Integration and segregation of prior outcomes in risky decisions

Two experiments were conducted to test if a prior outcome influences the likelihood to accept a current gamble. Undergraduate students participating as subjects imagined that they on a fictitious betting day at a horse-race track did not gamble in the prior race, that they gambled and won, or that they gambled and lost. Subjects rated in one session the likelihood of gambling in the current race. In another session they rated how satisfied they would be with not gambling, with winning, and with losing, respectively. The results of Experiment 1 showed that as compared to no prior outcome the ratings of likelihood of gambling increased after a gain and decreased after a loss. This was explained by the assumption, supported by the ratings of satisfaction, that the prior outcome only affected the satisfaction with the expected loss of the current choice, making it less negative after a gain and more negative after a loss. These results were replicated in Experiment 2. In addition, if subjects did not know with certainty the outcome of a previous choice to gamble, the likelihood to accept the current gamble and the ratings of the satisfaction with its expected outcomes were largely unaffected.     

Constrained canonical correlation

This paper explores some of the problems associated with traditional canonical correlation. A response surface methodology is developed to examine the stability of the derived linear functions, where one wishes to investigate how much the coefficients can change and still be in an -neighborhood of the globally optimum canonical correlation value. In addition, a discrete (or constrained) canonical correlation method is formulated where the derived coefficients of these linear functions are constrained to be in some small set, e.g., {1, 0, –1}, to aid in the interpretation of the results. An example concerning the psychographic responses of Wharton MBA students of the University of Pennsylvania regarding driving preferences and life-style considerations is provided.Wayne S. DeSarbo, Robert Jausman, Shen Lin, and Wesley Thompson are all Members of Technical Staff at Bell Laboratories. We wish to express our gratitude to the editor and reviewers of this paper for their insightful remarks.     

A bayesian procedure for mastery decisions based on multivariate normal test data

A Bayesian framework for making mastery/nonmastery decisions based on multivariate test data is described in this study. Overall, mastery is granted (or denied) if the posterior expected loss associated with such action is smaller than the one incurred by the denial (or grant) of mastery. An explicit form for the cutting contour which separates mastery and nonmastery states in the test score space is given for multivariate normal test scores and for a constant loss ratio. For multiple cutting scores in the true ability space, the test score cutting contour will resemble the boundary defined by multiple test cutting scores when the test reliabilities are reasonably close to unity. For tests with low reliabilities, decisions may very well be based simply on a suitably chosen composite score.This work was performed pursuant to Grant NIE-G-78-0087 with the National Institute of Education, Department of Health, Education, and Welfare, Huynh Huynh, Principal Investigator. Points of view or opinions stated do not necessarily reflect NIE positions or policy and no official endorsement should be inferred. The assistance of Joseph C. Saunders is gratefully acknowledged. The author is indebted to an anonymous referee who pointed out several computational errors in the earlier versions of the paper.     

Mixtures of conditional mean- and covariance-structure models

Models and parameters of finite mixtures of multivariate normal densities conditional on regressor variables are specified and estimated. We consider mixtures of multivariate normals where the expected value for each component depends on possibly nonnormal regressor variables. The expected values and covariance matrices of the mixture components are parameterized using conditional mean- and covariance-structures. We discuss the construction of the likelihood function, estimation of the mixture model with regressors using three different EM algorithms, estimation of the asymptotic covariance matrix of parameters and testing for the number of mixture components. In addition to simulation studies, data on food preferences are analyzed.The authors are grateful to Donald B. Rubin and Michael E. Sobel for critical reading of a first draft of this paper and to three anonymous reviewers ofPsychometrika for their helpful comments and suggestions. The research of the first and the third author was supported by a grant from the Deutsche Forschungsgemeinschaft.     

Perceived control,nature of risk information and risk taking. An experimental test of a simple taxonomy of uncertainty

Uncertainty may be categorized along two dimensions: (1) the nature of probabilistic information (i.e. frequency information, about the outcomes of similar situations in the past, versus process information, about the way(s) in which a future loss might occur), and (2) the degree of personal control (i.e. the extent to which an activity's outcomes depend on internal factors (e.g. knowledge, skills) versus external (e.g. chance) factors). The effects of variations in both dimensions on people's risk-taking tendency were experimentally studied. In a computerized task, subjects had to stop a fast-moving symbol before it passed a target line. Success yielded a financial gain, failure led them into a ‘penalty task’ with the possibility of a considerable loss. On each trial subjects chose among 10 risk levels (varying symbol speeds): low levels resulted in small but almost sure gains, high levels yielded larger but less probable gains. Across subgroups of subjects, three penalty task characteristics were varied: (1) the actual loss probability, (2) the external versus internal determination of outcomes, and (3) the available risk information. Major findings were: (a) subjects did not set a lower risk level, but they did appear to be more attentive (i.e. they failed fewer trials) when the actual loss probability was lower; (b) internal outcome determination resulted in more failed trials (lesser caution); (c) subjects ignored frequency information, but were sensitive to process information. Results are discussed in terms of effort allocation for controlling risk during task performance.     

