The importance of ‘mass’ in line drawings of faces

Previous research (Davies, Ellis and Shepherd, 1978; Rhodes, Brennan and Carey, 1987) has shown that accurate line drawings of familiar faces are identified rather poorly. However, artists can produce lifelike portraits with pen and ink, and Pearson and Robinson (1985) described an automatic method for producing computer-drawn sketches (‘cartoons’) of faces which appear very similar to those produced by a human artist. In this paper we show that subjects can identify famous faces depicted in such computer-drawn ‘cartoons’ almost as well as full grey-scale images. The cartoon algorithm comprises two components. One component draws lines at the locations of intensity changes corresponding to luminance valleys and edges (the ‘valledge’ detector). The other component applies a ‘threshold’ to the original intensity distribution, and replaces any area darker than threshold with black. Thus the full cartoon contains both ‘line’ and ‘mass’. Neither the valledges nor the threshold components alone were as well identified as full cartoons containing both components. The results suggest that the addition of the threshold component adds significantly to the identifiability of line drawings of faces.     

Counsellors' responses to clients' non-attendance at counselling sessions

Although there has been some study of clients who miss psychotherapy appointments, very little attention has been given to therapists' responses in such circumstances. A questionnaire survey of student counsellors working in British higher education institutions invited them to consider whether or not they would initiate any further contact with each of eight hypothetical non-attenders, described in case vignettes. It was predicted that responses would be influenced both by the subjects' theoretical approach to counselling and by their level of interpersonal dependency, as measured with the Interpersonal Dependency Inventory (Hirschfeld et al., 1977). It was found that the majority of counsellors were not subject to institutional guidelines about contacting non-attenders, and that responses ranged from contacting none to contacting all of the clients. Psychodynamic counsellors were found to be significantly more likely to initiate contact than were humanistic counsellors. There were no significant correlations between contact scores and intepersonal dependency scores but women were significantly more likely to initiate contact than were men. Many of the psychodynamic subjects had commented on their therapeutic use of written communications to interpret non-attendance to their clients.     

A framework for ML estimation of parameters of (mixtures of) common reaction time distributions given optional truncation or censoring

We present a framework for distributional reaction time (RT) analysis, based on maximum likelihood (ML) estimation. Given certain information relating to chosen distribution functions, one can estimate the parameters of these distributions and of finite mixtures of these distributions. In addition, left and/or right censoring or truncation may be imposed. Censoring and truncation are useful methods by which to accommodate outlying observations, which are a pervasive problem in RT research. We consider five RT distributions: the Weibull, the ex-Gaussian, the gamma, the log-normal, and the Wald. We employ quasi-Newton optimization to obtain ML estimates. Multicase distributional analyses can be carried out, which enable one to conduct detailed (across or within subjects) comparisons of RT data by means of loglikelihood difference tests. Parameters may be freely estimated, estimated subject to boundary constraints, constrained to be equal (within or over cases), or fixed. To demonstrate the feasibility of ML estimation and to illustrate some of the possibilities offered by the present approach, we present three small simulation studies. In addition, we present three illustrative analyses of real data.     

EZ does it! Extensions of the EZ-diffusion model

In this rejoinder, we address two of Ratcliff’s main concerns with respect to the EZ-diffusion model (Ratcliff, 2008). First, we introduce “robust-EZ,” a mixture model approach to achieve robustness against the presence of response contaminants that might otherwise distort parameter estimates. Second, we discuss an extension of the EZ model that allows the estimation of starting point as an additional parameter. Together with recently developed, user-friendly software programs for fitting the full diffusion model (Vandekerckhove & Tuerlinckx, 2007; Voss & Voss, 2007), the development of the EZ model and its extensions is part of a larger effort to make diffusion model analyses accessible to a broader audience, an effort that is long overdue.     

The power to detect sex differences in IQ test scores using Multi-Group Covariance and Means Structure Analyses

Research into sex differences in general intelligence, g, has resulted in two opposite views. In the first view, a g-difference is nonexistent, while in the second view, g is associated with a male advantage. Past research using Multi-Group Covariance and Mean Structure Analysis (MG-CMSA) found no sex difference in g. This failure raised the question whether the g-difference is truly absent or whether MG-CMSA lacked statistical power to detect it. The present study used the likelihood ratio test to investigate the power to detect a g-difference in the WAIS-III factor structure with MG-CMSA. Various situations were examined including those reported in the literature. Results showed that power varies greatly among different scenarios. The scenarios based on previous results were associated with power coefficients of about 0.5–0.6. Implications of these findings are discussed in the light of research into sex differences in IQ.     

Heteroscedastic one‐factor models and marginal maximum likelihood estimation

In the present paper, a general class of heteroscedastic one‐factor models is considered. In these models, the residual variances of the observed scores are explicitly modelled as parametric functions of the one‐dimensional factor score. A marginal maximum likelihood procedure for parameter estimation is proposed under both the assumption of multivariate normality of the observed scores conditional on the single common factor score and the assumption of normality of the common factor score. A likelihood ratio test is derived, which can be used to test the usual homoscedastic one‐factor model against one of the proposed heteroscedastic models. Simulation studies are carried out to investigate the robustness and the power of this likelihood ratio test. Results show that the asymptotic properties of the test statistic hold under both small test length conditions and small sample size conditions. Results also show under what conditions the power to detect different heteroscedasticity parameter values is either small, medium, or large. Finally, for illustrative purposes, the marginal maximum likelihood estimation procedure and the likelihood ratio test are applied to real data.     

The Price of Pain and the Value of Suffering

ABSTRACT— Estimating the financial value of pain informs issues as diverse as the market price of analgesics, the cost-effectiveness of clinical treatments, compensation for injury, and the response to public hazards. Such valuations are assumed to reflect a stable trade-off between relief of discomfort and money. Here, using an auction-based health-market experiment, we show that the price people pay for relief of pain is strongly determined by the local context of the market, that is, by recent intensities of pain or immediately disposable income (but not overall wealth). The absence of a stable valuation metric suggests that the dynamic behavior of health markets is not predictable from the static behavior of individuals. We conclude that the results follow the dynamics of habit-formation models of economic theory, and thus, this study provides the first scientific basis for this type of preference modeling.     

Facial identity and facial expression matching in 5–12‐year‐old children and adults

Facial identity and facial expression matching tasks were completed by 5–12‐year‐old children and adults using stimuli extracted from the same set of normalized faces. Configural and feature processing were examined using speed and accuracy of responding and facial feature selection, respectively. Facial identity matching was slower than face expression matching for all age groups. Large age effects were found on both speed and accuracy of responding and feature use in both identity and expression matching tasks. Eye region preference was found on the facial identity task and mouth region preference on the facial expression task. Use of mouth region information for facial expression matching increased with age, whereas use of eye region information for facial identity matching peaked early. The feature use information suggests that the specific use of primary facial features to arrive at identity and emotion matching judgments matures across middle childhood. Copyright © 2009 John Wiley & Sons, Ltd.