My professor actually spent 3 hours to teach me about your hypothesis testing #1,2,3,4 but I swear I can't even understand a thing. Thanks to your videos, they taught me everything in like half an hour and literally saved me from my final exam.
Thanks so much for making these series, made a daunting topic so much less intimidating. Do you have videos on two sample independent/dependent testing?
Isn't Null Hypothesis supposed to say Mu is bigger or equals to 6? If the alternative Hypothesis is saying Mu is less than 6 then i think the other one should say Mu is bigger is equals to 6!
A null hypothesis set up as an equality represents the most conservative (or least favorable) configuration of the test. The reason is that this setup has the highest probability of a Type I error i.e. rejecting the null hypothesis when it is in fact true. In the example above, you could plot the distributions for mu = 6, mu = 7 and so on and you would find that the probability of Type 1 errors is highest for mu = 6. Therefore, the convention to set up the null hypothesis is to choose the configuration with the greatest probability of Type 1 errors which is done by using the equality sign. You would not get an incorrect answer if you set up the null hypothesis as an inequality. However, when we are testing a hypothesis for possible rejection, if we reject under conditions where the chances of a Type I error are the highest, it automatically follows that the inference holds good for all the other values (since they would have lower probabilities of a Type I error). So, the common practice is to simply test the null hypothesis under the equality condition.
Question: Step 4 says not to reject H0 hypothesis as t and p > 5% significance. But Step 5 says we don't have necessary evidence that delivery times is less than 6 hours. I thought step 4 creates necessary evidence that delivery time is less than 6 hours. Isn't it?
Yaa, Since we don't have enough evidence to reject the NULL Hypothesis, We are saying delivery times is not less than 6 hours. (Even though sample mean = 5.6 hrs, i.e less than 6 hrs). That's what I think is the reason
There's an explanation to this! Mean estimator is considered to be unbiased estimator of the population mean! theoretically its proven that sample mean = population mean! However SD of a sample is a biased estimator! when you find out sample VARIANCE! It equals to (population variance/n) .. therefore it is considered to be biased! hence it in order to remove this biasness, we divide it by n-1 rather than! And the next question would be by minus 1 and not 2 or 3.. It has to do with degrees of freedom for which I'm still searching for logical answers! Furthermore, if the sample size is more than 30, then you divide in by n instead of n-1
Hi sir plz can yiu tell me in this question which test will be used. Three items have following values 12,32,21. Doea the mean of three values differ significantly with the population mean is 56 with level of significance 0.05
Can you help me with this : The Radio Advertising Bureau of statistics reports in Radio Facts, that in 2013 the mean number of radios per Kenya household was 5.6. A random sample of 45 Kenya households taken this year yields the data recorded in Data Desk on number of radios owned. Does the data provide sufficient evidence to conclude that this year’s mean number of radios per Kenya households has changed from the 2013 mean of 5.6?
My professor actually spent 3 hours to teach me about your hypothesis testing #1,2,3,4 but I swear I can't even understand a thing.
Thanks to your videos, they taught me everything in like half an hour and literally saved me from my final exam.
definitely reading your message with an implied full stop after the word "thing" ;)
The link for "why divide by n-1" is here: ruclips.net/video/wpY9o_OyxoQ/видео.html
"VIDEO LINKS TO COME!" I assume they'll be here any minute now. XP
Thanks so much for making these series, made a daunting topic so much less intimidating. Do you have videos on two sample independent/dependent testing?
Isn't Null Hypothesis supposed to say Mu is bigger or equals to 6? If the alternative Hypothesis is saying Mu is less than 6 then i think the other one should say Mu is bigger is equals to 6!
A null hypothesis set up as an equality represents the most conservative (or least favorable) configuration of the test. The reason is that this setup has the highest probability of a Type I error i.e. rejecting the null hypothesis when it is in fact true. In the example above, you could plot the distributions for mu = 6, mu = 7 and so on and you would find that the probability of Type 1 errors is highest for mu = 6. Therefore, the convention to set up the null hypothesis is to choose the configuration with the greatest probability of Type 1 errors which is done by using the equality sign.
You would not get an incorrect answer if you set up the null hypothesis as an inequality. However, when we are testing a hypothesis for possible rejection, if we reject under conditions where the chances of a Type I error are the highest, it automatically follows that the inference holds good for all the other values (since they would have lower probabilities of a Type I error). So, the common practice is to simply test the null hypothesis under the equality condition.
"VIDEO LINKS TO COME!"
They never came...
I cant find the link?
Question: Step 4 says not to reject H0 hypothesis as t and p > 5% significance.
But Step 5 says we don't have necessary evidence that delivery times is less than 6 hours.
I thought step 4 creates necessary evidence that delivery time is less than 6 hours. Isn't it?
Yaa, Since we don't have enough evidence to reject the NULL Hypothesis, We are saying delivery times is not less than 6 hours. (Even though sample mean = 5.6 hrs, i.e less than 6 hrs).
That's what I think is the reason
How can I derive this pValue without using Excel?
Hii Very nice video, Thanks for making it :)
Hi Zed, amazing video! Thanks! One quick question - SD of sample uses (n-1) but why sample mean is always divided by n (not (n-1)
There's an explanation to this! Mean estimator is considered to be unbiased estimator of the population mean! theoretically its proven that sample mean = population mean!
However SD of a sample is a biased estimator! when you find out sample VARIANCE! It equals to (population variance/n) .. therefore it is considered to be biased! hence it in order to remove this biasness, we divide it by n-1 rather than! And the next question would be by minus 1 and not 2 or 3.. It has to do with degrees of freedom for which I'm still searching for logical answers!
Furthermore, if the sample size is more than 30, then you divide in by n instead of n-1
How did you get the p value without excel
Hi sir plz can yiu tell me in this question which test will be used.
Three items have following values 12,32,21. Doea the mean of three values differ significantly with the population mean is 56 with level of significance 0.05
How did you get the -1.833, because I can't find it on the t-distribution table?
i cant believe i just saw his channel get 200k Subs, Im proud of you Justin
is there a rule of thumb of sample size for normal approximation
5:34, its supposed to be 's' and not sigma when writing the formula. Though doesn't matter since it has been solved correctly
Loved this video series.
Sir
Why is it the significance 5% placed to the left?
thank you so much! i forgot the SSD formula!!!
I still did not understand how we know population variance is known or unknown...plz explain again
YOUR SAVING MI A**! THANKS
What if I dont have the standard deviation for either population? Can I still run a significance test?
What's the value of x?--not x-bar but x?
Great video
why are you using sample standard deviation?
No links!
Thank you so much. This is very helpful.
Love the British Accent !
STRAYA M8
@@zedstatistics Wow, just love how you typed the sentence to emphasise your scottish dialect.
Thank you😍
THANK YOU
Thank u so much ❤, my final exam is tomorrow and I wasn't understand these things. You help me a lot, with your simplicity of explaining.
Can you help me with this : The Radio Advertising Bureau of statistics reports in Radio Facts, that in 2013 the mean number of radios per Kenya household was 5.6. A random sample of 45 Kenya households taken this year yields the data recorded in Data Desk on number of radios owned. Does the data provide sufficient evidence to conclude that this year’s mean number of radios per Kenya households has changed from the 2013 mean of 5.6?
Great ousome understanding
how did you get(5.6) mean and (2.27) std values,i dint understand can you please expain it
Use excel
@@royalty5042 you cant use excel during the test that's the problem
@@jadonferrance8940 Oh, we use computer in exams. do you want to know the how did he get values 🙂.?
@@Love-et1zt yes please i would appreciate it
great