This is probably the most straightforward explanation about Standard Deviation, Standard Error, Confidence Intervals, and Error Bars. Thank you so much!
Hi great video. I’m writing a practice paper - and the question is asking me to explain (using info from the graph provided) weather this is a significant difference between (the average temp rise) it’s a bar graph with standard deviation error bars. The biggest bar has a massive error bar with the middle of it meeting the top of the bar. I think that one has a large variation in results - so there is a significant difference? Another bar is very small and has a small bar. But the error line almost takes up a third of the bar. So this one is much harder to tell. Maybe there is a significant difference maybe there isn’t??? Okay so this is a hard question and I’m wondering if anyone can lead me in the right direction. Thank you :)
For the error bars is it important to consider the underlying probability distribution? I assume that for the confidence interval somebody already thought about that in order to get such interval.
I am confident in these interpretations for normal distributions, I'll have to think about it a bit to see if the same is true for other situations. I'm not sure error bars are a wise choice in other situations, given how much people are used to seeing them associated with normal distributions. The concept of "Standard error" is "standard" because it is closely tied to the rules that "standard" normal curves produce (see our video here: th-cam.com/video/WpkTm47thG4/w-d-xo.html ). Hope this helps!
Confusing. Why not use one example, stick lots of different kinds of error bars on it, and show what the different bars would mean using the same backdrop. Your like begging earned you a dislike too.
Thanks so much @Dazzletoad. Your comment is engagement which promotes us on TH-cam (and more than cancels out your dislike). I can tell you are confused and genuinely want to help. In teaching, it is better to give more than one example and that's especially important with a topic like error bars when they are used in graphs that can look radically different. As I'm sure you know, we need many types of graphs to represent information in the way that is most accurate. It's important to use a variety of graph types so that learners can recognize error bars in a variety of contexts.
This is probably the most straightforward explanation about Standard Deviation, Standard Error, Confidence Intervals, and Error Bars. Thank you so much!
most simple explanation ! made my life and thesis easier
Very helpful.. nice explanation
Thank you for the helpful video!
So helpful! Subscribed
Amazing explanation!! Thank you a lot
Well explained. Thanks a lot!
Thanks for sharing!
Does the length of the error bar equal twice the error?
very helpful thank you
Hi great video. I’m writing a practice paper - and the question is asking me to explain (using info from the graph provided) weather this is a significant difference between (the average temp rise) it’s a bar graph with standard deviation error bars. The biggest bar has a massive error bar with the middle of it meeting the top of the bar. I think that one has a large variation in results - so there is a significant difference? Another bar is very small and has a small bar. But the error line almost takes up a third of the bar. So this one is much harder to tell. Maybe there is a significant difference maybe there isn’t??? Okay so this is a hard question and I’m wondering if anyone can lead me in the right direction.
Thank you :)
Did you find the answer? I think the best way is to do a t-test.
For the error bars is it important to consider the underlying probability distribution?
I assume that for the confidence interval somebody already thought about that in order to get such interval.
I am confident in these interpretations for normal distributions, I'll have to think about it a bit to see if the same is true for other situations. I'm not sure error bars are a wise choice in other situations, given how much people are used to seeing them associated with normal distributions. The concept of "Standard error" is "standard" because it is closely tied to the rules that "standard" normal curves produce (see our video here: th-cam.com/video/WpkTm47thG4/w-d-xo.html ). Hope this helps!
Good content 👍🏻
Thank you
great video
what about percentage uncertainty error bars
how do we do that "formally"? 3:00 I mean how do we calculate the error without error balls and eye balls?😁merci
haha I loved your confession
You are amazing thank you so much!!!!
Thanks Paula!
Confusing. Why not use one example, stick lots of different kinds of error bars on it, and show what the different bars would mean using the same backdrop.
Your like begging earned you a dislike too.
Thanks so much @Dazzletoad. Your comment is engagement which promotes us on TH-cam (and more than cancels out your dislike).
I can tell you are confused and genuinely want to help. In teaching, it is better to give more than one example and that's especially important with a topic like error bars when they are used in graphs that can look radically different. As I'm sure you know, we need many types of graphs to represent information in the way that is most accurate. It's important to use a variety of graph types so that learners can recognize error bars in a variety of contexts.
More examples enhance understanding, so it's good he used more than one example
Thank you