Dear Greg,

Thank you very much for your kind reply.


Yes, it indeed makes sense to account for "NLR", which we didn't considered 
before. I rechecked the raw data and then have another doubt.


Let's take participant No.100206 for example. In the integrated behavioral data 
table, the "WM_Task_2bk_Face_Acc" score is 81.25, which indicates that there 
are some NLR trials for him/her. However, in the raw trial-level data (20 
trials in total, attached), there is no "NonResp" trials, and the Stim.RT were 
all recorded within the limitation of 2.5s. The Stim.ACC has 3 errors, 
therefore the accuracy is 17/20=0.85. 


I've no idea where I made a mistake for this difference. Could you please help 
figure it out? Thanks a lot.


Best regards,
Xinyang





At 2019-03-13 01:39:29, "Burgess, Gregory" <gburg...@wustl.edu> wrote:
Hello Xinyang,


Most likely, you haven’t accounted for “NLR”, which stands for “no logged 
responses”. If the response period timed out without an overt response, we 
can’t be sure whether the participant would have made a correct response or an 
error response. Since NLRs are neither correct or error trials, they’re omitted 
from the accuracy calculations like missing data might be handled. 


For example, if a condition had 4 NLRs, 13 correct responses and 3 error 
responses, they could have an accuracy rate of 81.25% = [ 100 x 13 / (20 - 4) ].


Hope this helps!

--Greg

____________________________________________________________________
Greg Burgess, Ph.D.
Senior Scientist, Human Connectome Project
Washington University School of Medicine
Department of Psychiatry
Phone: 314-362-7864
Email: gburg...@wustl.edu


On Mar 12, 2019, at 7:44 AM, Xinyang Liu <xinyang_ie...@163.com> wrote:


Dear HCP experts,


Hi. We are currently analyzing the behavioral accuracy scores from the Working 
Memory(WM) tfMRI task. However, we found an inconsistency between computed 
accuracy scores from the WM trial-level raw data and the "WM_Task_2bk_*_Acc" 
analyzed data columns in the integrated behavioral data file downloaded from 
the HCP website (named "unrestricted"). 


We attached an example in this email, showing that the average values of "LR" 
and "RL" raw scores are not equal to the analyzed values, although they have a 
trend to match each other. There are 20 trials (LR an RL) together in each 
task. However, the analyzed scores like "68.75" or "81.25" does not seem to be 
computed based on 20 trials as a total number. Was there any deletion of trials 
during processing?  If so, what is the reason to do this?


Another thing we found is that, for some participants, like No.104012 (not 
shown in the example file), they have raw behavioral data in the data boxes, 
but the results were not provided in the analyzed behavioral data file. What is 
the reason for that?


We would be very appreciated to receive any feedbacks. Thank you very much!


Best regards,
Xinyang





 


_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

<Example of accuracy inconsistency.xlsx>



 

The materials in this message are private and may contain Protected Healthcare 
Information or other information of a sensitive nature. If you are not the 
intended recipient, be advised that any unauthorized use, disclosure, copying 
or the taking of any action in reliance on the contents of this information is 
strictly prohibited. If you have received this email in error, please 
immediately notify the sender via telephone or return mail.
_______________________________________________
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users
subID   task_name       trial_number    trial.RT        trial.ACC
100206  WM_2-Back_Face_RL       1       741     1
100206  WM_2-Back_Face_RL       2       1202    1
100206  WM_2-Back_Face_RL       3       772     1
100206  WM_2-Back_Face_RL       4       840     1
100206  WM_2-Back_Face_RL       5       881     0
100206  WM_2-Back_Face_RL       6       1019    0
100206  WM_2-Back_Face_RL       7       806     1
100206  WM_2-Back_Face_RL       8       736     1
100206  WM_2-Back_Face_RL       9       910     1
100206  WM_2-Back_Face_RL       10      1005    1
100206  WM_2-Back_Face_LR       1       535     1
100206  WM_2-Back_Face_LR       2       537     1
100206  WM_2-Back_Face_LR       3       759     1
100206  WM_2-Back_Face_LR       4       759     1
100206  WM_2-Back_Face_LR       5       693     1
100206  WM_2-Back_Face_LR       6       889     1
100206  WM_2-Back_Face_LR       7       687     1
100206  WM_2-Back_Face_LR       8       763     1
100206  WM_2-Back_Face_LR       9       743     1
100206  WM_2-Back_Face_LR       10      989     0

Reply via email to