In addition to employing eye-tracking to control for maintenance of fixation, we examined small fixational eye movements and microsaccades during presentation of stimuli. Recent work suggests that these miniature eye movements underlie important perceptual functions (Martinez-Conde et al., 2006). If participants with an ASD exhibit problems in eye movement Selumetinib control, then it is possible that not only saccades but also fixational eye movements would be affected. Rate, amplitude and orientation of microsaccades during fixation were determined using a publicly available Matlab toolbox (Engbert & Kliegl, 2003). In addition, we determined

the standard deviation of eye-gaze along the horizontal and vertical axes for valid fixation trials. This measure complements the microsaccade analysis regarding

slower movements of the eye, which do not reach the velocity threshold of the microsaccade analysis. Based on findings in previous VEP studies on visual processing in ASD, we mainly focus our analysis on the timeframe of the P1 component (e.g. Boeschoten et al., 2007; Sutherland & Crewther, 2010). The amplitudes of VEP and VESPA P1 components were examined using an independent sample t-test. As response latencies increase with decreasing stimulus contrast (Reich et al., 2001), we expected delayed responses to Magno VESPA stimuli. Therefore, we did not define one P1 timeframe for all experimental conditions but used a 20-ms window centered on the individual peak amplitude Branched chain aminotransferase in the time range 130–170 ms for Magno VESPA peripheral and 90–130 ms for all other conditions. learn more In addition to planned comparisons, we ran post hoc statistical tests for detecting additional differences (for other VEP/ VESPA components such as C1, N1). For the first 300 ms of the evoked responses, running two-tailed t-tests between the experimental groups were applied. A time range of the response was termed significantly different if at least nine consecutive time points were different with P < 0.05, based on the autocorrelation of the signals of interest (Guthrie & Buchwald,

1991). Evoked responses were further examined by using relative amplitudes. To do so, the amplitude of the peripheral response in the P1 timeframe was divided by the one for central stimulation. Such a transformation removed any influence of the variability in individual response amplitudes and served as an index for peripheral visual processing. Relative peripheral P1 amplitudes as well as measures of behavioral performance, reaction time and eye movements were examined using mixed linear models (spss 21.0, IBM Corp., Armonk, NY) using experimental group, age (rounded to integer), stimulus type and laterality (where applicable) as factors. Main effects and two-way interactions between factors were computed.