Millisecond Forums

Data inconsistencies

https://forums.millisecond.com/Topic30934.aspx

By amanuel2 - 2/25/2021

Hello!
I am working with the preprogrammed face-name associative memory exam (https://www.millisecond.com/download/library/facenameassociativememoryexam/) but I seem to be having some issues with the data it gives me.
I have made some edits to the code and split it into two parts, before the delay, and after the delay.
I have manually recorded my results and compared them to the data from the test and there are some inconsistencies.
Before the delay: The CRN (correctly recalled names) value it is giving me is larger than the number of names that I actually correctly recalled, and the CRO (correctly recalled occupations) continually gives me "nan". I think the number I am getting for CRN is a combination of correct names and occupations (e.g., if I got 5 names correct and 4 occupations the CRN would be 9).
After the delay: The CRN30 (correctly recalled names after a 30-minute delay), I am getting the correct number, however, for CRO30 (correctly recalled occupations after a 30-minute delay) I am once again getting "nan".

Perhaps it is something with my edits, so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Does anyone know why this might be happening or what I can do to fix it?
By Dave - 2/25/2021

amanuel2 - 2/25/2021
Hello!
I am working with the preprogrammed face-name associative memory exam (https://www.millisecond.com/download/library/facenameassociativememoryexam/) but I seem to be having some issues with the data it gives me.
I have made some edits to the code and split it into two parts, before the delay, and after the delay.
I have manually recorded my results and compared them to the data from the test and there are some inconsistencies.
Before the delay: The CRN (correctly recalled names) value it is giving me is larger than the number of names that I actually correctly recalled, and the CRO (correctly recalled occupations) continually gives me "nan". I think the number I am getting for CRN is a combination of correct names and occupations (e.g., if I got 5 names correct and 4 occupations the CRN would be 9).
After the delay: The CRN30 (correctly recalled names after a 30-minute delay), I am getting the correct number, however, for CRO30 (correctly recalled occupations after a 30-minute delay) I am once again getting "nan".

Perhaps it is something with my edits, so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Does anyone know why this might be happening or what I can do to fix it?

> so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Could you please provide the data files -- raw and summary -- for that.
By Dave - 2/25/2021

Dave - 2/25/2021
amanuel2 - 2/25/2021
Hello!
I am working with the preprogrammed face-name associative memory exam (https://www.millisecond.com/download/library/facenameassociativememoryexam/) but I seem to be having some issues with the data it gives me.
I have made some edits to the code and split it into two parts, before the delay, and after the delay.
I have manually recorded my results and compared them to the data from the test and there are some inconsistencies.
Before the delay: The CRN (correctly recalled names) value it is giving me is larger than the number of names that I actually correctly recalled, and the CRO (correctly recalled occupations) continually gives me "nan". I think the number I am getting for CRN is a combination of correct names and occupations (e.g., if I got 5 names correct and 4 occupations the CRN would be 9).
After the delay: The CRN30 (correctly recalled names after a 30-minute delay), I am getting the correct number, however, for CRO30 (correctly recalled occupations after a 30-minute delay) I am once again getting "nan".

Perhaps it is something with my edits, so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Does anyone know why this might be happening or what I can do to fix it?

> so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Could you please provide the data files -- raw and summary -- for that.

Nevermind, I believe I've spotted the bug. There's a typo in <surveypage CR> :

Note: surveypage.CR (cued recall)
* presents a randomly selected target face
* presents 2 textboxes: a) one to enter the corresponding name b) one to enter the corresponding occupation
<surveypage CR>
/ ontrialbegin = [
    values.index = list.itemnumbers.nextvalue;
    values.targetFace = list.faces_itemnumbers.nextvalue;
    values.targetName = list.names_itemnumbers.nextvalue;    
    values.targetOccupation = list.occupations_itemnumbers.nextvalue;            
]
/ stimulusframes = [1 = faceB]
/ questions = [1=CRN,CRO]
/ nextbuttonposition = (50%, 95%)
/ finishlabel = "submit"
/ showbackbutton = false
/ showpagenumbers = false
/ ontrialend = [
    values.correctResponse_CRN = tolower(item.names.item(values.targetName));
    values.response_CRN = tolower(textbox.CRN.response);
    if (values.response_CRN == values.correctResponse_CRN){
        values.correct_CRN = 1;
    } else {
        values.correct_CRN = 0;
    };
    list.CRN_accuracy.insertitem(values.correct_CRN, 1);
    
    values.correctResponse_CRO = tolower(item.occupations.item(values.targetOccupation));
    values.response_CRO = tolower(textbox.CRO.response);
    if (values.response_CRO == values.correctResponse_CRO){
        values.correct_CRO = 1;
    } else {
        values.correct_CRO = 0;
    };
    list.CRN_accuracy.insertitem(values.correct_CRO, 1);    
]
/ fontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ itemfontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ responsefontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
</surveypage>

needs to be

Note: surveypage.CR (cued recall)
* presents a randomly selected target face
* presents 2 textboxes: a) one to enter the corresponding name b) one to enter the corresponding occupation
<surveypage CR>
/ ontrialbegin = [
    values.index = list.itemnumbers.nextvalue;
    values.targetFace = list.faces_itemnumbers.nextvalue;
    values.targetName = list.names_itemnumbers.nextvalue;    
    values.targetOccupation = list.occupations_itemnumbers.nextvalue;            
]
/ stimulusframes = [1 = faceB]
/ questions = [1=CRN,CRO]
/ nextbuttonposition = (50%, 95%)
/ finishlabel = "submit"
/ showbackbutton = false
/ showpagenumbers = false
/ ontrialend = [
    values.correctResponse_CRN = tolower(item.names.item(values.targetName));
    values.response_CRN = tolower(textbox.CRN.response);
    if (values.response_CRN == values.correctResponse_CRN){
        values.correct_CRN = 1;
    } else {
        values.correct_CRN = 0;
    };
    list.CRN_accuracy.insertitem(values.correct_CRN, 1);
    
    values.correctResponse_CRO = tolower(item.occupations.item(values.targetOccupation));
    values.response_CRO = tolower(textbox.CRO.response);
    if (values.response_CRO == values.correctResponse_CRO){
        values.correct_CRO = 1;
    } else {
        values.correct_CRO = 0;
    };
    list.CRO_accuracy.insertitem(values.correct_CRO, 1);    
]
/ fontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ itemfontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ responsefontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
</surveypage>

And, similarly, in <surveypage CR30>

Note: surveypage.CR (cued recall - 30minutes delay)
* presents a randomly selected target face
* presents 2 textboxes: a) one to enter the corresponding name b) one to enter the corresponding occupation
<surveypage CR30>
/ ontrialbegin = [
    values.index = list.itemnumbers.nextvalue;
    values.targetFace = list.faces_itemnumbers.nextvalue;
    values.targetName = list.names_itemnumbers.nextvalue;    
    values.targetOccupation = list.occupations_itemnumbers.nextvalue;            
]
/ stimulusframes = [1 = faceB]
/ questions = [1=CRN,CRO]
/ nextbuttonposition = (50%, 95%)
/ finishlabel = "submit"
/ showbackbutton = false
/ showpagenumbers = false
/ ontrialend = [
    values.correctResponse_CRN30 = tolower(item.names.item(values.targetName));
    values.response_CRN30 = tolower(textbox.CRN.response);
    if (values.response_CRN30 == values.correctResponse_CRN30){
        values.correct_CRN30 = 1;
    } else {
        values.correct_CRN30 = 0;
    };
    list.CRN30_accuracy.insertitem(values.correct_CRN30, 1);
    
    values.correctResponse_CRO30 = tolower(item.occupations.item(values.targetOccupation));
    values.response_CRO30 = tolower(textbox.CRO.response);
    if (values.response_CRO30 == values.correctResponse_CRO30){
        values.correct_CRO30 = 1;
    } else {
        values.correct_CRO30 = 0;
    };
    list.CRN_accuracy.insertitem(values.correct_CRO30, 1);    
]
/ fontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ itemfontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ responsefontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
</surveypage>

needs to be

Note: surveypage.CR (cued recall - 30minutes delay)
* presents a randomly selected target face
* presents 2 textboxes: a) one to enter the corresponding name b) one to enter the corresponding occupation
<surveypage CR30>
/ ontrialbegin = [
    values.index = list.itemnumbers.nextvalue;
    values.targetFace = list.faces_itemnumbers.nextvalue;
    values.targetName = list.names_itemnumbers.nextvalue;    
    values.targetOccupation = list.occupations_itemnumbers.nextvalue;            
]
/ stimulusframes = [1 = faceB]
/ questions = [1=CRN,CRO]
/ nextbuttonposition = (50%, 95%)
/ finishlabel = "submit"
/ showbackbutton = false
/ showpagenumbers = false
/ ontrialend = [
    values.correctResponse_CRN30 = tolower(item.names.item(values.targetName));
    values.response_CRN30 = tolower(textbox.CRN.response);
    if (values.response_CRN30 == values.correctResponse_CRN30){
        values.correct_CRN30 = 1;
    } else {
        values.correct_CRN30 = 0;
    };
    list.CRN30_accuracy.insertitem(values.correct_CRN30, 1);
    
    values.correctResponse_CRO30 = tolower(item.occupations.item(values.targetOccupation));
    values.response_CRO30 = tolower(textbox.CRO.response);
    if (values.response_CRO30 == values.correctResponse_CRO30){
        values.correct_CRO30 = 1;
    } else {
        values.correct_CRO30 = 0;
    };
    list.CRO30_accuracy.insertitem(values.correct_CRO30, 1);    
]
/ fontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ itemfontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
/ responsefontstyle = ("Arial", 3%, false, false, false, false, 5, 1)
</surveypage>
By amanuel2 - 2/25/2021

Dave - 2/25/2021
amanuel2 - 2/25/2021
Hello!
I am working with the preprogrammed face-name associative memory exam (https://www.millisecond.com/download/library/facenameassociativememoryexam/) but I seem to be having some issues with the data it gives me.
I have made some edits to the code and split it into two parts, before the delay, and after the delay.
I have manually recorded my results and compared them to the data from the test and there are some inconsistencies.
Before the delay: The CRN (correctly recalled names) value it is giving me is larger than the number of names that I actually correctly recalled, and the CRO (correctly recalled occupations) continually gives me "nan". I think the number I am getting for CRN is a combination of correct names and occupations (e.g., if I got 5 names correct and 4 occupations the CRN would be 9).
After the delay: The CRN30 (correctly recalled names after a 30-minute delay), I am getting the correct number, however, for CRO30 (correctly recalled occupations after a 30-minute delay) I am once again getting "nan".

Perhaps it is something with my edits, so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Does anyone know why this might be happening or what I can do to fix it?

> so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Could you please provide the data files -- raw and summary -- for that.

Here are the data files for part 1 (before delay) and part 2 (after delay). Just to note, I also edited to code to have 12 people instead of 16 and they are different photos with different names and occupations. Other than that, the study runs the same as the original.
By Dave - 2/25/2021

amanuel2 - 2/25/2021
Dave - 2/25/2021
amanuel2 - 2/25/2021
Hello!
I am working with the preprogrammed face-name associative memory exam (https://www.millisecond.com/download/library/facenameassociativememoryexam/) but I seem to be having some issues with the data it gives me.
I have made some edits to the code and split it into two parts, before the delay, and after the delay.
I have manually recorded my results and compared them to the data from the test and there are some inconsistencies.
Before the delay: The CRN (correctly recalled names) value it is giving me is larger than the number of names that I actually correctly recalled, and the CRO (correctly recalled occupations) continually gives me "nan". I think the number I am getting for CRN is a combination of correct names and occupations (e.g., if I got 5 names correct and 4 occupations the CRN would be 9).
After the delay: The CRN30 (correctly recalled names after a 30-minute delay), I am getting the correct number, however, for CRO30 (correctly recalled occupations after a 30-minute delay) I am once again getting "nan".

Perhaps it is something with my edits, so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Does anyone know why this might be happening or what I can do to fix it?

> so I manually recorded my data for the unedited version and still got inconsistent results ("nan" for both CRO and CRO30).

Could you please provide the data files -- raw and summary -- for that.

Here are the data files for part 1 (before delay) and part 2 (after delay). Just to note, I also edited to code to have 12 people instead of 16 and they are different photos with different names and occupations. Other than that, the study runs the same as the original.

Please fix the two bugs I've listed here https://www.millisecond.com/forums/FindPost30936.aspx in your modified script and try again.