George L. Wolford

The principled control of false positives in neuroimaging

An incredible amount of data is generated in the course of a functional neuroimaging experiment. The quantity of data gives us improved temporal and spatial resolution with which to evaluate our results. It also creates a staggering multiple testing problem. A number of methods have been created that address the multiple testing problem in neuroimaging in a principled fashion. These methods place limits on either the familywise error rate (FWER) or the false discovery rate (FDR) of the results. These principled approaches are well established in the literature and are known to properly limit the amount of false positives across the whole brain. However, a minority of papers are still published every month using methods that are improperly corrected for the number of tests conducted. These latter methods place limits on the voxelwise probability of a false positive and yield no information on the global rate of false positives in the results. In this commentary, we argue in favor of a principled approach to the multiple testing problem--one that places appropriate limits on the rate of false positives across the whole brain gives readers the information they need to properly evaluate the results.

Neural correlates of interspecies perspective taking in the post-mortem Atlantic Salmon: an argument for multiple comparisons correction

With the extreme dimensionality of functional neuroimaging data comes extreme risk for false positives. Across the 130,000 voxels in a typical fMRI volume the probability of a false positive is almost certain. Correction for multiple comparisons should be completed with these datasets, but is often ignored by investigators. To illustrate the magnitude of the problem we carried out a real experiment that demonstrates the danger of not correcting for chance properly. GLM RESULTS

Expressive disclosure and health outcomes in a prostate cancer population.

Objective: This pilot study explored the feasibility and the efficacy of a brief, well-defined psychosocial intervention (expressive disclosure) in improving behavioral, medical, immunological, and emotional health outcomes in men with diagnosed prostate cancer. Method: Thirty prostate cancer patients receiving outpatient oncology care were randomized into experimental (disclosure) and control (non-disclosure) groups. All had been previously treated by surgery or radiation within the last 4 years and were being monitored without further intervention for change in PSA levels. Psychological and physical health surveys were administered and peripheral blood for PSA levels and immune assays was obtained upon study enrollment and again at 3 and 6 months post enrollment. Multivariate analyses were used to examine how the expressive disclosure impacted the hypothesized domains of functioning: physical and psychological symptoms; health care utilization; and immunocompetence. Results: Compared to controls, patients in the expressive disclosure condition showed improvements in the domains of physical symptoms and health care utilization, but not in psychological variables nor in disease relevant aspects of immunocompetence. Conclusions: Study results support the feasibility of an expressive disclosure intervention for men with prostate cancer. The intervention was well accepted by this population, and participation/adherence was quite high. Results provide only limited support for the hypothesis that a written emotional disclosure task can positively impact health outcomes in a cancer population. However, this pilot study may have lacked adequate power to detect possible intervention benefits. Further studies with larger samples are needed to better assess the interventions impact on psychological well-being and immunocompetence.

Lateral masking as a function of spacing

It is argued that lateral masking is a composite of several processes. These processes include response competition, distribution of attention, perceptual grouping, and feature (contour) interaction. Three experiments were carried out in an attempt to isolate some of the components. In the first two experiments, it was shown that feature interaction dominates at close spacing but other processes dominate at wider spacing. The third experiment showed that at least part of the effect of perceptual grouping appears to be information provided about target location.

Processing of unattended visual information

A visual analog of auditory selective attention paradigms is described. Using that analog, we examined the amount of processing required to monitor unattended information. With the materials used, there was no evidence that any processing capacity was required to monitor the unattended information. Memory for the unattended information and the recognition of one’s name were examined to provide additional evidence concerning the fate of the unattended information.

Extensive Individual Differences in Brain Activations Associated with Episodic Retrieval are Reliable Over Time

The localization of brain functions using neuroimaging techniques is commonly dependent on statistical analyses of groups of subjects in order to identify sites of activation, particularly in studies of episodic memory. Exclusive reliance on group analysis may be to the detriment of understanding the true underlying cognitive nature of brain activations. In the present study, we found that the patterns of brain activity associated with episodic retrieval are very distinct for individual subjects from the patterns of brain activity at the group level. These differences go beyond the relatively small variations due to cyctoarchitectonic differences or spatial normalization. We quantify this individual variability by cross-correlating volumes of brain images. We demonstrate that individual patterns of brain activity are reliable over time despite their extensive variability. We suggest that varied but reliable individual patterns of significant brain activity may be indicative of different cognitive strategies used to produce a recognition response. We believe that individual analysis in conjunction with group analysis may be critical to fully understanding the relationship between retrieval processes and underlying brain regions.

Transformations for Within-Subject Designs: A Monte Carlo Investigation

We explored the use of transformations to improve power in within-subject designs in which multiple observations are collected for each S in each condition, such as reaction time and psychophysiological experiments. Often, the multiple measures within a treatment are simply averaged to yield a single number, but other transformations have been proposed. Monte Carlo simulations were used to investigate the influence of those transformations on the probabilities of Type I and Type II errors. With normally distributed data, Z and range correction transformations led to substantial increases in power over simple averages. With highly skewed distributions, the optimal transformation depended on several variables, but Z and range correction performed well across conditions. Correction for outliers was useful in increasing power, and trimming was more effective than eliminating all points beyond a criterion.

Evidence for feature perturbations.

Feature perturbations were proposed by Wolford (1975) to explain the pattern of errors generated by subjects in experiments involving speeded presentations. These experiments were carried out to test the notion of feature perturbations in a fairly direct fashion. The logic of the experiments involved the creation of an artificial alphabet in which feature perturbations, if they occurred, would lead to predictable error patterns. The results of the three experiments supported the notion of feature perturbations. Evidence was also found for errors which resulted from items that were correctly identified but that were reported in an incorrect order. These latter errors were shown to be distinct from errors resulting from feature perturbations.

A comparative study of trauma and posttraumatic stress disorder prevalence in epilepsy patients and psychogenic nonepileptic seizure patients.

Summary: Purpose: This study tests the hypothesis that trauma histories, including histories of physical and sexual abuse, and posttraumatic stress disorder (PTSD) are more prevalent in psychogenic non–epileptic seizure (NES) patients than in epilepsy patients.

Retinal location and string position as important variables in visual information processing

Three experiments were conducted to isolate the effects of retinal locus and string position in tachistoscopic letter recognition. Retinal locus proved to be an important variable even when its range was restricted to less than a degree from the center of the fovea. Performance was maximal at the center of the fovea, dropping off rapidly to about 1.5 deg from the center. From that distance on, the decline in performance was quite gradual. String position was also an important factor. Retinal locus and string position interacted in such a way that the end positions were less affected by retinal locus than the middle positions. It was also found that processing order, as distinct from report order, was a significant component of the string position effect.