Stability of Variables Derived From Measures of Multisensory Function in Children With Autism Spectrum Disorder

Abstract

Children with autism spectrum disorder (ASD) display differences in multisensory function as quantified by several different measures. This study estimated the stability of variables derived from commonly used measures of multisensory function in school-aged children with ASD. Participants completed: a simultaneity judgment task for audiovisual speech, tasks designed to elicit the McGurk effect, listening-in-noise tasks, electroencephalographic recordings, and eye-tracking tasks. Results indicate the stability of indices derived from tasks tapping multisensory processing is variable. These findings have important implications for measurement in future research. Averaging scores across repeated observations will often be required to obtain acceptably stable estimates and, thus, to increase the likelihood of detecting effects of interest, as it relates to multisensory processing in children with ASD.

Keywords: autism; measurement; multisensory; psychometrics; reliability; stability.

Figure 1.

Representative temporal binding window (TBW) derived for a participant in the study from a single observation. The proportions of perceived synchrony are normalized, such that the maximum value is set to 1, and are fit to two psychometric functions, one for trials wherein visual stimuli precede the auditory stimuli (right line) and one for trials wherein auditory stimuli precede the visual stimuli (left line). The vertical dotted lines represent the point at which each line reaches the .75 threshold for perceived synchrony (the horizontal dotted line; i.e., –25.0 ms and 200.0 ms for the depicted example). The TBW is the distance between these two values (the distance between the two vertical dotted lines; i.e., 225.0 ms).

Figure 2.

Example of the areas of interest (AOIs) used in the eye-tracking measure (i.e., in the English infant-directed multisensory speech condition). Proportion of time looking to each AOI was calculated as the time spent looking at the area (mouth or eyes)/time spent looking at the broader face during stimulus presentation.

Figure 3.

Stability of variables derived from eye-tracking measure of attention to multisensory speech presented in (A) English and (B) Spanish. Variables derived from eye tracking have high stability (g = .8) with one to two observations. ID = Infant-directed speech; AD = Adult-directed speech. Generalizability coefficients are observed for two samples (i.e., day 1 and 2) and projected beyond two samples based on variance estimates derived via ANOVA carried out on observed data. The a priori threshold for acceptable stability was set at g = .8 (dashed line).

Figure 4.

Stability of variables derived from psychophysics measures. Two variables derived from psychophysical measures met our a priori threshold for stability of g = .8 within two observations. AV = audiovisual; ID = identification accuracy; TBW = temporal binding window. Generalizability coefficients are observed for two samples (i.e., day 1 and 2) and projected beyond two samples based on variance estimates derived via ANOVA on observed data. The a priori threshold for acceptable stability was set at g =.8 (dashed line).

Figure 5.

Stability of variables derived from event-related potential (ERP) measure. P2 amplitude is the most stable of the ERP variables in both the (A) auditory-only (AO) condition and (B) audiovisual (AV) conditions. N1 = negative deflection of the ERP occurring in the a priori specified timeframe between 100 ms and 140 ms post-stimulus onset as measured at Cz, P2 = positive deflection of the ERP occurring in the a priori specified timeframe between 160 ms and 240 ms as measured at Cz. Generalizability coefficients are observed for two samples (i.e., day 1 and 2) and projected beyond two samples based on variance estimates derived via ANOVA carried out on observed data. The a priori threshold for acceptable stability was set at g = .8 (dashed line). Note that the consistent g coefficients of 0 for AO N1 amplitude and AV P2 latency suggest that it may not be possible to obtain stable estimates for these variables in school-age children with ASD even with repeated sampling.