What is the difference using this method and by running KS test under Legacy dialog? Wich is the most exact? I ask this Because i get defferent Sig values for KS in my sample, using these 2 ways of testing normality. I read about this and what confuses me is that some say things like this: "when you know mean and SD of your sample, you should do KS non param under legacy dialog... etc", but dont we always know mean and SD since we always have data .... ?? Thanks!
Hi Dr. Todd. Because the KS and SW test give different results for Pre-Test, and you said SW is a better test measurement, can we say the data for Pre-Test is NOT normally distributed?
thank u dr. for ur kind help...but I have a question? if the data not under a normal distribution. what should we do in this case? ...I mean should we do the research again to make it normal or not ? ...
I did a pre post test in a case study with total population sampling. Both sets of data shows that they are not normally distributed. What analysis should I use to compare means? Also in your case where pretest is normally distributed and postest is not. How would you compare the means?
But why would we trust K-S test when in large small even small differences can be assumed to be significant and in small sample size they lack power to detect normality?
Can I use K-S test for variables with different sample sizes (e.g. n=243 and n= 360)? I'm studying on seasonal variation in size structure of crabs. Thanks:)
+Xsentos If the the distribution is significantly different from the normal one, than we are supposed to use nonparametric tests, since there is no homogenity in variances.
HI! in the case the pre test is non- normally distributed and thepost test is normally distributed, should I use parametric or non-parametric tests to compare the means?
Using a non-parametric here is an option, but it comes at the cost of statistical power (the ability to detect a difference that is actually there). You may want to consider a data transformation first.
Thank you for your great demonstration, dear Dr. Grande, It would greatly facilitate our learning process--and I would appreciate immensely--if you could share some sample files of both EXCEL and SPSS.
Thank you for your insightful video! I'd like to ask you one question, please. I run normality tests for a variable in prisoners (so special, no general population). No wonder, I had the null hypothesis rejected. Shall I use parametric or non parametric tests for my analysis? Have in mind that I am an undegraduate student and I have not been taught non parametric tests!
Thank you for this great video. Sir, just a curious question. Now after you have run, the KS test and confirmed your data is normally distributed, Whats next? Thank you
what next? what is your mean? the KS ( Normality Test ) is give us information that our data normal our not, then after you get the result now you know that your data can be in parametrik or non parametrik, and each parametrik or non parametrik have their way to get what you need ( T test / Mann-Whitney or anything else ) Sorry bad english
what if my pretest is already 0.000 as well as my post test 0.000. how will that apply to my pretest and my hypothesis? thanks. Lost in the jungle of statistics
+Jepoy Ferraren I think that just applies to the type of tests used in the research. If the distribution is statistically different from the normal one, then we should test our hypotheses using nonparametric tests. But the hypotheses remain the same, we just use the different way to test them.
You are awesome!
thank you so much, you're 100x's easier to understand than my professor
You are quite welcome - thanks for watching -
Finally understand the H0 and H1 hypothesis for this test... Thanks!
Thank you! This video really helps me understanding the K-S test.
Seriously! Thank you!
What is the difference using this method and by running KS test under Legacy dialog? Wich is the most exact? I ask this Because i get defferent Sig values for KS in my sample, using these 2 ways of testing normality. I read about this and what confuses me is that some say things like this: "when you know mean and SD of your sample, you should do KS non param under legacy dialog... etc", but dont we always know mean and SD since we always have data .... ?? Thanks!
Thank you! This is a very helpful video.
Thanks for your video, well done
Hi Dr. Todd. Because the KS and SW test give different results for Pre-Test, and you said SW is a better test measurement, can we say the data for Pre-Test is NOT normally distributed?
thank u dr. for ur kind help...but I have a question? if the data not under a normal distribution. what should we do in this case? ...I mean should we do the research again to make it normal or not ? ...
I did a pre post test in a case study with total population sampling. Both sets of data shows that they are not normally distributed. What analysis should I use to compare means?
Also in your case where pretest is normally distributed and postest is not. How would you compare the means?
Thanks for the video!
What a great video, do you have any advice on how to report K-S test results APA style?
+louise laing did you ever work out how to do it?
Hello, I couldn't find anything online or in stats books about how to write it up so I just reported that the variables were significant (p
After testing the normality in posttest and not having a normal distribution, what should we do?
But why would we trust K-S test when in large small even small differences can be assumed to be significant and in small sample size they lack power to detect normality?
I mean should we not use them in conjunction with PP Plots, Histograms, the values of Skewness and Kurtosis, Z-Skew and Z-Kurtosis?
Thank you! Very helpful.
Dear Sir !
Please tell me that how to calculate the Minimum value from data by using K.S Test?
If K-S test show the normal distribution but Shapiro show that it is non-normal distribution. How to choose between these 2 test???
Can I use K-S test for variables with different sample sizes (e.g. n=243 and n= 360)? I'm studying on seasonal variation in size structure of crabs. Thanks:)
if we consider K-S results, which type of test we should use to compare pretest and posttest, parametric or non parametric tests?
+Xsentos If the the distribution is significantly different from the normal one, than we are supposed to use nonparametric tests, since there is no homogenity in variances.
HI! in the case the pre test is non- normally distributed and thepost test is normally distributed, should I use parametric or non-parametric tests to compare the means?
Using a non-parametric here is an option, but it comes at the cost of statistical power (the ability to detect a difference that is actually there). You may want to consider a data transformation first.
Thank you for your great demonstration, dear Dr. Grande,
It would greatly facilitate our learning process--and I would appreciate immensely--if you could share some sample files of both EXCEL and SPSS.
Thank you for your insightful video! I'd like to ask you one question, please. I run normality tests for a variable in prisoners (so special, no general population). No wonder, I had the null hypothesis rejected. Shall I use parametric or non parametric tests for my analysis? Have in mind that I am an undegraduate student and I have not been taught non parametric tests!
non parametric
Thank you for this great video. Sir, just a curious question. Now after you have run, the KS test and confirmed your data is normally distributed, Whats next? Thank you
what next? what is your mean? the KS ( Normality Test ) is give us information that our data normal our not, then after you get the result now you know that your data can be in parametrik or non parametrik, and each parametrik or non parametrik have their way to get what you need ( T test / Mann-Whitney or anything else )
Sorry bad english
Thanks so much Sir
what if my pretest is already 0.000 as well as my post test 0.000. how will that apply to my pretest and my hypothesis? thanks. Lost in the jungle of statistics
+Jepoy Ferraren I think that just applies to the type of tests used in the research. If the distribution is statistically different from the normal one, then we should test our hypotheses using nonparametric tests. But the hypotheses remain the same, we just use the different way to test them.
+Sanja Bulog thanks very much. I used the wilcoxon t test since my data was ordinal. Thank you for the help.
you are looking for a black cat in the dark room
Also known as vodka test!