compiled responses for processing data in eprime

[Michael J. Crowley]

Dear list,

Thank you all for your thorough responses.  What follows is a compilation of the
various methods each of you suggested for analyzing raw data from E-Prime. 
We’ve appended the responses from all those who contributed to this query. 
Tony Zuccolotto’s response addresses the main points for typical data handling
(merge, exclude/filter unnecessary variables, reuse preferred column
arrangement as data files are added), but the many users offered specific
solutions and extensions that may be of interest to the list.  I’ve appended
each of these below.

Thanks to you all for your thorough responses.

Mike Crowley

Michael J. Crowley, Ph.D.
Yale Child Study Center
230 South Frontage Rd.
New Haven, CT 06520


From Kristie Willhoit:

Date:  Thu, 23 Jun 2005 16:52:54 -0500
From:  Kristie Willhoit <kwillhoit at uwf.edu>
To:  eprime at mail.talkbank.org
Subject:  RE: processing eprime data

I, too, use Excel and SPSS for analyzing my data, but I prep it using
DataAid first.  You can use the Filter function to remove any NULL values
from a specific column by setting the range not equal to NULL.  Note,
however, that this will exclude all data in that trial from analysis.  I
also use the analyze function in DataAid to run the analyses I need
(strictly reaction and movement time means in my case), then export the
results to Excel, then import that file to SPSS (after making it
SPSS-friendly, as Dave mentioned).  Hope this helps!

Kristie

Kristie Willhoit
Research Associate
University of West Florida
Division of Health, Leisure, and Exercise Science


From Leisha Wharfield:

Date:  Fri, 24 Jun 2005 11:53:09 -0700
From:  Leisha Wharfield <leisha at decisionresearch.org>
Cc:  eprime at mail.talkbank.org
Subject:  Re: processing eprime data

I do the same. Filter, copy to Excel. Our data analyst imports to SPSS &
takes it from there.

Leisha


From Jordan Bigio:

From:  Jordan Bigio <j-bigio at northwestern.edu>
To:  Tony Zuccolotto <anthony.zuccolotto at pstnet.com>, Paul Warren
<Paul.Warren at vuw.ac.nz>, eprime at mail.talkbank.org
Subject:  RE: processing eprime data

I do this as well. Hide columns I don't want, then arrange the ones I do.
Copy. Paste to excel. Save, and run in SPSS with SPSS syntax for stats we
want. Then, save outfiles as excel files, and paste them back into my
behavioral stats file (excel as well).

Jordan

From Georg Odenthal:

From:  Georg Odenthal <odenthal at soz.psychologie.uni-konstanz.de>
To:  eprime at mail.talkbank.org
Subject:  RE: processing eprime data

To be honest I usually never use the data files E-Prime creates
automatically. Since I work a lot with E-Basic to create special
questionnaires like open ended text input, text fields or variable
scales or animation I usually open a new textfile for writing and
then create an E-Basic code that puts all the data I need into this
file after each trial. I sometimes go back to the edat files to
double-check a timing issue or whether the response keys were coded
correctly, but most of the time I don't care much about the edat files.

A lot of the data I collect in my studies are variables that are
defined temporarily in an E-Basic code object. To get E-Prime to put
these variable out to the .edat file you need to add the attribute to
the current object and then update the value of the attribute every
time something in E-Basic or a response has changed this value. This
can be quite tedious.

Since I often write programs for other people the creation of an own
datafile is necessary. The people I write the programs for like it if
they can just open SPSS and then import a text file with one row for
each subject and X columns for the variables or the response data.
It's also more convenient for me, because then I don't have to
convert the edat files into something the researcher finds more useful.

Georg

Georg Odenthal, Dipl.-Psych.
Social Psychology and Motivation
Box D39
University of Konstanz
78457 Konstanz, Germany


From Tony Zuccolotto:

Date:  Thu, 23 Jun 2005 17:44:25 -0400
From:  Tony Zuccolotto <anthony.zuccolotto at pstnet.com>
To:  Paul Warren <Paul.Warren at vuw.ac.nz>, eprime at mail.talkbank.org
Subject:  RE: processing eprime data

If you are not familiar with the feature, you may be able to save a
little bit of work by using E-DataAid's "Tools|Arrange Columns..."
dialog.

Specifically, you can click the Remove All button on the dialog to
remove (hide) all the columns, optionally alphabetize the list of
variables, and then double click just the variables that you want to add
them back in.

If you save the data file it will bring back that same column
arrangement the next time you load it, and you can then use the File |
Export... option to send the data elsewhere.

If you do this early on in your master merged file, save it, and keep
reusing it then E-DataAid should save your preferred column arrangement
and keep it as the default for that merged file as new data files are
added in.

