| Jun Bao and Richard Shillcock |
| Tue 20 Sep 2016, 11:00 - 12:00 |
| IF 4.31/4.33 |
If you have a question about this talk, please contact: Gareth Beedham (gbeedham)
Richard Shillcock, Florian Bolenz, Sarnali Basak & Alasdair Morgan
Understanding sex differences in IQ: ‘Hemisphericity’ in computational models simulates greater male variance in IQ
How can we account for the reported difference in standard deviation between comprehensive male and female distributions of IQ scores
(Deary et al., 2003)? Males are significantly over-represented at both ends of the scale. Perhaps the data reflect two qualitatively different underlying distributions? Perhaps there is a genetic explanation based on the X chromosome? An answer has not been forthcoming.
We report a series of neural network simulations in which the observed data are robustly generated by variation in a single dimension. The status of the modelling and the entities in the model are discussed, together with directions for future work.
Deary, I. J., Thorpe, G., Wilson, V., Starr, J. M., & Whalley, L. J. (2003). Population sex differences in IQ at age 11: The Scottish mental survey 1932. Intelligence, 31(6), 533-542.
Jun Bao
Physics of Eye Movements in Large Corpus of reading behaviours
This project aims to provide a reliable event detection algorithm for eye movement data in reading. Event detection is the process of parsing raw gaze coordinates (sample points produced by an eyetracker) into segments of eye movement events (saccade, fixation, smooth pursuit, blink, etc.). Previously researchers did not realise there is a wobbling of pupil and lens at the end of each saccade (postsaccadic oscillation, lasting for 30ms-100ms); the current research tries to analyse the idiosyncratic of postsaccadic oscillation for both the left and right eye, and then singles out postsaccadic oscillation as an independent eye event, hence delineates new boundaries for saccade and fixation. In the process of research, we extracted saccade-to-fixation trajectory from raw eye tracker data, applied time adjustment and trajectory outlier detection algorithm to sets of trajectories, and then used a damping system to model eye movement trajectories in an attempt to decompose the trajectories into pupil movement and eyeball movement.