Thank you for the video, can you help me how to prove that is unbiased in this question? Question: Compare the average height of employees in Google with the average height in the United States, do you think it is an unbiased estimate? If not, how to prove it is not matched?
Hi Sia, Unfortunately, without knowing whether you have paired or unpaired samples, or see your data - I cannot answer your question. Please follow the logic, especially at the beginning, shown in the video to determine the type of test you need. Dr E.
@@EugeneOLoughlin I have daily data, arranged by weekdays, that means Monday, Tuesday, Wednesday, etc. Example, I want to test whether Monday is significant different from Tuesday. Can I say that, my data is unpaired data?
Hi Sia, Again - if you follow the logic in my video you will have to first decide whether your data are unpaired or not. So if you have two groups - one measured on Monday, and the other on Tuesday, then this will be an unpaired test. If you have one group and measure them on Monday and again on Tuesday - this is a paired test. Dr E.
@@EugeneOLoughlin Sorry, I'm still confused. Actually I have a time series data, which sorted by day, which means Monday, Tuesday, Wednesday, etc. That means I have 7 variable now. If I want to know whether there is any significant difference between Monday and Tuesday, which type of t test should I use? Paired or unpaired t test?
But what value is considered equal or unequal variance? I can't seem to get a straight answer on this. If I have two populations n1=340, sd1=4.3 and n2=235, sd2=5.8 are these equal or unequal?
Hi Lindsey, You need to first conduct a test to check if variance is equal or not - this is called an F Test. Please see my "How To..." video to show you how this works: ua-cam.com/video/-ukqBN5sQSI/v-deo.html Hope this helps, Dr E.
Good afternoon. I have a question, For a t-test my sample size should be than 30 right, but what if I have a categorical data with a sample size of 25000 what test will be appropriate
Hi @olaoluwaomoyebi6772, A sample size of 30-50 is sufficient for a t-test. Larger size is OK, but 25,000 is very big. Be sure that you choose the correct parametric or non-parametric test that is appropriate for your data. Some websites, such as www.scribbr.com/statistics/statistical-tests/, may be useful to help you determine what is best for you. Dr E.
Hi Norjana, At 4:50 in the video I suggest that the Paired t Test is the one to choose for a "before" and "after test. this is the same as pre/post. So this is the one to choose. Hope this helps, Dr E.
Hi Omar, Each of the three types of t Tests described in this video are described by me in separate videos - please check these out in my channel.: ua-cam.com/channels/jXUIvMeiEUna3KTQtYmScA.html Thank you for watching my video. Dr E.
I really enjoy Eugene's Irish accent. It makes learning fun again :D
A saint from the Statistics realm. Thanks!
Very helpful to both teachers and students with explicit explanation
Thank you! really clear and easy explanation
Great concised presentation!
Vers useful Thanks!
This is valuable 7 min, thank you
Thank you so much, you explained it so well.
Thank You very much this was very helpful.
Beautifully put.. Thanks a ton
great video! very clear.
amazing my concepts are related to t test
In addition to the sample means in the numerator, what are the other two variables u1 and u2
Hypothesized difference in mean
does it have specific value for closer /deviate from 1 at 1:49
Thank you for the video, can you help me how to prove that is unbiased in this question? Question: Compare the average height of employees in Google with the average height in the United States, do you think it is an unbiased estimate? If not, how to prove it is not matched?
which type of t-test to choose when i want to find whether there is any significant difference across the weekday?
Hi Sia,
Unfortunately, without knowing whether you have paired or unpaired samples, or see your data - I cannot answer your question. Please follow the logic, especially at the beginning, shown in the video to determine the type of test you need.
Dr E.
@@EugeneOLoughlin I have daily data, arranged by weekdays, that means Monday, Tuesday, Wednesday, etc. Example, I want to test whether Monday is significant different from Tuesday. Can I say that, my data is unpaired data?
Hi Sia,
Again - if you follow the logic in my video you will have to first decide whether your data are unpaired or not. So if you have two groups - one measured on Monday, and the other on Tuesday, then this will be an unpaired test. If you have one group and measure them on Monday and again on Tuesday - this is a paired test.
Dr E.
@@EugeneOLoughlin thanks for the explanation. =)
@@EugeneOLoughlin Sorry, I'm still confused. Actually I have a time series data, which sorted by day, which means Monday, Tuesday, Wednesday, etc. That means I have 7 variable now. If I want to know whether there is any significant difference between Monday and Tuesday, which type of t test should I use? Paired or unpaired t test?
But what value is considered equal or unequal variance? I can't seem to get a straight answer on this. If I have two populations n1=340, sd1=4.3 and n2=235, sd2=5.8 are these equal or unequal?
Hi Lindsey,
You need to first conduct a test to check if variance is equal or not - this is called an F Test.
Please see my "How To..." video to show you how this works: ua-cam.com/video/-ukqBN5sQSI/v-deo.html
Hope this helps,
Dr E.
Thank you!
Good afternoon. I have a question, For a t-test my sample size should be than 30 right, but what if I have a categorical data with a sample size of 25000 what test will be appropriate
Hi @olaoluwaomoyebi6772,
A sample size of 30-50 is sufficient for a t-test. Larger size is OK, but 25,000 is very big. Be sure that you choose the correct parametric or non-parametric test that is appropriate for your data. Some websites, such as www.scribbr.com/statistics/statistical-tests/, may be useful to help you determine what is best for you.
Dr E.
Which type of t-test if I want to know the difference between Pre-test and Post-test?
Hi Norjana,
At 4:50 in the video I suggest that the Paired t Test is the one to choose for a "before" and "after test. this is the same as pre/post. So this is the one to choose.
Hope this helps,
Dr E.
may i know you are using references from what book ?
Is it the same for z test ?
it was a good example but lack of applying it in your question was a little annoying
Hi Omar,
Each of the three types of t Tests described in this video are described by me in separate videos - please check these out in my channel.:
ua-cam.com/channels/jXUIvMeiEUna3KTQtYmScA.html
Thank you for watching my video.
Dr E.
rest in peace earbuds users ... but thanks!
fyi: there's a popup at 4:59... ;-)
@@janlukasr.2141 That scared the shit out of me lol
👍
Good grief that was confusing