Thanks Jacqueline for your explanation
In this semester I have increased the options from 4 to 5 (a to e) because I have the problem of cheating and multiple streams but I only have 10 questions
In the TRAT the scoring I use is 4 first correct, 2 second try correct, 1 third try correct, 0 final try. I don't take account of the probability of getting correct because I have supposed that random sorting of the key makes each question independent
Please can you satisfy my curiosity why you use 55% as your non-guessing threshold as I haven't see that discussed. Always happy to learn more about the mechanics!
Warm regards/gary
________________________________________
From: Team-Based Learning [[log in to unmask]] on behalf of J.A. Vos [[log in to unmask]]
Sent: Monday, 18 July 2016 22:55
To: [log in to unmask]
Subject: Re: iRAT and tRAT scores correcting for guessing?
Hi,
I’m sorry, my question is not clear enough . I’ll clarify it.
When a student answers a one-best-answer item with options and he/she knows nothing about the subject, he/she can always guess what the right answer is by putting one of the options right. When an exam consists of one-best-answer items we are used to correct the exam score for guessing. A student normally pass the exam when he answers 55% of the items correct. Each right answer is one point, a wrong answer is no point. Because of the probality of guessing for exams with one-best-answer items a student should answer more items right than just the 55% of the items to pass the exam. Exactly, the score should be the guessing score plus 55% of the max score minus the guessing score.
An example:
A exam consists of 40 one-best-answer items, all with 4 options. Each right answer gives 1 point. The max score is 40 points.
When a student answers the questions only by guessing he can get an exam score of 10 points (= ¼ * 40).
To pass the exam the student should answer 26,5 items right (= 10 + 55% * (40-10)).
In our TBL-module all the RAT questions are one-best-answer items with 4 options.
In the iRAT a student can guess an answer. There is a chance of 25% that the answer is right.
Also in the tRAT the team can guess the answer, for the first answer there is a chance of 25% that the answer is right. When the first answer is wrong, a team can try again to give the right answer. For the second effort there are still 3 options left, so there is a chance of 33% that the second answer is right. And so on for the third effort.
When the iRAT and the tRAT were exams, we should set a pass score taking into account the guessing score.
And now my question:
What’s your way of calculating the score to pass (or the grades of) the iRAT/tRAT and what’s your opinion in taking into account the score for guessing? Is there any difference between the iRAT and the tRAT?
It would be very helpful when I can gain insight in your practice.
I hope this explanation will clarify my question.
Kind regards, Jacqueline
Van: Team-Based Learning [mailto:[log in to unmask]] Namens Sibley, James Edward
Verzonden: vrijdag 15 juli 2016 15:36
Aan: [log in to unmask]
Onderwerp: Re: iRAT end tRAT correcting for guessing?
Hi
Can you explain a bit more about "correct scores for guessing" I am not not quite sure what that is.....are you talking about confidence testing where student spread their points across all dis tractors ...or something else
Take care
Jim
Jim Sibley
Author of Getting Started with Team-Based Learning
Visit my website at learntbl.ca<http://learntbl.ca>
Sorry for brief message -sent from my iPad
On Jul 15, 2016, at 6:25 AM, J.A. Vos <[log in to unmask]<mailto:[log in to unmask]>> wrote:
Dear collegues,
I wonder what your practice is in calculating the iRAT and the tRAT grade of a TBL-module. Are you used to correct the scores for guessing?
Kind regards, Jacqueline Vos
the Netherlands
________________________________
AMC Disclaimer : https://www.amc.nl/disclaimer
________________________________
________________________________
To unsubscribe from the TEAMLEARNING-L list, please click here.<https://lists.ubc.ca/scripts/wa.exe?SUBED1=TEAMLEARNING-L&A=1>
Further information about the UBC Mailing Lists service can be found on the UBC IT website.
________________________________
To unsubscribe from the TEAMLEARNING-L list, please click here.<https://lists.ubc.ca/scripts/wa.exe?SUBED1=TEAMLEARNING-L&A=1>
Further information about the UBC Mailing Lists service can be found on the UBC IT website.
________________________________
To unsubscribe from the TEAMLEARNING-L list, please click here.<https://lists.ubc.ca/scripts/wa.exe?SUBED1=TEAMLEARNING-L&A=1>
Further information about the UBC Mailing Lists service can be found on the UBC IT website.
########################################################################
To unsubscribe from the TEAMLEARNING-L list, please click the following link:
https://lists.ubc.ca/scripts/wa.exe?SUBED1=TEAMLEARNING-L&A=1
Further information about the UBC Mailing Lists service can be found on the UBC IT website.
|
