|
|
|
Forum Newbie
Group: Forum Members
Last Login: 10/26/2009 4:59:55 PM
Posts: 1,
Visits: 3
|
|
Hello all,
I'm hoping to get some insight/recommendations regarding complex filtering of trials during an Emotional Stroop paradigm.
I'm using the E-Prime v1.1 software package. I've used E-Run v1.1.4.6 to present single words in a color response task. There 10 words from five different semantic categories(a,b,c,d,e), yielding 50 words total. Each word is presented once in each of 3 colors, yielding 150 trials. Each of these word-color stimuli are presented randomly, without replacement. You may recognize this as a "mixed presentation" emotional Stroop task.
For filtering and analysis, I'm using E-DataAid v1.1.4.1. (Outdated, yes.)
---------------------------------
Problem 1:
I would like to filter out all trials except those occurring directly after the presentation of "catagory a" words.
Problem 2:
I would like to filter out all trials except those occurring directly after presentation of "category a" or "category b."
--------------------------------
I have not had any luck achieving these goals myself (using DataAid v1.1.4.1). I understand this problem could have been avoided by presenting the words in a fixed, pseuedo-random sequence. Since, I've already collected 300+ data files (using randomization w/o replacement), I'm hoping someone may know of a solution. In order to better direct your feedback, I will propose some possible solutions. I need you to tell me if such a solution exists!
--------------------------------
Potential Solution 1:
Using the current software (DataAid v1.1.4.1), there is some way to filter out trials, such that only trials occurring directly after "catagory a" words remain. If you're familiar with this version and know of a solution, please respond with "Actual Solution 1."
Potential Solution 2:
Using a newer version of DataAid software, there is some way to filter out trials, such that only trials occurring directly after "catagory a" words remain. Since I have no experience with newer versions of DataAid, I can only hope that there are new filtering options. If you're aware of a solution using a different version of DataAid, please respond with "Actual Solution 2." I someone could confirm this, I would be interested in purchasing an updated version of DataAid.
Potential Solution 3:
There is some 3rd party software that can analyze files created by E-Prime. (Tsk tsk.) If you're aware of some other software that may be able to perform this complex filtering, please respond with "Actual Solution 3."
Of course, I welcome any other ideas you may have!
Thanks,
-Dan
|
|
|
|
|
Forum Member
Group: Forum Members
Last Login: 6/23/2010 7:08:51 PM
Posts: 25,
Visits: 63
|
|
| Export data to Microsoft Excel. Use Excel functions to define new variables based on whatever combinations and conditionals you want. Analyze data using these newly created variables. I always use Excel as an intermediate step to prepare data for statistical analysis.
|
|
|
|
|
Forum Member
Group: Forum Members
Last Login: 6/23/2010 7:08:51 PM
Posts: 25,
Visits: 63
|
|
| I have to admit that I might underestimate DataAid; I don't do much with it. And maybe I misunderstand what you're after. As I understand it, you simply want each trial identified by the category of the previous trial. I see that there is a way to create a new data column in DataAid. I do not see, however, how one can define a cell in this new column according to values in other columns and rows. I think you just want a new "PrevTrialCat" column to show the category of the previous trial, if that previous trial also is in the same block. Perhaps someone more familiar with DataAid can explain how make it do this; it's a simple spreadsheet function in Excel. Do you need something more complicated?
|
|
|
|
|
Forum Member
Group: Forum Members
Last Login: 6/23/2010 7:08:51 PM
Posts: 25,
Visits: 63
|
|
| Maybe I should wait for you to explain more, but I just thought of a clumsy way you might be able to accomplish this with dataAid (there's surely a better way). Create a new data column; call it "prevTrialCat". Copy the entire column of the merged data with the Category identifier. Paste in the new column but displaced one row. Then filter the first trial of each block or subject (since the previous trial is irrelevant). Of course the data must be properly sorted by subject, block, and trial. I haven't tried the copy and paste in DataAid, but surely you can do that. Would this work?
|
|
|
|