A simple multivariate probabilistic model for preferential and triadic choices

Multidimensional probabilistic models of behavior following similarity and choice judgements have proven to be useful in representing multidimensional percepts in Euclidean and non-Euclidean spaces. With few exceptions, these models are generally computationally intense because they often require numerical work with multiple integrals. This paper focuses attention on a particularly general triad and preferential choice model previously requiring the numerical evaluation of a 2n-fold integral, wheren is the number of elements in the vectors representing the psychological magnitudes. Transforming this model to an indefinite quadratic form leads to a single integral. The significance of this form to multidimensional scaling and computational efficiency is discussed.The authors would like to thank Jean-Claude Falmagne and Norman Johnson for suggestions and advice concerning quadratic forms.     

Multivariate analysis with linearizable regressions

We study the class of multivariate distributions in which all bivariate regressions can be linearized by separate transformation of each of the variables. This class seems more realistic than the multivariate normal or the elliptical distributions, and at the same time its study allows us to combine the results from multivariate analysis with optimal scaling and classical multivariate analysis. In particular a two-stage procedure which first scales the variables optimally, and then fits a simultaneous equations model, is studied in detail and is shown to have some desirable properties.     

Sequential social network data

A new method is proposed for the statistical analysis of dyadic social interaction data measured over time. The data to be studied are assumed to be realizations of a social network of a fixed set of actors interacting on a single relation. The method is based on loglinear models for the probabilities for various dyad (or actor pair) states and generalizes the statistical methods proposed by Holland and Leinhardt (1981), Fienberg, Meyer, & Wasserman (1985), and Wasserman (1987) for social network data. Two statistical models are described: the first is an associative approach that allows for the study of how the network has changed over time; the second is a predictive approach that permits the researcher to model one time point as a function of previous time points. These approaches are briefly contrasted with earlier methods for the sequential analysis of social networks and are illustrated with an example of longitudinal sociometric data.Research support provided by National Science Foundation Grant #SES84-08626 to the University of Illinois at Urbana-Champaign and by a predoctoral traineeship awarded to the second author by the Quantitative Methods Program of the Department of Psychology, University of Illinois at Urbana-Champaign, funded by ADAHMA, National Research Service Award #MH14257. We thank the editor and three anonymous referees for helpful comments.This paper is based on research presented at the 1986 Annual Meeting of the Psychometric Society, Toronto, Ontario, June, 1986.     

The impact of variability and prechoice experience on taking safety measures: The case of security updates

The goal of this research is to clarify the conditions that trigger reluctance to take cost-effective safety measures. We present three experiments. In two of the experiments, the participants were asked to operate a simulated system for 20 periods, each with 10 trials. They could “update the system” to eliminate the risk of a “security failure” that led to a loss of 100 points. The updating cost was either fixed (at 10 points) or variable (initially 10 points, and some probability of free—0 points—updates). The optimal strategy prescribed updating at the first opportunity. Another experiment focused on one-shot decision under risk. The results highlight two factors that reduce the tendency to update and impair performance: cost variability and prechoice experience. Importantly, we show that the negative impact of cost variability is the product of two tendencies. First, experiencing periods with free updates slowed learning to select the optimal policy. Second, in many cases, the participants behaved as if they plan to update when the cost of updating is low but forget to do so. The results suggest that security can be enhanced by asking users to select a default updating policy before gaining experience and by replacing “free updates” with automatic updates. Information concerning the existence of automatic updates reduced manual updating, but this effect was eliminated by experience.     

投资决策进化心理学的研究：预期的私人资金分配和父母对子女的差异性精力投入

投资决策的进化心理学研究着眼于辨认人类获得进化适应的特定环境中经常出现的典型性风险,探寻为了应对这些风险而进化出的信息处理机制,并验证现时的社会因素和个体因素对这些心理机制的激活或抑制作用。在研究一中,被试预测了与自己同龄的男人或女人如何分配一笔中彩的奖金给自己和其他可能的受益人。研究发现：（1）钱数的分配大体由亲缘关系的疏密程度决定;（2）两性被试都假想男性比女性更慷慨,但实际上男性表现得更自利;（3）女性被试预测男性中奖人的金钱分配比男性被试预测女性中奖人的金钱分配更为准确;（4）女性被试的受益人更多,分享的社会范围更广。研究二探讨了父母对子女投入精力的不同取决于家庭的相对财富而非绝对财富的进化心理学假说。用哺乳与否和生育间隔期为测量指标,研究结果显示：（1）家庭实际收入影响父母对子女的总投入;（2）与邻里家庭相比,父母对于自己家庭相对收入的认知影响了对子女有别的差异性精力投入。基于男性普遍在财富和生育数量上比女性有更大的变异度,投资儿子比投资女儿更具博弈性。两项研究表明,人类的理性决策既受限于社会关系又适应于相对的财富状况