Mr Patrick Craston
Research Student
|
|
|
Publications
My publications are available from the Computer Science department publications repository.
PhD Project Summary
I am interested in how humans allocate attention over time. Although we are constantly processing and at least temporarily storing all information that is transferred from sensory inputs, how are we able to separate the relevant from the irrelevant?
During my Phd, my main focus was on the Attentional Blink phenomenon [Raymond et al., 1992], which seems to reflect some of the underlying principles of how humans distribute attentional resources over time.
This paradigm is typically
structured as follows: A person is exposed to a stream of digits,
that are presented in the same spatial location and at a very rapid
rate (typically 10-20 items/second). There are two target letters in this stream, which are
to be reported at the end of the stream. Performance in reporting a second
target (T2) is strongly impaired if it follows an identified and reported first target (T1) within a time window of 200-600 ms. Nevertheless, if T2 immediately follows T1 (within 100ms - at lag 1), T2 detection is almost unaffected, which has been called
lag 1 sparing. However, in
this case T1 detection performance suffers. Therefore, at short lags,
there is an accuracy trade-off between detection of the
targets, rather
than sparing, as usually suggested.
For a simplified demo of a typical Attentional Blink experiment, click here.
The ST2 Model of temporal attentional and working memory, developed by Howard Bowman and Brad Wyble, reproduces the empirical results mentioned above. It is based on the Dual Stage Model by Chun and Potter (1995) and the Types-Tokens Account of Working Memory proposed by Nancy Kanwisher (1987). During my PhD, I tested predictions derived from the ST2 Model using EEG techniques, in particular by means of Event-Related Potential (ERP) and time-frequency analysis. Furthermore, we generated virtual ERPs using the neural network implementation of the ST2 Model and compared these to results from EEG studies.
I contributed to the EPSRC Salience Project.
Thesis
Download my PhD thesis as pdf (12MB)
Recent publications
- The attentional blink reveals serial working memory encoding: Evidence from virtual & human event-related potentials. Patrick Craston, Brad Wyble, Srivas Chennu, and Howard Bowman. Journal of Cognitive Neuroscience, 21(3), 550-566, 2009.
- Attention Increases the Temporal Precision of Conscious Perception: Verifying the Neural-ST2 Model Srivas Chennu, Patrick Craston, Brad Wyble, and Howard Bowman. PLoS Computational Biology, 2009, in press
- The influence of target discriminability on the time course of attentional selection. Srivas Chennu, Patrick Craston, Brad Wyble, and Howard Bowman. In N. A. Taatgen and H. van Rijn (editors), Proceedings of the 31st Annual Conference of the Cognitive Science Society, Cognitive Science Society, 2009, 1506-1511
- The delayed consolidation hypothesis of all-or-none conscious perception during the attentional blink, applying the ST2 framework. Howard Bowman, Patrick Craston, Srivas Chennu and Brad Wyble. In N. A. Taatgen and H. van Rijn (editors), Proceedings of the 31st Annual Conference of the Cognitive Science Society, Cognitive Science Society, 2009, 2152-2157
- A reciprocal relationship between bottom-up trace strength and the attentional blink bottleneck: Relating the LC-NE and ST2 models. H. Bowman, B. Wyble, S. Chennu, and P. Craston. Brain Research, 1202:25-42, April 2008.
- Transient attentional enhancement during the attentional blink: ERP correlates of the ST2 model. Srivas Chennu, Patrick Craston, Brad Wyble, and Howard Bowman. In Robert M. French and Elizabeth Thomas (editors), From Associations to Rules: Connectionist Models of Behavior and Cognition, volume 17 of Progress in Neural Processing, January 2008, World Scientific.
Miscellaneous
- I was supervised by Prof. Howard Bowman
- My PhD was funded by the Engineering and Physical Sciences Research Council
Research Interests
I am a member of the following research groups:

