My point was that after autorecoding, you could use VARSTOCASES combine as you originally wanted to do, as long as the autorecode definitionis the same for all the variables.
If you wanted to combine these variables some other way, you should look at the LOOP, DO REPEAT and VECTOR commands in the Command Syntax Reference. The Programming and Data Management pdf has a chapter on data management that might also be helpful. You can download that book via Help > Documentation in PDF Format
--
Original Message:
Sent: 9/13/2023 8:04:00 AM
From: Ash
Subject: RE: Combining Several Performance Task Trials into Designated Categories.
Hi Jon,
Thank you for the prompt suggestion. I have had a look at autorecoding and run some tests, but it seems that I cannot state that variables 1-12 fall under one group and they just get treated as 12 new variables (which I think is almost like the standard recoding process). Perhaps I am going about this wrong. Do you happen to know of step-by-step tutorials you would recommend for marking several variables as one category for analysis?
Just to add in an edit as I think this was not clear, sorry - I am trying to retain the reaction time data within each cell. So participant 1 had a reaction time across all trials, but only parts of those trials were in response to certain images. This is where the restructuring fails, because I cannot retain the psychometric data for each participant as this will only have 12 cells of data compared to the 48 cells of data for the reaction time task.
Thank you again for the help so far.
------------------------------
Ash
------------------------------
Original Message:
Sent: Mon September 11, 2023 10:04 AM
From: Jon Peck
Subject: Combining Several Performance Task Trials into Designated Categories.
You could write code to collect values from different variables, but you can restructure your strings, which might be easier. Use autorecode to make them numeric and go from there. If there is a pattern in the image variables that you need to preserve, you might try using a template with autorecode.
--
Original Message:
Sent: 9/10/2023 5:53:00 PM
From: Ash
Subject: Combining Several Performance Task Trials into Designated Categories.
Hi there,
I am hoping that someone might be able to help me with data analysis/structuring.
My data is exploring whether two psychometrics (treated as separate variables, not combined) can effect task performance (reaction time and error rate) when exposed to certain stimuli in a given task. In this task, I have 48 trials where trials 1-12 show one image, trials 13-25 show a second image, and trials 26-48 show a third image. As a brief example, my data looks something like this:
| participantID |
Age |
Trial1 |
Trial2 |
Trial3 |
Trial4 |
Trial5 |
Trial6 |
Trial1 RT |
Trial2 RT |
Trial 3 RT |
Trial 4 RT |
Trial 1 Error |
Trial 2 Error |
Trial 3 Error |
Psychometric1 Mean Score |
Psychometric2 Mean Score |
| 1 |
55 |
Image A |
Image A |
Image A |
Image B |
Image B |
Image B |
.45 |
.52 |
.28 |
.65 |
Y |
N |
N |
5 |
3 |
| 2 |
26 |
Image A |
Image A |
Image A |
Image B |
Image B |
Image B |
.38 |
.48 |
.61 |
.88 |
N |
Y |
N |
2 |
1 |
Unfortunately, I cannot restructure the data from wide to long because the lengths of the variables I want to restructure are not the same (I have around 8 other variables I would ideally like to explore on top of this). This would also duplicate the psychometric scores (each participant would have 12 inputs instead of just 1 for the mean score) which I think would effect any analysis. I tried manual entry as far as I could but, with over 150 participants, this would take several hours of entry that is open to human error.
Is there any way that I can tell SPSS to categorise specific trials that are spread across multiple variables? I have tried re-coding, transposing, manual entry and aggregating but I think I am just not able to figure this one out myself. I am not used to running Syntax, but am open to learning.
If this is not possible, then I would be extremely grateful to hear what analysis ideas anyone might have as an alternative approach. Any help that anyone could provide would be greatly appreciated. Thank you.
------------------------------
Ashleigh H-S
------------------------------