Brendan Halpin

Optimal matching analysis and life course data: the importance of duration

The optimal matching (OM) algorithm is widely used for sequence analysis in sociology. It has a natural interpretation for discrete-time sequences but is also widely used for life-history data, which are continuous in time. Life-history data are arguably better dealt with in terms of episodes rather than as strings of time-unit observations, and in this article, the author examines whether the OM algorithm is unsuitable for such sequences. A modified version of the algorithm is proposed, weighting OM’s elementary operations inversely with episode length. In the general case, the modified algorithm produces pairwise distances much lower than the standard algorithm, the more the sequences are composed of long spells in the same state. However, where all the sequences in a data set consist of few long spells, and there is low variability in the number of spells, the modified algorithm generates an overall pattern of distances that is not very different from standard OM.

Unified BHPS work-life histories: combining multiple sources into a user-friendly format

Longitudinal data is often difficult to use, and continuous histories collected in a panel are a particularly unfriendly case. This paper reports an exercise to re-organise the British Household Panel Studys work-life history data into a format more convenient for analysis. The British Household Panel Study collects extensive labour market history information from its respondents, both during the panel period and retrospectively from labour market entry. That this information is of necessity stored in multiple locations, and of varying levels of detail, has made use somewhat inconvenient. This paper describes an exercise to bring the labour market information together in a more convenient format. It also considers some of the problems of retrospective and panel longitudinal data, and discusses issues of recall error and measurement error. The data files described are available through the UK Data Archive. Longitudinal Data, Work-Life Histories, Recall Bias, Measurement Error.

Union Dissolution in the United Kingdom

Abstract: This article uses recent panel data to explore the dynamics of divorce in the United Kingdom. The findings are consistent with the independence hypothesis, but reveal little impact of gender-role attitudes or domestic division of labor. The article also finds a robust effect of children in raising the risk of divorce.

Simulation in Sociology

Simulation has a long and checkered history in areas of substantive interest to sociology, from before Forresters model of overpopulation to up-to-the-minute approaches based on complexity theory or distributed artificial intelligence. Although in some respects it has failed to live up to its inflated promise, it offers nonetheless a very useful paradigm. Moreover, advancing simulation technology offers some advantages, particularly the modelling of macro-micro links too complex to deal with linguistically or mathematically. This article briefly reviews the history of simulation in sociology and goes on to consider in more detail specific areas such as system dynamics, cellular automata, iterated game theory, distributed artificial intelligence, neural networks, multilevel simulation, simulation of social networks and organizations, and policy-oriented tax-benefit microsimulation. It concludes with a consideration of the role of statistics in simulation and the very good potential for expanded use of simulation in sociology.

Who marries whom in Great Britain

We investigate educational assortative mating, or homogamy, by modelling the hazard of entry to first marriage for a sample of residents of Great Britain. Using marital, and imputed educational, life-histories drawn from the British Household Panel Study, we estimate first a set of competing-risk models where the outcome variable is defined as respectively hypogamy, homogamy and hypergamy. Age, educational participation and cohort have very strong effects, in directions that may be expected. Time gap between leaving education and marrying shows some signs of the hypothesised effect (that a greater time-lag means less homogamy). Cohort differences suggest men are decreasingly likely to marry down and women decreasingly likely to marry up. We also fit a second set of competing-risk models, where the outcomes are marriage to individuals of specific educational levels. We argue that these are more stable models, and that they provide more insight into the actual marriage patterns. In both sets of models we attempt to control for the changing opportunity structure by including estimates of the educational distribution of single people in an appropriate age range. While this is obviously a necessary control variable, and while it has strong effects on other covariates, its own parameter estimates are hard to interpret. We cannot therefore claim that the effect of changing marginal distributions has been fully removed, but we feel our estimates are nonetheless improved.

Three narratives of sequence analysis

How do we relate the distance between two sequences, as given by an algorithm such as optimal matching, to sociologically meaningful notions of similarity and dissimilarity? This has been controversial in sequence analysis. Attention must be paid to how the algorithm operates, and to what sort of distances it generates in empirical practice. We can think of algorithms as giving distinct “narratives” of similarity, derived either formally or heuristically from their operation. This paper compares such narratives for several measures, including the optimal matching algorithm, a duration-weighted combinatorial subsequence algorithm and a time-warping algorithm. The algorithms have different ways of identifying similarity and of accounting for similarity displaced in time. Optimal matching has a narrative of string-editing and alignment, and is best adapted for discrete-time sequences; combinatorial methods focus on common order and produce radically different dissimilarities, differentiating strongly between simple and complex sequences, and treat time as sequential but not scaled; the time-warping algorithm has a narrative of locally warping the time axis, and while it is structurally similar to OM, it can be thought of as more suited to processes in continuous time. Moreover, because the time-warping parameters can be varied over a wide range, it provides a bridge between algorithms like OM, for which time is a (flexible) scale, and combinatorial subsequence algorithms for which time is scale-less order. With this and its continuous-time foundation, time-warping offers a real alternative to OM for lifecourse sequences.

Children, Standard of Living and Distributions in the Family

The standard of living of children depends on (i) family income; (ii) production within the family; (iii) economies of scale within the family; and (iv) distributions within the family. We use 1976 and 1986 Family Expenditure Survey information on (i), models of housework time based on time-budget surveys to estimate (ii), and equivalence scales to take account of (iii).

A Spatial Economic Perspective on Higher Education Choices

This chapter considers the spatial economics of higher education participation in Ireland, focussing on a range of spatial factors that impact the choices of school leavers. It employs cluster analysis and innovative data visualisation techniques to identify and illustrate important school-to-higher-education transition patterns. It also utilises GIS techniques to identify the extent of inequalities in geographic accessibility to higher education. The implications of these inequalities for higher education participation decisions are discussed with reference to a number of recent Irish studies that have considered the spatial economics of higher education participation. Particular attention is paid to the current restructuring of the higher education system in Ireland and to the student maintenance grant system.