Dear CONN users,
I am writing having some doubts about the correctness of Denoising and 1st level Analysis steps that I have conducted on my data.
My experiment consisted of 2 phases – encoding and retrieval (after almost 3 weeks), using event-related design.
During encoding session, my participants memorized stimuli related to 3 emotion conditions – DISGUST, FEAR, NEUTRAL. Here I wanted to analyze only stimuli that were later correctly recognized.
During retrieval session, they were asked to indicate if a presented stimulus is NEW or OLD, again related to 3 emotion conditions. Again I wanted to analyze only correct responses.
This way, I had 3 conditions in encoding (durations: 4s in sessions 1, 6s in session 2) and 6 conditions in recognition (durations: 7s in sessions 1 and session 2). TR was the same - 2.5s in both parts of the experiment. Also, all the preprocessing steps were exactly the same in both sessions.
However, I can see huge differences between sessions in Conn:
- DENOSING: Histogram of original data is skewed in encoding, but not really skewed in recognition; also the preview looks fragmented in recognition
- 1st level ANALYSIS: in recognition each fragment of brain surface is either blue or red showing some effects, it looks strange.
I was wondering if during recognition something went wrong and what would be the reason, either in the 1st level analysis or denoising step.
Please find attached the comparison of denoising and 1st level analysis for my encoding and recognition for two subjects.
Best wishes,
Monika
I am writing having some doubts about the correctness of Denoising and 1st level Analysis steps that I have conducted on my data.
My experiment consisted of 2 phases – encoding and retrieval (after almost 3 weeks), using event-related design.
During encoding session, my participants memorized stimuli related to 3 emotion conditions – DISGUST, FEAR, NEUTRAL. Here I wanted to analyze only stimuli that were later correctly recognized.
During retrieval session, they were asked to indicate if a presented stimulus is NEW or OLD, again related to 3 emotion conditions. Again I wanted to analyze only correct responses.
This way, I had 3 conditions in encoding (durations: 4s in sessions 1, 6s in session 2) and 6 conditions in recognition (durations: 7s in sessions 1 and session 2). TR was the same - 2.5s in both parts of the experiment. Also, all the preprocessing steps were exactly the same in both sessions.
However, I can see huge differences between sessions in Conn:
- DENOSING: Histogram of original data is skewed in encoding, but not really skewed in recognition; also the preview looks fragmented in recognition
- 1st level ANALYSIS: in recognition each fragment of brain surface is either blue or red showing some effects, it looks strange.
I was wondering if during recognition something went wrong and what would be the reason, either in the 1st level analysis or denoising step.
Please find attached the comparison of denoising and 1st level analysis for my encoding and recognition for two subjects.
Best wishes,
Monika