This is one of the best videos about the subject I've seen. Very well structured and explained. Thank you.

Thank you the video really helped me in understanding the concept and apply it in my work

Thanks for the nice vid, very easy to follow. I especially liked when the kids came in asking for mama and you had a hard time not laughing!

Working Mom's Rock!!! Thanks for the video!

This is great. Very clear. And it made me laugh.

Such a helpful video, thank you very much!

thanks for the vid

Its great explantion and loved children little voices in the back!

Greetings! Thank you for your upload, about which I would like to agree with others that it is instructional and well-paced, especially for non-statisticians, such as myself. I'd like to ask a "naive" question. For a repeated measures interventional experiment, on a SINGLE population (one group, same population pre vs.post), where EIGHT DEPENDENT variables, at TWO LEVELS each and further divided into two sublevels (with pre post level for each; total of 36) as well as one Level with no sublevel (total 8) (grand total of 36+8 = 44 ) coud be analysed using Manova IFF a high level order of measurement (interval or ratio) is used, can one still attempt to use MANOVA for ORDINAL type measures as concerns ALL the variables and obtained from a LIKERT scale (same 5-item scaling, but DIFFERENT responses, depending on the sublevel), given that recent research seems to be in favour of allowing the use of ANOVA in such cases (please see the reference below). If NOT, WHAT then can be used as an "equivalent" to MANOVA, for a "one-shot" analysis of such ORDINAL variables (there is a technique known as ODA, but I'd like an equivalent in "classical" statistics, if possible). Thank you very much. G. Electronic Journal of Business Research Methods Volume 12, Issue 1, 2014, Pages 16-26 The impact of non-equidistance on Anova and alternative methods (Article) Lantz, B. Chalmers University of Technology, Gothenburg, Sweden Abstract The normality assumption behind ANOVA and other parametric methods implies not only mound shape, symmetry, and zero excess kurtosis, but also that data are equidistant. This paper uses a simulation approach to explore the impact of non-equidistance on the performance of statistical methods commonly used to compare locations across several groups. These include the one-way ANOVA and its robust alternatives, the Brown-Forsythe test, and the Welch test. We show that non-equidistance does affect these methods with respect to both significance level and power, but the impact differs between the methods. In general, the ANOVA is less sensitive to non-equidistance than the other two methods are and should therefore be the primary choice when analyzing potentially non-equidistant data.

Thank you for the presentation. my dependent variable consists of 4 likert-scale questions for 1 item( customer satisfaction), and a nominal variable (bussiness education). I made a composite variable with my dependent variable (mean of the 4 questions). To check the significance, i have used chi-sqiare, but the p-value is much much more greater than 0.05, and also for all my independent nominal variables. Can you please suggest me an approptiate measure of significance with the type of variable i have. Thank you.

I really am asking for someone to help me!! I've discovered a possible math connection at the heart of hollywood. Does 3x + 1 ring any bells? The 'Collatz Conjecture'?

Hi I'm wondering if I can use Cramer's V for nominal and ordinal (attitudes/satisfaction)? You've said it can be used for nominal by ordinal, but most textbooks said it can only be used for nominal data (so shows on spss). I'm very confused please help. Thank you

how about spearman rank correlation for measuring association ??

Can ordinal data used for chi square test ... (I think only nominal / categorical data is used for chi square test) pls clarify