Why High-Order Polynomials Should Not Be Used in Regression Discontinuity Designs

It is common in regression discontinuity analysis to control for third, fourth, or higher-degree polynomials of the forcing variable. There appears to be a perception that such methods are theoretically justified, even though they can lead to evidently nonsensical results. We argue that controlling for global high-order polynomials in regression discontinuity analysis is a flawed approach with three major problems: it leads to noisy estimates, sensitivity to the degree of the polynomial, and poor coverage of confidence intervals. We recommend researchers instead use estimators based on local linear or quadratic polynomials or other smooth functions.     

Abandon Statistical Significance

ABSTRACT

We discuss problems the null hypothesis significance testing (NHST) paradigm poses for replication and more broadly in the biomedical and social sciences as well as how these problems remain unresolved by proposals involving modified p-value thresholds, confidence intervals, and Bayes factors. We then discuss our own proposal, which is to abandon statistical significance. We recommend dropping the NHST paradigm—and the p-value thresholds intrinsic to it—as the default statistical paradigm for research, publication, and discovery in the biomedical and social sciences. Specifically, we propose that the p-value be demoted from its threshold screening role and instead, treated continuously, be considered along with currently subordinate factors (e.g., related prior evidence, plausibility of mechanism, study design and data quality, real world costs and benefits, novelty of finding, and other factors that vary by research domain) as just one among many pieces of evidence. We have no desire to “ban” p-values or other purely statistical measures. Rather, we believe that such measures should not be thresholded and that, thresholded or not, they should not take priority over the currently subordinate factors. We also argue that it seldom makes sense to calibrate evidence as a function of p-values or other purely statistical measures. We offer recommendations for how our proposal can be implemented in the scientific publication process as well as in statistical decision making more broadly.     

Friendly foes under friendly fire: a risk-based approach to benefit society failures in West Kent, 1860–1885

ABSTRACT

Much of the extensive historiography of nineteenth-century English friendly societies has focused on their capacity to create social capital and their predisposition towards insolvency, although more recently historians have argued that the ability of affiliated orders to raise levies made them financially stronger than previously supposed. However, given that societies still failed regularly, in what circumstances was it impossible to mount such rescues, and why? This article answers this question through a case study of three small ‘local’ societies in West Kent. Using the perspective of risk and risk transfer, it suggests they all suffered from a serious combination of membership and competitor risk which, when mishandled, also caused significant reputational damage to both societies and their managers. A further major tension was that many local opinion formers were starting to demand higher standards of accountability, including from friendly society trustees. The article also looks at the survival strategies of individual managers and members, arguing that – during financial crises – self-interest exerted powerful forces on those societies. In each case, the evidence reveals that often longstanding local relationships, based on mutual trust and shared values, were fractured so completely that raising levies or other bailouts to keep them afloat proved impossible.     

Scientific Misconduct in Social Welfare Research: Preventive Lessons from Other Fields

This article examines instances of scientific misconduct which have occurred in recent years from an international perspective and explores their implications for the burgeoning social work research enterprise. Although the majority of cases of scientific misconduct which have received international publicity are confined to the hard sciences, there are important implications to be derived for the growing productivity in social welfare research. A comprehensive search was conducted of press reports for the time frame of 1995–2002 for both US and international stories about research misconduct. Using qualitative analysis, themes related to scientific conduct were identified. These cases reveal serious consequences to individual researchers and to their employing institutions. The motivating factors for the wrongdoing are applicable to researchers of any discipline and include perceived pressure to document positive outcomes and the rush to publish findings. The authors urge development of standards of research conduct specific to social work and social welfare and increased dialogue about the issues involved.     

Industry culture and strategic response: the case of the british bus industry

This article is concerned to demonstrate the utility of the notion of culture at the industry level. A multiple-case study of nine British bus companies highlights the way in which entrenched views and practices which were common among the nine companies restricted the range of strategic thinking among them. Many of these views and practices are emblematic of traditional bus company perspectives and are common throughout the industry. The reasons for the “hold” that the industry culture has on managers are examined. These include such factors as the impact of leadership and the inheritance of pre-deregulation features.     

Thinking with suffering

Abstract

This article provides a critical review of literature on ‘social suffering’. Analytical attention is focused upon the ways in which writers struggle to bring ‘meaning’ to this topic. All sense that there is always something in events of extreme suffering that resists conceptualisation and defies analysis. This problem of establishing a language for ‘thinking with suffering’ is explored with reference to the works of Hannah Arendt, Paul Ricoeur and Max Weber. An agenda for sociological research is proposed which focuses on the struggle to make sense out of the phenomenon of suffering as a force of cultural innovation. In this context, it is suggested that what is most interesting here is the evidence to suggest that, when faced with the ‘brute fact’ of a world where there appears to be too much suffering, people are always moved to make this phenomenon productive for thought and action.     

Using the ‘risk factor paradigm’ in prevention: lessons from the evaluation of Communities that Care

This paper discusses results from an evaluation of the UK-based Communities that Care programme. This ‘risk and protective’ programme was set up in 1997 by the Joseph Rowntree Foundation with a central aim of exploring if such an approach could be successful in the UK context. Communities that Care puts into operation the ‘risk factor paradigm’ by providing evidence of risk and protection for local communities that help them target resources for the reduction of future social problems. Results show that such a model looks ‘promising’ although though there are substantial implementation issues that need to be addressed if it is to be successful over a longer time frame.     

Assembling an economic actor: the agencement of a Hedge Fund

Michel Callon has conceptualised economic actors as made up of socio‐technical agencements: collectives of human beings, technical devices, algorithms, and so on. This article reports a pilot, partially observational study of a hedge fund, a category of actor in financial markets that is of growing importance but that has so far attracted little attention in economic sociology. It draws on that study, and on interviews with other financial market practitioners, to delineate what is involved in viewing such an actor as made up of an agencement, and discusses the merits of doing so.     

AVERAGE PREDICTIVE COMPARISONS FOR MODELS WITH NONLINEARITY, INTERACTIONS, AND VARIANCE COMPONENTS

In a predictive model, what is the expected difference in the outcome associated with a unit difference in one of the inputs? In a linear regression model without interactions, this average predictive comparison is simply a regression coefficient (with associated uncertainty). In a model with nonlinearity or interactions, however, the average predictive comparison in general depends on the values of the predictors. We consider various definitions based on averages over a population distribution of the predictors, and we compute standard errors based on uncertainty in model parameters. We illustrate with a study of criminal justice data for urban counties in the United States. The outcome of interest measures whether a convicted felon received a prison sentence rather than a jail or non-custodial sentence, with predictors available at both individual and county levels. We fit three models: (1) a hierarchical logistic regression with varying coefficients for the within-county intercepts as well as for each individual predictor; (2) a hierarchical model with varying intercepts only; and (3) a nonhierarchical model that ignores the multilevel nature of the data. The regression coefficients have different interpretations for the different models; in contrast, the models can be compared directly using predictive comparisons. Furthermore, predictive comparisons clarify the interplay between the individual and county predictors for the hierarchical models and also illustrate the relative size of varying county effects.