Lasso回归：从解释到预测

传统的最小二乘回归法关注于对当前数据集的准确估计, 容易导致模型的过拟合, 影响模型结论的可重复性。随着方法学领域的发展, 涌现出的新兴统计工具可以弥补传统方法的局限, 从过度关注回归系数值的解释转向提升研究结果的预测能力也愈加成为心理学领域重要的发展趋势。Lasso方法通过在模型估计中引入惩罚项的方式, 可以获得更高的预测准确度和模型概化能力, 同时也可以有效地处理过拟合和多重共线性问题, 有助于心理学理论的构建和完善。     

Effects of stop signal modality, stop signal intensity and tracking method on inhibitory performance as determined by use of the stop signal paradigm

In Experiment 1, the effects of stop signal modality on the speed and efficiency of the inhibition process were examined. Stop signal reaction time (SSRT) and inhibition function slope in an auditory stop signal condition were compared to SSRT and inhibition function slope in a visual stop signal condition. It was found that auditory stop signals compared to visual stop signals enhanced both the speed and efficiency of stopping. The modality effects were attributed to differences in the neurophysiological processes underlying perception. However, Experiment 2 demonstrated that the modality difference was larger for 80 dB(A) auditory stop signals than 60 dB(A) auditory stop signals. This effect was reconciled with the suggestion that loud tones are more capable of eliciting immediate arousing effects on motor processes than weak tones and visual stimuli. The second purpose of the present investigation was to explore the utility (and potential advantages) of an alternative way of setting stop signal delay relative to mean reaction time (MRT). The method that was suggested compensates for inter-individual differences in primary task reaction speed by setting stop signal delays as proportions of the subjects' MRT.     

Psychometric Modeling of response speed and accuracy with mixed and conditional regression

Human performance in cognitive testing and experimental psychology is expressed in terms of response speed and accuracy. Data analysis is often limited to either speed or accuracy, and/or to crude summary measures like mean response time (RT) or the percentage correct responses. This paper proposes the use of mixed regression for the psychometric modeling of response speed and accuracy in testing and experiments. Mixed logistic regression of response accuracy extends logistic item response theory modeling to multidimensional models with covariates and interactions. Mixed linear regression of response time extends mixed ANOVA to unbalanced designs with covariates and heterogeneity of variance. Related to mixed regression is conditional regression, which requires no normality assumption, but is limited to unidimensional models. Mixed and conditional methods are both applied to an experimental study of mental rotation. Univariate and bivariate analyzes show how within-subject correlation between response and RT can be distinguished from between-subject correlation, and how latent traits can be detected, given careful item design or content analysis. It is concluded that both response and RT must be recorded in cognitive testing, and that mixed regression is a versatile method for analyzing test data.I am grateful to Rogier Donders for putting his data at my disposal.     

Evaluating the Impact of Genetic Counseling and Testing with Signal Detection Methods

One measure of the impact of genetic counseling and testing (GCT) is the extent to which it fosters behavioral change that is consistent with mutation status. We describe and illustrate how two different signal detection methods, receiver operating characteristic (ROC) analysis and recursive partitioning, can be used in this context to evaluate the impact of GCT. We analyzed real screening behavior data obtained in the 12 months following GCT for Hereditary Nonpolyposis Colon Cancer (HNPCC) using these two different signal detection approaches. Each approach demonstrated that GCT had an impact on behavioral outcomes, and was effective in fostering behavioral outcomes appropriate to mutation status. The ROC approach demonstrated that GCT was effective because mutation positive and mutation negative individuals could be distinguished on the basis of the number of recommended screening behaviors. The recursive partitioning approach demonstrated that GCT was effective because there were generally high rates of adherence to screening guidelines among subjects. The recursive partitioning technique also identified four subgroups of subjects, each with distinct characteristics, for which tailored interventions could be developed to increase rates of adherence to screening guidelines. Signal detection methods are easily implemented and are useful techniques for evaluating the impact of GCT.     

The Duty to Eradicate Global Poverty: Positive or Negative?

In World Poverty and Human Rights, Thomas Pogge argues that the global rich have a duty to eradicate severe poverty in the world. The novelty of Pogges approach is to present this demand as stemming from basic commands which are negative rather than positive in nature: the global rich have an obligation to eradicate the radical poverty of the global poor not because of a norm of beneficence asking them to help those in need when they can at little cost to themselves, but because of their having violated a principle of justice not to unduly harm others by imposing on them a coercive global order that makes their access to the objects of their human right to subsistence insecure. In this paper, I claim that although Pogge is right in arguing that negative duties are crucial in an account of global justice, he is wrong in saying that they are the only ones that are crucial. Harming the global poor by causing their poverty provides a sufficient but not a necessary condition for the global rich to have a duty of justice to assist them. After engaging in a critical analysis of Pogges argument, I conclude by suggesting the need for a robust conception of cosmopolitan solidarity that includes positive duties of assistance which are not mere duties of charity, but enforceable ones of justice.     

The Role of Apparent Constraints in Normative Reasoning: a Methodological Statement and Application to Global Justice

The assumptions that are made about the features of the world that are relatively changeable by agents and those that are not (constraints) play a central role in determining normative conclusions. In this way, normative reasoning is deeply dependent on accounts of the empirical world. Successful normative reasoning must avoid the naturalization of constraints and seek to attribute correctly to agents what is and is not in their power to change. Recent discourse on global justice has often come to unjustified conclusions about agents obligations due to a narrow view of what is changeable and by whom.I would like to thank for their helpful comments Christian Barry, Rudiger Bittner, Darrel Moellendorf and Thomas Pogge.     

Signal-detection analyses of conditional discrimination and delayed matching-to-sample performance

Quantitative analyses of stimulus control and reinforcer control in conditional discriminations and delayed matching-to-sample procedures often encounter a problem; it is not clear how to analyze data when subjects have not made errors. The present article examines two common methods for overcoming this problem. Monte Carlo simulations of performance demonstrated that both methods introduced systematic deviations into the results, and that there were genuine risks of drawing misleading conclusions concerning behavioral models of signal detection and animal short-term memory.     

A formal model for structuring local perceptions of environmental space

The problem of constructing computational models of space that mimic those found in human spatial reasoning is addressed. This paper extends a formal model that addressed point objects in a 2D world to include the presence of line objects and barriers. The problem consists of determining an appropriate global model for the spatial configuration given a sequence of local views or observations. A formal model is proposed based on constraint network theory. The model leads to the identification of stable regions within which perceptual information about the environment changes slowly, and zones of transition within which the perceptual information changes more quickly. A special case of transition zones is a gateway, which may be viewed as a kind of discontinuity or singularity in the model. It is expected that the model obtained will be used to mimic human mental representations of navigable outdoor environments.     

Toward a New Concept of Global Morality

The human community faces today the most serious challenges ever to have confronted the planet in the areas of health, environment, and security. Science and technology are essential for responding to these challenges. More is needed, however, because science is not equipped to deal adequately with the values dimensions and the political issues that accompany the challenges. For an adequate response, there must be cooperative effort by scientists and statespersons, informed for moral leadership by the religious wisdom that is available. The religious communities can provide this spiritual dimension, thereby fulfilling their traditional role, but it will require their coming to terms with the character of the scientific and technological base of contemporary culture. The paper lays the conceptual groundwork for understanding these issues.