What can Statistics Contribute to a Causal Understanding?

Abstract.  We discuss the concept of causality in a broad manner. The traditional attitude in statistics has been to shy away from the causality concept. In recent years, however, a more proactive attitude to the causality concept has developed among statisticians. On the one hand, there is the school of counterfactual causality directed towards forming a better basis for clinical trials and epidemiology. On the other hand, there is the mechanistic view of causality. These developments are discussed and set into a larger context, where the often ignored role of time is emphasized. An extension of path analysis to stochastic processes is briefly presented. Causality is furthermore discussed in relation to genetic studies and to the emerging field of systems biology. Statisticians should also relate to the philosophical content of causality, especially that found in the foundations of physics.     

深层水泥搅拌桩加固软土地基的设计与施工

概述深层水泥搅拌桩在加固软土地基中有关单桩承载力的确定、置换率与复合地基承载力的关系、桩长与沉降量的关系以及深层水泥搅拌桩在施工中的一些经验体会．     

Evaluation of Atmospheric Transport Models for Use in Phase II of the Historical Public Exposures Studies at the Rocky Flats Plant

Five atmospheric transport models were evaluated for use in Phase II of the Historical Public Exposures Studies at the Rocky Flats Plant. Models included a simple straight-line Gaussian plume model (ISCST2), several integrated puff models (RATCHET, TRIAD, and INPUFF2), and a complex terrain model (TRAC). Evaluations were based on how well model predictions compared with sulfur hexafluoride tracer measurements taken in the vicinity of Rocky Flats in February 1991. Twelve separate tracer experiments were conducted, each lasting 9 hr and measured at 140 samplers in arcs 8 and 16 km from the release point at Rocky Flats. Four modeling objectives were defined based on the endpoints of the overall study: (1) the unpaired maximum hourly average concentration, (2) paired time-averaged concentration, (3) unpaired time-averaged concentration, and (4) arc-integrated concentration. Performance measures were used to evaluate models and focused on the geometric mean and standard deviation of the predicted-to-observed ratio and the correlation coefficient between predicted and observed concentrations. No one model consistently outperformed the others in all modeling objectives and performance measures. About 75% of the maximum hourly concentration predictions were within a factor of 5 of the observations. About 64% of the paired and 80% of the unpaired time-averaged model predictions were within a factor of 5 of the observations. The overall performance of the RATCHET model was somewhat better than the other models. All models appeared to experience difficulty defining plume trajectories, which was attributed to the influence of multilayered flow initiated by terrain complexities and the diurnal flow patterns characteristic of the Colorado Front Range.     

The Analysis of Designed Experiments and Longitudinal Data by Using Smoothing Splines

In designed experiments and in particular longitudinal studies, the aim may be to assess the effect of a quantitative variable such as time on treatment effects. Modelling treatment effects can be complex in the presence of other sources of variation. Three examples are presented to illustrate an approach to analysis in such cases. The first example is a longitudinal experiment on the growth of cows under a factorial treatment structure where serial correlation and variance heterogeneity complicate the analysis. The second example involves the calibration of optical density and the concentration of a protein DNase in the presence of sampling variation and variance heterogeneity. The final example is a multienvironment agricultural field experiment in which a yield–seeding rate relationship is required for several varieties of lupins. Spatial variation within environments, heterogeneity between environments and variation between varieties all need to be incorporated in the analysis. In this paper, the cubic smoothing spline is used in conjunction with fixed and random effects, random coefficients and variance modelling to provide simultaneous modelling of trends and covariance structure. The key result that allows coherent and flexible empirical model building in complex situations is the linear mixed model representation of the cubic smoothing spline. An extension is proposed in which trend is partitioned into smooth and non-smooth components. Estimation and inference, the analysis of the three examples and a discussion of extensions and unresolved issues are also presented.     

A comparison of population average and random-effect models for the analysis of longitudinal count data with base-line information

The generalized estimating equation (GEE) approach to the analysis of longitudinal data has many attractive robustness properties and can provide a 'population average' characterization of interest, for example, to clinicians who have to treat patients on the basis of their observed characteristics. However, these methods have limitations which restrict their usefulness in both the social and the medical sciences. This conclusion is based on the premise that the main motivations for longitudinal analysis are insight into microlevel dynamics and improved control for omitted or unmeasured variables. We claim that to address these issues a properly formulated random-effects model is required. In addition to a theoretical assessment of some of the issues, we illustrate this by reanalysing data on polyp counts. In this example, the covariates include a base-line outcome, and the effectiveness of the treatment seems to vary by base-line. We compare the random-effects approach with the GEE approach and conclude that the GEE approach is inappropriate for assessing the treatment effects for these data.     

