Performing an independent samples t test in Excel

[@aaronrichmond374]

Thank you Ken, your video was precise and helped me make my sample t test in 2024

[@slcolt45]

THANK YOU for explaining the scientific notation issue with the p-value. Greatly appreciated!

[@soapyroz4471]

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 😭

[@mukularora9128]

Your last information of scientific notation really helped me i was stuck on the same thank you.

[@connieekenwa3291]

Authentic! Thank you very much, Ken.

[@radheymohan6968]

Thanks ❤

[@hjdc7422]

YOU SAVED OUR STUDY

[@juanmartinsantos5382]

You really helped me solve my problem in excel, thank you so much

[@Lyn-eg3id]

what if the sample sizes are unequal? How do you know when to assume unequal variances?

[@tubeymoto]

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.

[@eesa6463]

Sir plz tell me there are 60 total number of observations why df is 50 instead it should be 58.

[@zaidoonibrahim8580]

thank you so much.. my question is.. if my data not normal distribution, what statistical test can I use

[@KenBlake]

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.

[@sardineRN]

Bloody helpful. Thanks!

[@ghayamohammed1674]

from where do we find the numbers for a and b?

[@laxmanbisht2638]

what is the significance of hypothesized mean difference box?

[@sandryl0109]

This is very helpful. Thank you!

[@rebeccadegracia8778]

Thank you!! So helpful!

[@tidalaustralia]

Excellent presentation

[@AnemicChia]

THAT'S AWESOME~~~!!!

[@baxterdifabrizio6418]

Excellent, thank you.

[@frankieroandMCR]

amazing, thank you

[@michelleluu13]

Thank you very much!! :D

[@gyogalakshmi1704]

Thank you sir

[@RimmyTim]

I get it now, thanks so much.

[@lidiyaivannikova5464]

t stat is t obt, right?

[@Chelley89]

thank you sooooooooooooooo much!!!

[@lidiyaivannikova5464]

Thanks!!!

[@marcdaniel208]

how to interpret?

[@mingtipparat]

thanks

[@waynegreaves9070]

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.