I appreciate these lessons for my stats class being on TH-cam. If we had TH-cam back in the day of my undergrad studies I may have passed Algebra and I'd be a lawyer by now. Everything happens for a reason....
We are using SPSS statistics in my class It sure has made doing statistics more easier then back in the old days. The old school statistics way was using the TI-84 calculator and finding a way to figure out calculations on how one got the answer back when I was still an undergrad student. I sure hated doing it back in the old days. When Exel and SPSS came out it sure made doing math much more simpler. Thanks for the upload bernstmj I cannot wait to see more of your videos.
Thanks, you explained it waaaay better than my spss teacher. However, I sometimes have to choose for "inequal variances assumed", but I don't know when. Could you (or anyone else) please explain? Thanks in advance :)
@Gasttag, you could also run it as a factorial ANOVA if you split people into categorical groups based on income and education (e.g., low, middle, high income).
Thanks you for making this video, it really helps in my dissertation. Besides, i like the way you enlarge the parts that you want to explain in detail which i feel that it makes the whole explanation more easy to understand because it is very clear. Once again, thank you so much for doing this. Appreciate that.
Great video! I became a victim when my university hired some homeless oldman to be our lecturer who apparently had no idea how to teach. Your 6-minute-video explained pretty much everything, and I love how you made the comparison between alien species lol. Which planet are you from!!
Same here man... I know the struggle lol. Really the only problem I have left is: when do I choose for equal variances NOT assumed? what's the choice based on?
the video is good - but you forgot to mention the fact to check out the "sig." BEFORE going to the first or second row... if the test was significant -> we are to go to the 2nd row, or go to the 1st row OTHERWISE
In the last comparison of 3 variables, I noticed that Levine's test for equality was significant in the second one (Income), does it mean that we should look at the bottom line i.e.; the line that says equal variances NOT assumed? This is something that I learnt in another video, is that correct? Granted the P-value is the same but should we be looking at the bottom line instead? Cheers.
+Michelle Long since you posted this michelle, but to support Julian. You seem to look at years of education which has p value .236. Julian asked about p value on income, which is significant. And I also learned in a video from how2stats that if this value is signifikant we must read from the second line (this is mostly for anyone new coming to this video and wonder about this). so I assume Julian is correct in his/her question
Mark is 100% right in theory. In practice, he's still right most of the time too. I made this video originally purely for internal purposes with my students and I briefly covered Levene's in class. I never thought people would watch the videos outside of my classes. Levene's test is a test of homogeneity of variances (is the variance in each of the two independent groups roughly equivalent). If the test is NON-SIGNIFICANT, that means they are similar variances, and you can proceed with the video as described (looking at the top row of the output for the t-statistic, df, and p-value). If Levene's is SIGNIFICANT, that means the variances are NOT SIMILAR and the assumption of homogeneity of variance as been violated. You then look at the BOTTOM row to look at your t-statistic. In theory, this is 100% what you should do. In practice, it's what you should do too (Mark is right, the video is not correct on this). However, in practice, it is very unusual for Levene's test to be significant AND for the t-statistic to be different in terms of your conclusion than if you read the top row (in other words, the top and bottom rows rarely differ even when the Levene's is significant. One more time, you should 100% look at Levene's and, if it is significant, follow the video above except looking at the bottom row. In practice, it rarely makes a difference but Mark (and others) who noted this are 100% correct. When I have time, I will remake this video (I will also remake the video for Repeated Measures ANOVA where I gloss over the same issue for Maulchly's test of Sphericity (for the same practical reasons above). Thanks for the comment Mark (and Alon below) and others.
Thanks. Occasionally screen enlarging were really helpful and effective
I appreciate these lessons for my stats class being on TH-cam. If we had TH-cam back in the day of my undergrad studies I may have passed Algebra and I'd be a lawyer by now. Everything happens for a reason....
Great stuff - clear, straight-forward and concise. Very helpful!
Thank you so much. I will have my statistics exam two days later and this is gonna save my grade. God bless you LOL
We are using SPSS statistics in my class It sure has made doing statistics more easier then back in the old days. The old school statistics way was using the TI-84 calculator and finding a way to figure out calculations on how one got the answer back when I was still an undergrad student. I sure hated doing it back in the old days. When Exel and SPSS came out it sure made doing math much more simpler. Thanks for the upload bernstmj I cannot wait to see more of your videos.
This has helped me so much over the past weeks...thanks so much :-)
Awesome Bernstmj... Keep it up, Great video I ever listen on youtube on statistical topics
Thanks, you explained it waaaay better than my spss teacher. However, I sometimes have to choose for "inequal variances assumed", but I don't know when. Could you (or anyone else) please explain? Thanks in advance :)
@Gasttag, you would need to do an multiple regression to do an interaction of education and income on health (as the DV).