A multilevel exploration of the role of interviewers in survey non-response

This paper illustrates the use of multilevel statistical modelling of cross-classified data to explore interviewers' influence on survey non-response. The results suggest that the variability in whole household refusal and non-contact rates is due more to the influence of interviewers than to the influence of areas. The results from separate logistic regression models are compared with the results from multinomial models using a polytomous dependent variable (refusals, non-contacts and responses). Using the cross-classified multilevel approach allows us to estimate correlations between refusals and non-contacts, suggesting that interviewers who are good at reducing whole household refusals are also good at reducing whole household non-contacts.     

Specification analysis in mixed hazard models and a test of crossing survival functions

The paper investigates diagnostic procedures for the specification of common hazard models in duration analysis. It is shown that under mixed hazard specifications the survival functions of different subgroups cannot cross. A nonparametric test for the crossing of two survival functions is provided and its applications in duration analysis are discussed. In particular, the proportional hazard model with unobserved heterogeneity (PHU) is investigated, and procedures are developed to test whether given data are consistent with the PHU model and whether they contain unobserved heterogeneity within the PHU specification. Examples in which crossing survivals are of substantive concern are discussed, including the dynamics of infectious diseases and the demand for vaccination.     

Limited opportunities: Adolescents’ access to social capital in secondary schools in three European countries

While previous research indicates that students benefit from their peers’ resources, little is known about access to social capital in the school context. Therefore, this study examines differential access to social capital – measured by friends’ socioeconomic status (SES), the number of books they have at home, and their reading habits – in secondary schools in Germany, the Netherlands, and Sweden. Relying on a large-scale dataset, I investigate the association between socioeconomic status, minority status, and social capital using complete friendship network information. I argue that social capital access is connected to a two-stage process consisting of school sorting and friendship selection. To differentiate between these two processes, I apply within-between random effects (REWB). The models show that friendship selection is much less relevant for access to social capital than school sorting. Results indicate that while high-SES students have better access to social capital across dimensions, access patterns for minority students are more nuanced.     

‘A perfect fit’ – Swedish midwives’ interest in continuity models of midwifery care

BackgroundMidwifery continuity models of care are highly recommended yet rare in Sweden, although approximately 50% of pregnant women request them. Before introducing and scaling up continuity models in Sweden, midwives’ attitudes about working in continuity models must be investigated.Objectiveto investigate Swedish midwives’ interests in working in midwifery continuity models of care and factors influencing the midwifery workforce’s readiness for such models.MethodsA cross-sectional online survey was utilised and information collected from a national sample of midwives recruited from two unions regarding background and work-related variables. Crude and adjusted odds ratios and logistic regression analysis were used in the analysis.ResultsA total of 2084 midwives responded and 56.1% reported an interest. The logistic regression model showed that respondents’ ages 24–35 years (OR 1.73) or 35–45 years (OR 1.46); years of work experience 0–3 years (OR 5.81) and 3–10 years (OR 2.04); rotating between wards or between tasks (OR 2.02) and working temporary (OR 1.99) were related to interest in continuity models. In addition, working daytime only (OR 1.59) or on a two-shift schedule (OR 1.93) was associated with such interest.ConclusionA sufficient number of midwives in Sweden appear to be interested in working in continuity models of midwifery care to align with women’s interest in having a known midwife throughout pregnancy, birth and postpartum period. Developing strategies and continuity models that will address the preferences of women in various areas of Sweden is important for offering evidence-based maternity services.     

Supporting the midwifery workforce: An evaluation of an undergraduate midwifery student employment model at a large tertiary maternity service in Victoria,Australia

BackgroundIn 2020, in response to major maternity workforce challenges exacerbated by the COVID-19 pandemic, the Victorian Department of Health implemented a number of workforce maximisation strategies, one of which was employment of undergraduate midwifery students called ‘Registered Undergraduate Student Of Midwifery’ (RUSOM).AimTo evaluate the RUSOM model implemented in a tertiary maternity service in Melbourne, Australia.MethodsA cross-sectional online survey was distributed to all RUSOMs and midwives at the study site in August 2021.FindingsTwenty of 26 RUSOMs (77%) and 110 of 338 permanent midwives (33%) responded. Both groups considered the model to be a positive workforce strategy that contributed to work readiness of students, and increased confidence and competence to practise. RUSOMS and midwives reported positives for the organisation including improving workload for midwives on the postnatal ward, enhancing quality of care and outcomes for women and babies, and the value of RUSOMs as team members. RUSOMs felt well supported, supervised and clinically and theoretically prepared. Both groups considered RUSOMs were underutilised, and that they could undertake additional duties, and both thought that the RUSOM model should continue.ConclusionThe model was highly valued by both RUSOMs and midwives. There was strong agreement that the model should continue and that the list of duties could be expanded. Given these findings, further research should explore the expansion and sustainability of RUSOMs in the maternity workforce.