Great video and thank for firstly speaking to Type 1 error & Type 2 error and why we don’t need to over complicate things. Going by a more intuitive approach helps.
I really enjoyed the lesson. Thanks for enlightening me on the context because the type one error is known to be worse, but after watching your video, I have to understand the context and be wiser regarding the conclusion and required action.
Hey doctor Jeff, having a strong feeling that you will rise fast and become one of the youtibe heriage of education. I believe that writing about a range of topics is a good way to improve one's understanding. I find your closing question insightful. Thank you for your lecture.
Researchers or statisticians often use Type I/Type II errors in technical papers and discussions. Practitioners in fields like medicine, machine learning, or quality control might use false positives and false negatives because they are more intuitive and relatable.
Good question, with no easy answer. The larger your sample, the less likely you are to experience both types of error, but that also largely depends on the variability of the data itself.
let us suupose there are 100 students in each school and we measure the height of all students and compare between two school, can we say we can get 100% errorless result?
It's a little difficult to make a claim about what is worst in science. False positives a re far more commons than negative positives, given those conditions, it's easier to run into false positive than with negative positives, because we do care about the sample size and the spread or variability of the data instead of caring about the randomness and uncertainty around false positives more frequently. It's kind of bias or something like that with the uncertainty.
The first example is bad. It helps with nothing. I've given it an hour with no clear understanding between them, nor perspective of what is true. You cannot ask this question and get a right answer 100% of the time.
Great video and thank for firstly speaking to Type 1 error & Type 2 error and why we don’t need to over complicate things. Going by a more intuitive approach helps.
I really enjoyed the lesson. Thanks for enlightening me on the context because the type one error is known to be worse, but after watching your video, I have to understand the context and be wiser regarding the conclusion and required action.
Hey doctor Jeff, having a strong feeling that you will rise fast and become one of the youtibe heriage of education.
I believe that writing about a range of topics is a good way to improve one's understanding. I find your closing question insightful. Thank you for your lecture.
Thank you for your kind words!
Researchers or statisticians often use Type I/Type II errors in technical papers and discussions.
Practitioners in fields like medicine, machine learning, or quality control might use false positives and false negatives because they are more intuitive and relatable.
you are by far the best at explaining this stuff
Finally this topic makes sense. Thank you!
Excellent explanation on these concepts especially to laymen. thank you so much
Thank you! So glad you found the explanation clear and useful!
Type I error (falsely rejecting a null hypothesis) and type II error (falsely accepting a null hypothesis).
I take it that the null hypothesis is considered a negative and the alternate hypothesis is considered a positive.
So out of 1000 student schools, how large would the sample sizes of each school have to be to minimize the possibility of a false positive/negative?
Good question, with no easy answer. The larger your sample, the less likely you are to experience both types of error, but that also largely depends on the variability of the data itself.
In information retrieval these are called recall and precision.
100% right. Thank you pointing that out!
Really good video, thank you
Great video. Many thanks.
Thank you!
Thank you
let us suupose there are 100 students in each school and we measure the height of all students and compare between two school, can we say we can get 100% errorless result?
Yes
And if you say no you have to prove it
thank you sir
you are the best
Well it depends on the benefits that come from the results , i mean newton theory had a beneficial outcome although it is not totally true
It's a little difficult to make a claim about what is worst in science. False positives
a re far more commons than negative positives, given those conditions, it's easier to run into false positive than with negative positives, because we do care about the sample size and the spread or variability of the data instead of caring about the randomness and uncertainty around false positives more frequently. It's kind of bias or something like that with the uncertainty.
The first example is bad. It helps with nothing. I've given it an hour with no clear understanding between them, nor perspective of what is true. You cannot ask this question and get a right answer 100% of the time.