FBI’s Most Wanted neuroscientist on imitation

Aafia Siddiqui was the FBI’s Most Wanted Woman for several years and is currently in US custody in New York, awaiting trial on charges that she is a terrorist and member of Al-Qaeda.

She is also a neuroscientist and co-authored a scientific paper in 2005 on the cognitive science of imitation learning.

Before her recent capture, which some sources claim may have actually happened five years ago in Pakistan, Siddiqui completed a PhD entitled ‘Separating the components of imitation’ at Brandeis University in the US.

Before that, she completed a Masters degree in neuroscience, also on imitation learning, and her 2005 paper is based on her work for this degree.

The paper describes three experiments that attempt to understand how our learning of seen actions is affected by delays, memory interference and visual interference.

Each experiment involved a pale red disc that followed an ‘invisible track’ on a computer monitor. The sort of track is illustrated in the diagram of the left, although in the actual experiment the participants just saw the disc.

In Siddiqui’s experiment, one group of participants used a trackpad to ensure that the cursor was within the disc at all times (a classic pursuit-tracking task), while another group had to wait until the disc had followed the route before trying to reproduce it from memory.

To look at the effect of complexity on imitation learning, some routes had only three straight lines, while others had up to seven.

Furthermore, some routes were repeated, while others appeared only once. This allowed the researchers to compare learning for identical routes (specific learning) with learning for the general task (skill learning).

The results showed that, unsurprisingly, participants were better at reproducing the simpler routes. What was more surprisingly though, was that practice-related improvement was only seen when participants watched the whole movement before starting, and then only on routes that were repeated.

Intriguingly, when interviewed after the experiment, the participants had no idea some routes were presented more than once, suggesting that this learning occurred without any conscious involvement.

A further experiment showed that delays of up to 6 seconds barely affected performance and that interfering with short-term memory by getting participants to do maths problems only made them a little worse.

Finally, the researchers ran an experiment where the disc appeared only at the beginning and end of each straight line, or when it was turning a corner. This had virtually no impact on performance. Participants were almost equally as good with much less information.

The research helps us understand the limits of learning when we need to copy a certain action sequence, be this tying shoelaces, swinging a golf club or learning tai chi.

The study suggests that for short action sequences we may be better off waiting until we watch the whole thing before attempting it ourselves.

Link to PubMed entry for paper.
Link to library record for Siddiqui’s PhD thesis.

One Comment

  1. Greg
    Posted October 15, 2010 at 3:04 pm | Permalink

    FYI, the correct URL for the thesis is: http://loufind.brandeis.edu/vufind/Record/MBB01000000000000000930469


Post a Comment

Required fields are marked *
*
*

Follow

Get every new post delivered to your Inbox.

Join 23,875 other followers