Analysis of autonomic reaction patterns

Using multilevel models for the analysis of event-related potentials

Multilevel modeling (MLM) is becoming increasingly accessible and popular in the analysis of event-related potentials (ERPs). In this article, we review the benefits of MLM for analyzing psychophysiological data, which often contains repeated observations within participants, and introduce some of the decision-making points in the analytic process, including how to set up the data set, specify the model, conduct hypothesis tests, and visualize the model estimates. We highlight how the use of MLM can extend the types of theoretical questions that can be answered using ERPs, including investigations of how ERPs vary meaningfully across trials within a testing session. We also address reporting practices and provide tools to calculate effect sizes and simulate power curves. Ultimately, we hope this review contributes to emerging best practices for the use of MLM with psychophysiological data.

Estimation of Treatment Effect and Measurement of Change

A variety of statistical procedures have been recommended for the evaluation of treatment effects in correlated sequential measures. Analysis of covariance is generally favored and recognized as an appropriate procedure although other techniques have been suggested as simpler and essentially equivalent. This paper examines several of these techniques and reports results of a Monte Carlo study comparing Type I error rate per experiment for the following procedures (1) analysis of variance of difference scores, (2) analysis of variance of final scores, and (3) analysis of covariance. It is shown that of these procedures only analysis of covariance produces error rates close to chance. Analysis of covariance is recommended for the assessment of treatment effects in data with pretreatment-posttreatment measures of the dependent variable and where experimental control is not possible.

Probing the response selection bottleneck with a cardiac measure: individual differences in strategy for a psychological refractory period task

We examined the coordination of processing streams when two reaction stimuli are presented with minimal temporal separation (the psychological refractory period, PRP, paradigm). We tested the hypothesis that individuals that grouped responses to the two stimuli would schedule response preparation later than those not-grouping their responses would. Performance measures were combined with a cardiac measure interpreted as an index of response inhibition. Forty-two college-aged participants performed an auditory choice task followed in 50, 100 or 400 ms by a visual choice reaction time (RT) task. Performance results replicated prior work. The timing of cardiac inter-beat intervals (IBIs) showed that participants grouping their reactions did appear to delay response preparation; while Non-Groupers promptly initiated response selection for the initial task. These results generally supported the importance of preparation for understanding the coordination of processing streams.

Relationships between the type A behavior pattern and phasic heart rate responses during a reaction time task

Type A and B college students performed a signalled reaction time task with a 17-s fixed preparatory internal (PI) while heart rate was recorded. The analysis of the first half of the PI indicated a triphasic heart rate response (deceleration-acceleration-deceleration) to the warning signal, followed by a return toward baseline. No group differences were found. During the latter half of the PI, the Type B subjects exhibited a time-locked heart rate deceleration which reached its maximum at the second immediately prior to the onset of the imperative stimulus. The Type A subjects exhibited maximum deceleration earlier in the PI. These findings are consistent with previous work indicating that Type A's have difficulty maintaining their attention over relatively long PI's of reaction time tasks.

Effect of the orienting response on sensory discriminability

The effects of orienting responses to different types of stimuli on acquisition of information were studied. Subjects underwent a standard habituation series of 15 trials. On Trial 16, they received one of following three stimuli: (a) no-change stimulus--same stimulus as habituation stimulus, (b) innocuous-change stimulus, (c) significant-change stimulus--the subject's own name. These orienting stimuli were followed 500 msec. later by an imperative stimulus (100 msec.) which contained a number of Chinese characters. The subjects were then unexpectedly asked to recall and recognize these characters. Recall and recognition were consistently superior in conditions with significant and innocuous change relative to no change, which indicated that the orienting response to unexpected stimuli is related to a generalized perceptual enhancement. The data pose problems for the position that the orienting response reflects only the passive 'registration' of the eliciting stimulus information. While digital pulse-amplitude data showed no difference among the conditions for the change trial, these conditions were differentiated by electrodermal and cardiac changes.

The development and validation of an analytical training program for medical suturing

Based on attitude questionnaires and film analysis. three training simulators, electromechanical (E), perceptual (P), and (E) + (P),were developed to teach suturing to medical students. Training effectiveness on these simulators was compared to the traditional training method. Intergroup comparisons were obtained by having nine subjects from each of the three experimental groups, subsequent to training, practice on porcine skin by placing 35 sutures similar to those o{ the traditional group (N = 9). Results indicate that all three experimental training methods provided for significantly better quality of performance, briefer training times. and less psychophysiological stress (as measured by sinus arrhythmia, chin and neck EMG, and skin conductance) than the traditional training method.

Measurement of the evoked cardiac response: the problem of prestimulus variability

This paper discusses the problem of prestimulus variability in cardiac activity and its implications for the measurement of evoked cardiac responses (ECRs). Two methods which have been proposed to reduce error arising from respiratory sinus arrhythmis (SA) were reviewed critically. The first approach attempts to control only for the error associated with the sampling of the prestimulus level. This is considered inappropriate since it does not take into account the continuation of SA into the poststimulus period. The second approach reviewed was time series analysis which provides an elegant statistical solution of the problem. Unfortunately, the application of time series analysis has not yet been evaluated for adult cardiac data. A third approach which utilizes an adjustment for SA by employing the actual prestimulus values of an SA cycle was proposed. The application of this technique in the case of a pseudostimulus demonstrated that correction for SA results in significantly smaller 'responses' with less variance than does a conventional procedure which does not take prestimulus variability into account.

Subvocalization: aural and EMG feedback in reading

Speech is a complex skill with auditory and kinesthetic components—sensory and motor—coordinated inter- and intramodally. In reading, vision joins these modalities. Functional relations of variables in reading for arithmetic problem-solving were explored for individuals who show much (HVs) compared with those who show little (LVs) speech-muscle activity (subvocalization) during silent reading. Higher relative sound intensities were used by LVs than by HVs in normal reading aloud but not when reading with delayed auditory feedback (DAF). Feedback training for the self-regulated reduction of electromyographic activity of speech muscle during silent reading was effective for both HVs and LVs. After feedback training the difference in relative sound intensities between HVs and LVs was eliminated in normal reading aloud and HVs showed smaller increases than LVs when reading with DAF. Latency to correct answer in silent problem-solving was shorter for LVs than for HVs before but not after feedback training. Variations in heart rate, eye-movement rate, regressive eye movements, reading speed, rate of integrated speech peaks, syllable rate, eye-voice span, eye movements per syllable, and correct answers were related to experimental conditions but not different for HVs and LVs. Data on the interrelations of variables are presented to suggest some tentative hypotheses. The correlation of eye-voice span with speech-muscle activity was related to levels of subvocalization in normal reading aloud but modified by DAF and feedback training. The measurement of speech-muscle activity, feedback training, general and developmental differences in the use of auditory versus kinesthetic cues by HVs and LVs, and implications of subvocalization for speech, memory, thought, and reading are discussed.