Thread Tools  
bradw
Junior Member
 
Join Date: Dec 2008
bradw is on a distinguished road Posts: 9
need help with data calcs & export, Posted January 5th, 2009 #1
Hi folks,

I'm new to the forum and Superlab and need some help. I have 'inherited' data from an Emotional Stroop task that I wish to clean, transform and then merge in SPSS.

For each participant, an introduction screen is first presented, followed by a practice/instruction screen, then 10 practice trials (using a subset of stimuli presented in the actual trials), another instruction 9'are you ready?') screen, then the 48 actual trials (4 stimulus word types, each with 6 words, and 4 corresponding control-neutral word sets of 6 words each, presented in a randomized order for each participant, but with each word assigned and recorded as a unique Trial Number across all participants... e.g., the word "worried" = trial no. 14, "tense"= trial 32, etc).

The data files for each participant display as follows when I open them in Excel. (I've deleted several irrelevant columns and rows for sake of brevity here.) In the data output, there are several different variables that are recorded in the first column (participant ID -- in this case "101W", followed by the task name, date, and trial name), only one of which I care about (ID), while the other variables of interest each have their own column:

------------
101W
stroop.xpt
Wed Jan 1, 12:33:00 2008


Trial Name Trial No. Error Code ReactionTime
intro 1 C 561243
practice 2 C 2062
red-worried 14 C 1001 (*this is the first practice trial)
* (9 more practice trial rows -- deleted here)
instructions 3 C 5000
blue-tense 32 E 1205
red-cat 20 C 782
* (46 more actual trials -- deleted here)
-----------

I need help coming up with a way (maybe a macros program or two? never used macros before) to do the following:

1) Screen/clean data for each participant:

a) Discard Instructions and Practice Trials (i.e. all Trial rows between the Trial Name “Introduction” row and “Instructions” row).

b) Discard overly fast/slow Trials: (i.e. Trial rows with Reaction Time <100 msec or >3000 msec)

c) Discard Trials with incorrect responses (i.e. Trial rows with Error Code = E)

d) Code as invalid any participant's data containing too many errors (i.e. less than 39 valid actual trials remaining)

2) Compute means to summarize each participant's data & merge into SPSS:

Desired SPSS format:
ID No. MeanRTW1 MeanRTW1cn MeanRTW2 MeanRTW2cn... (8 total summary variables)

Where Mean RTW1=Mean Reaction Time for Type 1 Words: i.e. sum of Reaction Time values for rows with corresponding Trial No. values of 12, 13, 74, 53, 54, and 31, divided by the total number of RT values (6, or less if trials were discarded)

Mean RTW1cn= Mean Reaction Time for Type 1 Control Words (i.e., corresponding to trials 58, 38, 39, 40, 80, 81).

Etc. for Word types 3 & 4 and their control word sets (total of 8 variables).


I tried the Superlab Data Viewer but didn't know what I was doing and couldn't find an instruction manual. I'm not sure what version of Superlab this program was written in but I presume it was an earlier one since data collection began in 2005. I would greatly appreciate any assistance you can provide.

Thanks!
Brad W
Reply With Quote
bradw
Junior Member
 
Join Date: Dec 2008
bradw is on a distinguished road Posts: 9
need help with data -- p.s., Posted January 5th, 2009 #2
In my post earlier, the columns spacing didn't display correctly for my sample excel file. Trying again:

---------------
101W
stroop.xpt
Wed Jan 1, 12:33:00 2008


Trial Name.....Trial No...... Error Code.......ReactionTime
introduction..... 1................C..................561243
practice.......... 2................C.................. 2062
red-worried.....14...............C...................1 001 (*this is the first practice trial)
* (9 more practice trial rows -- deleted here)
instructions......3................C.............. ......5000
blue-tense.......32...............E.................... 1205
red-cat............20..............C.................. ...782
* (46 more actual trials -- deleted here)
---------------

Thanks! BW
Reply With Quote
Hisham
Founder
 
Join Date: Aug 2006
Location: San Pedro, CA
Hisham is on a distinguished road Posts: 1,124
Posted January 5th, 2009 #3
Hi Brad,

It sounds like you may need Excel macros to clean up the data -- not something that we have the expertise to help with. I recommend that you give the Cedrus Data Viewer another try. It has the ability to flag records automatically, e.g. when RT > 3000 msec. Data files can then be exported or merged while omitting the flagged records.

Data Viewer was tested with SuperLab 4.0 data files but the 2.0 data files will probably work with it.

Hisham.
Reply With Quote
bradw
Junior Member
 
Join Date: Dec 2008
bradw is on a distinguished road Posts: 9
macros & data viewer, Posted January 6th, 2009 #4
Thanks, Hisham. Sounds like I'll need to find some help with macros. Any suggestions on where to look/inquire?

Re. flagging, is there a manual or instructions on how to use the Cedrus Data Viewer?

Regards,
BradW
Reply With Quote
Hisham
Founder
 
Join Date: Aug 2006
Location: San Pedro, CA
Hisham is on a distinguished road Posts: 1,124
Posted January 6th, 2009 #5
I can't help with Excel macros -- sorry.

There is no manual for Cedrus Data Viewer other than what's on the web. Key points:
  • Click on the [+] icon in the bottom left of the main window to add data files to the list.

  • Once added, you can select one, two, or any number of files at the same time. Data Viewer will display the records in the selected files only. The summary stats will also apply to the selected files only.

  • From the Rows menu, select Flag Special. You will see a variety of options for having Data Viewer automatically flag records for you.

  • Later, when you click on the Merge Selected Files button on the lower right, a small dialog (window) will popup. It offers the option to include or exclude the flagged rows.
I hope this helps.
Reply With Quote
bradw
Junior Member
 
Join Date: Dec 2008
bradw is on a distinguished road Posts: 9
help with data calcs, Posted January 6th, 2009 #6
Thanks, Hisham. I'll keep searching for help with Excel macros too.
Reply With Quote
bradw
Junior Member
 
Join Date: Dec 2008
bradw is on a distinguished road Posts: 9
good help with macros for superlab data, Posted January 9th, 2009 #7
For anyone who may run into similar challenges, I received excellent assistance on www.excelforum.com with developing macros to simultaneously screen/clean, summarize, and merge my participants' Superlab data.

Regards,
BradW
Reply With Quote
Thread Tools
Posting Rules
Forum Jump:
© Copyright Cedrus Corporation, 1991-2014.
All times are GMT -7. The time now is 11:48 PM.