Bayesian and maximum likelihood estimation of hierarchical response time models

Hierarchical (or multilevel) statistical models have become increasingly popular in psychology in the last few years. In this article, we consider the application of multilevel modeling to the ex-Gaussian, a popular model of response times. We compare single-level and hierarchical methods for estimation of the parameters of ex-Gaussian distributions. In addition, for each approach, we compare maximum likelihood estimation with Bayesian estimation. A set of simulations and analyses of parameter recovery show that although all methods perform adequately well, hierarchical methods are better able to recover the parameters of the ex-Gaussian, by reducing variability in the recovered parameters. At each level, little overall difference was observed between the maximum likelihood and Bayesian methods.     

Long-term effects of prior cocaine exposure on Morris water maze performance

Cocaine addiction is associated with long-term cognitive alterations including deficits on tests of declarative/spatial learning and memory. To determine the extent to which cocaine exposure plays a causative role in these deficits, adult male Long-Evans rats were given daily injections of cocaine (30 mg/kg/day x 14 days) or saline vehicle. Three months later, rats were trained for 6 sessions on a Morris water maze protocol adapted from Gallagher, Burwell, and Burchinal [Gallagher, M., Burwell, R., & Burchinal, M. (1993). Severity of spatial learning impairment in aging: development of a learning index for performance in the Morris water maze. Behavioral Neuroscience, 107, 618-626]. Rats given prior cocaine exposure performed similarly to controls on training trials, but searched farther from the platform location on probe trials interpolated throughout the training sessions and showed increased thigmotaxis. The results demonstrate that a regimen of cocaine exposure can impair Morris water maze performance as long as 3 months after exposure. Although the impairments were not consistent with major deficits in spatial learning and memory, they may have resulted from cocaine-induced increases in stress responsiveness and/or anxiety. Increased stress and anxiety would be expected to increase thigmotaxis as well as cause impairments in searching for the platform location, possibly through actions on ventral striatal dopamine signaling.     

Motivated information processing in group judgment and decision making.

This article expands the view of groups as information processors into a motivated information processing in groups (MIP-G) model by emphasizing, first, the mixed-motive structure of many group tasks and, second, the idea that individuals engage in more or less deliberate information search and processing. The MIP-G model postulates that social motivation drives the kind of information group members attend to, encode, and retrieve and that epistemic motivation drives the degree to which new information is sought and attended to, encoded, and retrieved. Social motivation and epistemic motivation are expected to influence, alone and in combination, generating problem solutions, disseminating information, and negotiating joint decisions. The MIP-G model integrates the influence of many individual and situational differences and combines insight on human thinking with group-level interaction process and decision making.     

Not So Enticing Reasons

A common view of the relation between oughts and reasons is that you ought to do something if and only if that is what you have most reason to do. One challenge to this comes from what Jonathan Dancy calls ‘enticing reasons.’ Dancy argues that enticing reasons never contribute to oughts and that it is false that if the only reasons in play are enticing reasons then you ought to do what you have most reason to do. After explaining how enticing reasons supposedly work and why accepting them may appear attractive, I firstly show why we are not committed to accepting them into our conceptual framework and then argue that no reasons work in the way enticing reasons are claimed to. Thus we should reject the category of enticing reasons entirely.

Absolute localization of vibrotactile stimuli on the torso

Vibrotactile mobility systems present spatial information such as the direction of a waypoint through a localized vibration on the torso. Using these systems requires the ability to determine the absolute location of the stimulus. Because data are available only on the ability to determine the relative location of stimuli on the torso, we developed a novel method for measuring absolute localization on the basis of triangulation. For 15 observers, we calculated the subjective location of visual and tactile stimuli on the frontal half of the torso. The size of the 95% confidence intervals around the subjective tactile locations is about the size of the stimuli (1.66 cm) and is slightly larger than that around the subjective visual locations (mean difference, 0.17 cm). The error in tactile judgments over and above that in the visual judgments is present only for locations near the body midline. When the subjective visual and tactile locations are not co-located, the difference can best be described by a shift along the radius from the body midaxis. The same holds for the differences between the veridical and the subjective locations. Therefore, the difference between the veridical and the subjective directions of a stimulus is small. The results make us believe that stimulus locations on the torso are coded in polar coordinates of which the angle is perceptual invariant and the distance is less important, probably because it varies with changes in, among other things, posture and breathing.     

Toward an integrative social identity model of collective action: a quantitative research synthesis of three socio-psychological perspectives

An integrative social identity model of collective action (SIMCA) is developed that incorporates 3 socio-psychological perspectives on collective action. Three meta-analyses synthesized a total of 182 effects of perceived injustice, efficacy, and identity on collective action (corresponding to these socio-psychological perspectives). Results showed that, in isolation, all 3 predictors had medium-sized (and causal) effects. Moreover, results showed the importance of social identity in predicting collective action by supporting SIMCA's key predictions that (a) affective injustice and politicized identity produced stronger effects than those of non-affective injustice and non-politicized identity; (b) identity predicted collective action against both incidental and structural disadvantages, whereas injustice and efficacy predicted collective action against incidental disadvantages better than against structural disadvantages; (c) all 3 predictors had unique medium-sized effects on collective action when controlling for between-predictor covariance; and (d) identity bridged the injustice and efficacy explanations of collective action. Results also showed more support for SIMCA than for alternative models reflecting previous attempts at theoretical integration. The authors discuss key implications for theory, practice, future research, and further integration of social and psychological perspectives on collective action.     

On use of student samples for scale construction

This research note responds to the question of whether a convenience sample of undergraduate students may be successfully utilized in concept development and in scale construction, and in what way the results are comparable to the findings of a representative national sample. The results of a Mokken analysis in both samples support the hypothesis that convenience samples have utility in concept development and in developing measures that can also be used in representative samples.