[ACT-R-users] Inter-subject variability in declarative memory

[Phil Beaman]

...but both approaches are valid on their own terms. It makes no sense to average over individual subject strategies, where these do differ, but it is an interesting question to ask whether individual differences in parameter values or in strategy choice provide the most parsimonious account a priori of observed subject differences. In some research areas (eg cognitive development) the assumption of a qualitative change between one population (age group) and another appears to be the default account of performance differences between the two groups, and this is sometimes worth questioning (recent papers by Jarrold and colleagues on the development of phonological representations in short-term memory, for example).  Similarly, if parameter differences are always and only the means used to simulate individual differences on a task, this is also worth challenging if such parameter manipulation is used unthinkingly.

P

Dr Philip Beaman AFBPsS
Senior Lecturer (Associate Professor) in Cognitive Psychology
School of Psychology & Clinical Language Sciences
Earley Gate, Whiteknights
Reading RG6 6AL

Tel: +44 (0) 118 3787637
Fax: +44 (0) 118 3786715

Staff web-page:
http://www.reading.ac.uk/psychology/about/staff/c-p-beaman.aspx
Cognitive Science Lab:
http://www.personal.rdg.ac.uk/~sxs98cpb/philip_beaman.htm

"I need truth, and some aspirin" -- Fernando Pessoa, 1931

________________________________________
From: act-r-users-bounces at act-r.psy.cmu.edu [act-r-users-bounces at act-r.psy.cmu.edu] on behalf of Coty Gonzalez [conzalez at andrew.cmu.edu]
Sent: 12 July 2012 11:30
To: Niels Taatgen
Cc: act-r-users at act-r.psy.cmu.edu
Subject: Re: [ACT-R-users] Inter-subject variability in declarative memory

Beyond "tweaking parameters": the point is to understand the contributions of each parameter and what each of them really mean for different populations and tasks.

Sent from Coty's iPad excuse typos and brevity.

On Jul 12, 2012, at 4:25 AM, Niels Taatgen <n.a.taatgen at rug.nl> wrote:

> David,
> Tweaking parameters to fit individual differences is a common approach, as the earlier responses have already indicated, but another option is to look at differences in working memory strategies. I strongly believe that many individual differences in cognitive control, including working memory, are due to people applying different strategies. Transfer experiments support this idea by showing that, for example, training working memory not only increases working memory capacity, but also reduces Stroop interference. I presented models of this during ICCM in Berlin, and have a draft paper about it (which I can mail you if you are interested).
> Niels
> On Jul 11, 2012, at 9:34 PM, David Reitter wrote:
>
>> Most ACT-R models seem to assume a "standard mind", an average cognitive system, with the same (global) parameters set for each simulated subject.  Particularly, the base-level decay parameter alpha (:bll) is not one to mess with - modelers usually leave it at 0.5.
>>
>> That said, humans differ in their memory abilities.
>>
>> I am interested in any empirical and modeling work that explores the distribution of declarative memory parameters (such as alpha and the rather model-dependent base-level constant).  Can we assume a normal distribution around alpha=0.5?  How wide is it?
>>
>> Much work has been done on short-term and working-memory capacity.  However, if you are aware of work relating ACT-R with its integrated perspective on short-term and long-term memory and inter-subject variability, please let me know.  If I get a lot of answers, I will post a summary here.
>>
>> Thanks,
>> David
>>
>>
>>
>> --
>> Dr. David Reitter
>> Research Psychologist, Department of Psychology, Carnegie Mellon University
>> 08/2012: Assistant Professor of Information Sciences and Technology, Penn State University
>> http://www.david-reitter.com
>>
>>
>> _______________________________________________
>> ACT-R-users mailing list
>> ACT-R-users at act-r.psy.cmu.edu
>> http://act-r.psy.cmu.edu/mailman/listinfo/act-r-users
>
> ===============================================
> Niels Taatgen - Professor
> University of Groningen, Artificial Intelligence
> web: http://www.ai.rug.nl/~niels     email: niels at ai.rug.nl
> Telephone: +31 50 3636435
> ===============================================
>
>
>
>
>
>
>
> _______________________________________________
> ACT-R-users mailing list
> ACT-R-users at act-r.psy.cmu.edu
> http://act-r.psy.cmu.edu/mailman/listinfo/act-r-users
>

_______________________________________________
ACT-R-users mailing list
ACT-R-users at act-r.psy.cmu.edu
http://act-r.psy.cmu.edu/mailman/listinfo/act-r-users