..not totally automated, but may save a few repetitive steps or a lot
of adding and deleting using Excel.

Tony


From Stephani Foraker:

Date:  Thu, 23 Jun 2005 16:42:43 -0500
From:  Stephani Foraker <sforaker at uchicago.edu>
To:  "Michael J. Crowley" <michael.crowley at yale.edu>
Subject:  Re: processing eprime data

I like to use data aid to pull out only my experimental items and the
columns of interest from the merged file - use the 'hide columns' feature
and filter feature to show only the info you want. I often use a condition
tag in the eprime program that runs the experiment, log that, and can do a
checklist for filtering on just the conditions that I want to look at. Then
I usually export that as a text, tab-delimited file which can be imported
to excel or spss easily. I usually do some preprocessing in excel, whose
pivot table function is fabulous for looking at various properties of the
data, and if working with reaction time type measures I can sort based on
min or max times, etc. The pivot tables also provide an excellent way to
produce just the cell means in just the configuration that you want for any
spss or other statistical package analyses - select, copy and paste into
the data anlaysis spreadsheet. And lastly, excel's table formatting is much
more flexible than spss's if that's what you use, so I like having my data
and means available in excel, too.

Stephani.


From Lena Katz:

Date:  Thu, 23 Jun 2005 17:35:11 -0400
From:  "Katz, Lena B." <katzlb at upmc.edu>
To:  eprime at mail.talkbank.org
Subject:  RE: processing eprime data

I haven't actually started collecting data yet (sorry for all the code
questions!).  But I'd love to see the source code for Dave's macros... we're
probably not all going to want to strip out the same data, but having a good
example would be nice.

Lena


From Paul Warren:

Date:  Fri, 24 Jun 2005 09:28:58 +1200
From:  Paul Warren <Paul.Warren at vuw.ac.nz>
To:  eprime at mail.talkbank.org
Subject:  RE: processing eprime data

I use a similar strategy to Dave, i.e. E-merge then Excel and then stats either
in Excel (with the statistiXL product) or in SPSS. But I haven't come to grips
with macros in Excel for stripping out the dross. Maybe Dave could make his
macros available?

Paul

Dr Paul Warren
School of Linguistics and Applied Language Studies
Victoria University of Wellington
New Zealand

tel +64 4 463 5631/5600
fax +64 4 463 5604


From Mark Casteel:

Date:  Thu, 23 Jun 2005 17:23:31 -0400
From:  "Mark A. Casteel" <mac13 at psu.edu>
To:  "Michael J. Crowley" <michael.crowley at yale.edu>
Subject:  Re: processing eprime data

One easy way to accomplish this is to use the Filter available
from the tools menu in E-DataAid. You can select as many filters as you
like. For each, hit the checklist button, and then simply check all of
those conditions which are applicable. You can then export your data to
SPSS or an EXCEL spreadsheet, and it will be limited to only those values
that are selected via the filter.

Mark


From David Hairston:

Date:  Thu, 23 Jun 2005 17:19:59 -0400
From:  David Hairston <dhair at wfubmc.edu>
To:  "Michael J. Crowley" <michael.crowley at yale.edu>, eprime at mail.talkbank.org
Subject:  RE: processing eprime data

As a general rule, I do not use E-Data Aid at all; I use SPSS for 90% of
my analyses, and Excel for transitioning the data between E-Data Aid and
SPSS. Personally since I know that I'll need everything in SPSS for
eventual stats, groups for figures, etc etc, I find it easier to go
there directly.

So, when the experiment is done, I immediately group it via E-Merge,
then export into Excel Test format... then I open it via Excel, and go
through and trim out all the stuff I don't want (Mike I know what you
mean about a LOT of irrelevant stuff). The nice thing about doing this
in Excel is that you can easily record a macro to delete all the things
you don't want, so you can do it again later w/ a few clicks. This also
gives me a chance to change variable names to be SPSS-friendly, etc.
Then I just port it back out into tab-delineated text, and import into
SPSS from there.

I, too, am curious to see what other folks do...

Unfortunately as a result I do not know how to handle the block/ NULL
data in E-data-Aid, sorry.

Dave

W. David Hairston
Neurobiology and Anatomy

Wake Forest University School of Medicine
Winston-Salem, NC 27157
(336) 716-4481 (lab)
http://www.wfubmc.edu/nba/XMODAL/






Please be aware that email communication can be
intercepted in transmission or misdirected. Please
consider communicating any sensitive information
by telephone, fax or mail. The information
contained in this message may be privileged and
confidential. If you are NOT the intended
recipient, please notify the sender immediately
with a copy to hipaa.security at yale.edu and destroy
this message.