to displace workers – but they won’t easily fill healthcare and education shortages -
  1. Michael J. Böhm
  2. Ben Etheridge
  3. Aitor Irastorza-Fadrique
• A new method to detect anomalous units with panel data and quantify their influence on the results -
  1. Annalivia Polselli
• Testing policy interventions using machine learning and panel data: the Double Machine Learning method for static panel models with fixed effects -
  1. Paul Clarke
  2. Annalivia Polselli
• What should we do about inter-survey variability in survey data harmonisation? A survey data recycling approach -
  1. Joonghyun Kwak
• University access: the role of background and Covid-19 throughout the application process -
  1. Emilia Del Bono
  2. Laura Fumagalli
  3. Angus Holford
  4. Birgitta Rabe

More MiSoc Explainers

Blog

• How does early socio-emotional development in childhood relate to later offending? -
  Paul Garcia Hinojosa
• Understanding the use of genetics in social science research -
  Anna Dearman
• Designing more robust experiments to improve education research -
  Joshua Fullard
• The value of keeping ethnic ties: why adherence to recommended COVID-19 health behaviours differs among young adults -
  Renee Luthra, Angus Holford, Adeline Delavande
• Are there ethnic differences in adherence to recommended health behaviours related to Covid-19? -
  Renee Luthra
• Hard hit: Covid-19 has changed many lives, but ethnic minorities and women, especially lone mothers, have been particularly affected. -
  Emilia Del Bono
• Income inequality and the very rich -
  Mike Brewer
• Subjective Expectations: Measurement and analysis of decision-making under uncertainty -
  Adeline Delavande
• UKMOD: a new, free-to-use, tax-and-benefit microsimulation model for the UK -
  Mike Brewer

More Blog Posts

Contact

ESRC Research Centre on Micro-Social Change (MiSoC)
Institute for Social and Economic Research
University of Essex
Wivenhoe Park
Colchester
Essex
CO4 3SQ
UK

Email: iser@essex.ac.uk
Follow us on X: @MiSoC_Essex