Thank you so much! Answered everything that I didn’t know. Even answered when I couldn’t look for the data analysis tool and explained what the results mean. I wish you were my teacher 😭
You can generally use t-test independent assuming unequal variances But, in-case you want to be sure then; You will have to perform an F - Test to find out whether to choose equal or unequal variances. 1.)Calculate variances for both samples 2.) F = Larger Variance/Smaller Variance -----> THIS IS THE F-STAT or F-SCORE 3.)DoF = count(obs)-1 [for column which gave larger variance] and DoF = count(obs)-1 [for column which gave smaller variance] 4.) Check your alpha in the question, download F-Distribution table w.r.t. that alpha level 5.) Larger Variance column DoF should be the DoF1. Check that in column of F-table Smaller Variance column DoF should be DoF2. Check that in row of F-table 6.) You will get a F-Critical. If your F-Score is lesser than or equal to the F-Critical, then go for the Equal Variance t-test Note: If you are doing it by data analysis tool pack, make sure that in variable 1 you enter the samples column, which gave you a larger variance.
When nonnormality is a problem, you can use the nonparametric test described here: real-statistics.com/non-parametric-tests/wilcoxon-rank-sum-test/. I don’t have a video for this situation, I’m afraid.
What quickly comes to my attention is that you have a total of 30 observation, yet you are doing a Student's t-test. For observation of n >= 30, why aren't you using the Normal Distribution? You have instead chosen to do a Student's t which is for n being a s mall number.
Thank you Ken, your video was precise and helped me make my sample t test in 2024
Thank you so much! Answered everything that I didn’t know. Even answered when I couldn’t look for the data analysis tool and explained what the results mean. I wish you were my teacher 😭
Your last information of scientific notation really helped me i was stuck on the same thank you.
Authentic! Thank you very much, Ken.
YOU SAVED OUR STUDY
You really helped me solve my problem in excel, thank you so much
what if the sample sizes are unequal? How do you know when to assume unequal variances?
You can generally use t-test independent assuming unequal variances
But, in-case you want to be sure then;
You will have to perform an F - Test to find out whether to choose equal or unequal variances.
1.)Calculate variances for both samples
2.) F = Larger Variance/Smaller Variance -----> THIS IS THE F-STAT or F-SCORE
3.)DoF = count(obs)-1 [for column which gave larger variance] and DoF = count(obs)-1 [for column which gave smaller variance]
4.) Check your alpha in the question, download F-Distribution table w.r.t. that alpha level
5.) Larger Variance column DoF should be the DoF1. Check that in column of F-table
Smaller Variance column DoF should be DoF2. Check that in row of F-table
6.) You will get a F-Critical. If your F-Score is lesser than or equal to the F-Critical,
then go for the Equal Variance t-test
Note: If you are doing it by data analysis tool pack, make sure that in variable 1 you enter the samples column, which gave you a larger variance.
Thanks ❤
Sir plz tell me there are 60 total number of observations why df is 50 instead it should be 58.
Bloody helpful. Thanks!
This is very helpful. Thank you!
what is the significance of hypothesized mean difference box?
Thank you!! So helpful!
THAT'S AWESOME~~~!!!
Excellent presentation
amazing, thank you
Excellent, thank you.
thank you so much.. my question is.. if my data not normal distribution, what statistical test can I use
When nonnormality is a problem, you can use the nonparametric test described here: real-statistics.com/non-parametric-tests/wilcoxon-rank-sum-test/. I don’t have a video for this situation, I’m afraid.
from where do we find the numbers for a and b?
I get it now, thanks so much.
Thank you very much!! :D
thank you sooooooooooooooo much!!!
t stat is t obt, right?
Thanks!!!
Thank you sir
how to interpret?
thanks
What quickly comes to my attention is that you have a total of 30 observation, yet you are doing a Student's t-test. For observation of n >= 30, why aren't you using the Normal Distribution? You have instead chosen to do a Student's t which is for n being a s
mall number.