LEMMA 1 research projects

Lemma logo Methodological and substantive research was conducted on the following themes:

 Growth trajectory or mixture models

The usual distributional assumption of normality for higher level random effects is over-restrictive. Nagin (1999: Psychological Methods, 4: 139-177) describes a formulation for growth curves where a discrete set of latent groups is posited and each individual has a membership distribution across the groups. Separate growth trajectories are modelled for each group.

Group-based latent trajectory models were used in the analysis of voting trends at the constituency level between 1950 and 2001. The same methodology was also used to model trends in global life expectancy and peri-natal maternal depression.  Outputs include Jen et al. (2009a; 2009b; 2010) and Johnston et al. (2007; 2009).

Publications

(Back to top)

Mental health and psychosocial development

This project was concerned with the development and application of models to child development and family relationship data, accounting for complex family structures and processes.  The work included extension of the social relations model to include genetic effects, and was in tandem with an ESRC Research Methods project ( Methodologies for Studying Families and Family Effects). Another project examined the impact of family disruption on children’s educational careers, using a simultaneous equations model to allow for selection effects. Outputs include Jenkins et al. (2005a), Jenkins et al. (2005b), O’Connor et al. (2006), Rasbash et al. (in press),  Steele et al. (2009), and Propper et al. (2005).

Publications

(Back to top)

Modelling school effects

Projects include: a study of pupil progress over time, disentangling pupil, school and neighbourhood effects; an examination of pupil residency, location and the related issues of school mobility and selection; a sensitivity analysis of the impact of missing data on estimates of school effects; and a study of the impact of school resources on pupil attainment.  Outputs include Leckie (2009), Goldstein et al. (2007), Peng et al. (2006), Steele et al. (2007), Thomas et al. (2007) and Leckie’s PhD thesis. Related work on the statistical limitations of the government’s school league tables (Leckie and Goldstein, 2009a, b) also appeared in the RSS magazine Significance (Goldstein and Leckie, 2008) are received considerable media coverage: BBC Radio 4, FT, Telegraph and TES.  This work was also disseminated to education policy makers at the ESRC-funded PLASC/NPD Users' Group (PLUG).

Publications

(Back to top)

Voting choice

This project considered the individual, household and neighbourhood determinants of voting abstention and party choice. The British Household Panel Study provides repeated binary measures on voting intention for individuals within households within areas at a variety of scales. Normally distributed individual level random effects are unrealistic as part of the mover/stayer problem, as are normally distributed household effects. Outputs include Johnston et al. (2007; 2009).

Publications

Edit this page