Quote:
Originally Posted by MeleaB
I think you owe it to everyone to apologise and just admit that you simply messed up the comparisons. It's fine to use an approximation but there's no excuse not to use the correct percentiles for that figure!!
How can you read everyone's responses (e.g. "I'm terrible", "Am I ******ed?", "Dang it, I thought I was better at probability than that!!!") and not feel the need to correct yourself. People have given up their time to help you and you've basically told them "Thanks, and you're really dumb."
And if you didn't realise the results were so far off, then how could you possibly not have!?
Hi MeleaB,
I am one of collaborators in this study responsible for programming the survey.
Thank you for your time taking the survey and kindly rectified the feedback information.
I am really sorry for the use of bad approximation in providing the feedback. The original idea was to have a guess of what general population would be like given that we only have a sample of 82 recreational poker players. The resultant approximation, however, was too much a boost in their performance scores (especially in the incoherence scores).
We didn't mean to provide misleading feedback and downplay anyone's real performances. However, the approximation doesn't serve our original purpose well and now it looks like a disastrous idea.
Adjusting the feedback scores halfway through the data collection process would create unnecessary confound in our study, which would basically create two subgroups of people whose feedback is based on different comparisons. To avoid this confound, we decided to let everyone have the same comparison, at least for the period of data collection.
In retrospect, we regret to use approximations and apology to people who have sacrificed times to take our survey. I hope the additional quantile information that Phil posted earlier can be somewhat helpful.
I also include a more detailed individual-level scatterplot of performance scores (each dot represents one participant):
Best regards,
Zhu
Last edited by jqz; 03-09-2020 at 06:03 PM.