response duration, but not with EPrime

[Sudevan, Padmanabhan]

Hello all,

Some general responses to this discussion thread:

First, regarding Matlab, I believe Geoff Loftus' lab at the University
of Washington uses a version of Matlab, along with some of their
modules, for experimentation. Although the Toolkits were written
separately from Matlab, my recollection is that they are available to
any researcher.

Second, I too understand that PST needs to be able to get revenues in
order to continue doing what they are doing, and I don't feel that the
costs involved, for a researcher at a small university (like myself) are
exorbitant. On the other hand, I think ClassMate with its currently
proposed semester fees will pose some difficulty for a campus like mine
because students want to keep textbook/software costs down and we have a
text rental system. Perhaps one could think of some arrangement where
the software could be made available to the university at an annual cost
for the use of a certain number of students -- and the university could
then charge the students a reduced fee. Walter Schneider did say at
Psychonomics in Minneapolis that they were still exploring alternate
ways of charging users for ClassMate.

Third, regarding Yoav's question on alternate systems, there is Intuit
Systems Millisecond which enables folks to write their own code. The
problem, as Brian points out, is that not everyone would want to write
their own code. Also E-Prime is a general system that is meant to run on
a variety of machines, and this generality is bought at some sacrifice,
in terms of the range of things that it can do -- viz., standalone
executable scripts, with attendant issues of driver compatibility etc.


Sudevan



P Sudevan
Professor of Psychology
Chair, Faculty Senate, UWSP


-----Original Message-----
From: eprime at mail.talkbank.org [mailto:eprime at mail.talkbank.org] On
Behalf Of Paul Gr
Sent: Sunday, January 16, 2005 4:48 PM
To: eprime at mail.talkbank.org
Cc: baranan at post.tau.ac.il
Subject: RE: response duration, but not with EPrime


Hi,

I'm not sure if it appropriate to discuss non-eprime issues in this
mailing 
list, but I think you're right when you say that deploying or
distributing 
runtime versions of EPrime scripts may be difficult or even impossible.
I 
also realize that software developing companies such as PST require
people 
buying their products to be able to continue their business. I really
hope 
that PST will offer royalty free runtime licences in the near future, so

EPrime users are not restricted anymore to run their scripts at their
own 
sites. Or even better: being able to create standalone exe-versions of 
scripts.

However, beside the licensing strategy, there are also some technical
issues 
that require attention. For instance, the current version of EPrime uses

special drivers to handle response devices. These drivers are required
for 
optimal response processing. So, even if there was a possibility to
create 
standalone executables, you would still have to deploy a proper runtime 
environment. (And don't forget DirectX..., which is a 'must have' for
this 
type of application.)

As far as I know, there are no commercially available alternatives,
which 
don't have the same deployment issues as EPrime. So, for real standalone

executables, you would have to get into  a suitable programming
environment 
and enter the low-level world of WIN32, GDI and DirectX APIs. I know of
some 
projects at other universities, where colleagues are developing their
own 
'time critical' software libraries. However, compared to using EPrime
and 
the like, you really require in-depth expertise and lots of time...

Let's hope PST (and other's) come up with a customer-friendly solution
for 
this problem. If not, then I expect some kind of open source project to 
arise within a year or so.

Best,
Paul

P.S.: about matlab and alternatives... beside accurate response
measurement, 
you would also require accurate stimulus delivery. Not sure if matlab
will 
pass that test.


>From: yoav Bar-Anan <baranan at post.tau.ac.il>
>To: eprime at mail.talkbank.org
>Subject: response duration, but not with EPrime
>Date: Sun, 16 Jan 2005 15:39:05 +0200
>
>Hi,
>
>
>
>I often prefer not to use E-Prime for programming experiments, because 
>a
>simple executable written in C++, VB or even HTML, is easier to deploy.

>However, when an experiment requires response-duration recording and 
>accurate timing of events, I always resort to E-Prime, because I know
they 
>invested much effort in accuracy.
>
>
>
>I wonder whether anyone can suggest other environments which offer 
>accurate
>timing (or tools that can easily produce accurate timing) that can be
ran 
>on most computers without installing anything but the experiment file.
>
>
>
>On the same note: What do you think about Matlab as an alternative for
>E-Prime?
>
>
>
>Thank you,
>
>Yoav

_________________________________________________________________
Nooit ongewenste berichten ontvangen: gebruik MSN Messenger 
http://messenger.msn.nl/