Can you poss. make one for dependent sample t-test? This actually really helped me so much, thank you!
@Gasttag, you could also run it as a factorial ANOVA if you split people into categorical groups based on income and education (e.g., low, middle, high income).
Hey there. I love your explanation and I hope you can make more videos on SPSS :))
Thanks alot for a clear and professional explanation
Thanks you for making this video, it really helps in my dissertation. Besides, i like the way you enlarge the parts that you want to explain in detail which i feel that it makes the whole explanation more easy to understand because it is very clear. Once again, thank you so much for doing this. Appreciate that.
This is an excellent description. I really appreciate it.
you explain everything so well! really really like your video! thank you!
This video was really a savior to me. Thank you so much!
great job..It's really clear to understand.really appreciated.
Your tutorials are perfect. Thank you.
Great video! I became a victim when my university hired some homeless oldman to be our lecturer who apparently had no idea how to teach. Your 6-minute-video explained pretty much everything, and I love how you made the comparison between alien species lol. Which planet are you from!!
Same here man... I know the struggle lol. Really the only problem I have left is: when do I choose for equal variances NOT assumed? what's the choice based on?
the video is good - but you forgot to mention the fact to check out the "sig." BEFORE going to the first or second row... if the test was significant -> we are to go to the 2nd row, or go to the 1st row OTHERWISE
Thanks dude it was really really helpful video ever :) u saved me cheers
@lanmurdo, thanks! I am glad you found it helpful.
Thank you for posting this!
thanks!
Thank you so much
Thank you !
For income, the variance differs, u cant perform t test for that
you are great , thank you very much
In the last comparison of 3 variables, I noticed that Levine's test for equality was significant in the second one (Income), does it mean that we should look at the bottom line i.e.; the line that says equal variances NOT assumed? This is something that I learnt in another video, is that correct? Granted the P-value is the same but should we be looking at the bottom line instead? Cheers.
+Julian Chan Thumbs up!!! You are right.
+Julian Chan I don't believe so, since the p-value .236 is great than .05 we can assume equal variances.
+Michelle Long since you posted this michelle, but to support Julian. You seem to look at years of education which has p value .236. Julian asked about p value on income, which is significant. And I also learned in a video from how2stats that if this value is signifikant we must read from the second line (this is mostly for anyone new coming to this video and wonder about this). so I assume Julian is correct in his/her question
if there is no group , is that mean that it is one sample test??
Thank you so much for uploading this!
May i ask how many minimum sample needed to use this independent sample t-test? Thanks for attention
Thanks its help us.
Best bass upload.
Hi If I want to use T test to compare male and female (self-esteem)
Which T test I must use?
The independent or what?
Thank you kind sir
This is great, thank you so much!
amazing discription
thank you!
Do you consider Levene's test?
THANK YOU!!
Thank you very much.
Don't use this example. He doesn't discuss Levene's test, which can be important.
Mark is 100% right in theory. In practice, he's still right most of the time too. I made this video originally purely for internal purposes with my students and I briefly covered Levene's in class. I never thought people would watch the videos outside of my classes.
Levene's test is a test of homogeneity of variances (is the variance in each of the two independent groups roughly equivalent). If the test is NON-SIGNIFICANT, that means they are similar variances, and you can proceed with the video as described (looking at the top row of the output for the t-statistic, df, and p-value). If Levene's is SIGNIFICANT, that means the variances are NOT SIMILAR and the assumption of homogeneity of variance as been violated. You then look at the BOTTOM row to look at your t-statistic.
In theory, this is 100% what you should do. In practice, it's what you should do too (Mark is right, the video is not correct on this). However, in practice, it is very unusual for Levene's test to be significant AND for the t-statistic to be different in terms of your conclusion than if you read the top row (in other words, the top and bottom rows rarely differ even when the Levene's is significant.
One more time, you should 100% look at Levene's and, if it is significant, follow the video above except looking at the bottom row. In practice, it rarely makes a difference but Mark (and others) who noted this are 100% correct. When I have time, I will remake this video (I will also remake the video for Repeated Measures ANOVA where I gloss over the same issue for Maulchly's test of Sphericity (for the same practical reasons above).
Thanks for the comment Mark (and Alon below) and others.
The fact that I recognized the error is amazeballs! These videos are great. Thank you.
thank you
helpful but frustrating to listen to as there were so many variables. it's just too lengthy just
to say
about t test
if any one help me