E-Prime 2.0 Data Features

[Paul Gr]

Hello Brandon,

There are a few practical issues about E-DataAid that could be improved:

1) E-DataAid can become very sloooow when using filters. Especially when 
used in combination with the analysis option.

2) When exporting analysis tables to SPSS I always have to rename column 
names. The latest version of SPSS is somewhat improved on the maximum 
allowed variable name, but dashes are just not allowed and filtered out. 
This is especially a problem when numbers are part of the name. (I.e. 
Total.ACC-11-1 will become Total.ACC111)

3) I know a lot of researchers who are using the analysis option to create a 
'case' based table for SPSS. That is: all relevant data of an experiment is 
collected on one line. The problem is that the content of the table can only 
be based on numerical values, not string values. I know that this is 
required for performing statistical calculations, but this limitation also 
holds if the table is designed in such a way that each cell is derived from 
a single value. Our current workaround is to encode each relevant string 
value with a corresponding numerical value and store this along with all 
other attributes. For example: a typed response character can be translated 
at runtime to a unique numerical value by using inline script. Offline 
translation of such data is also possible by using excel and copying and 
pasting columns.

Let me finish with a compliment: E-Merge and E-DataAid are straightforward 
but powerful utilities, which do exactly what they claim: provide an easy 
mechanism to prepare datasets for further analysis. I also like the 
annotation option, which allows me to track any manual changes in the data.

kind regards,
Paul Groot
Vrije Universiteit Amsterdam



>From: Brandon Cernicky <brandon_cernicky at yahoo.com>
>To: eprime at mail.talkbank.org
>Subject: E-Prime 2.0 Data Features
>Date: Tue, 28 Jun 2005 19:06:05 -0700 (PDT)
>
>The last couple weeks had a number of threads in
>relation to data handling in E-Prime.  The recent
>E-Prime 2.0 Feature Survey placed its emphasis on many
>of the design and implementation features and did not
>focus on any data handling improvements for E-Merge
>and E-DataAid.
>
>Please feel free to discuss any items you would like
>to see for E-Prime 2.0 in regards to E-Merge and
>E-DataAid.
>
>Improvements to export in general?
>Improvements to export to Excel?
>Improvements to export to SPSS?
>Improvements to export to XML?
>
>PST is certainly not going to try to reinvent the
>stats wheel and understand there are many stats
>packages available and exporting to them is intended.
>Are there any areas where E-Merge or E-DataAid could
>be improved to make it so that bouncing in-between
>packages would be reduced?
>
>
>-Brandon
>
>~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>Brandon S. Cernicky
>Senior Software Engineer
>Psychology Software Tools
>
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam?  Yahoo! Mail has the best spam protection around
>http://mail.yahoo.com
>