00:02 The video explains testing for normality. 01:00 Normality tests check if data fits a normal distribution 01:56 P-value and normal distribution 02:50 Normality tests help to check distribution types 03:52 Large sample size can lead to rejecting the assumption of normal distribution based on small p-values. 04:53 Use QQ plot for better assessment of normality 05:49 Using data tab for testing normal distribution 06:52 Understanding the importance of normality test in data analysis Crafted by Merlin AI.
Can I ask a question about normal distribution as part of the assumptions in ANOVA please? Are we looking at normal distribution for the whole set of data or for each independent variable group? Say we're doing a one way ANOVA without repeated measures and we have 3 groups (A, B, C) who had one dependent variable measure (weight). Should we run normality test for each of A, B, and C ? Thank you
Hi the concept of "degrees of freedom" as commonly discussed in statistics (e.g., in the context of a t-test or chi-squared test) isn't directly applicable to the Shapiro-Wilk test in the same way. The Shapiro-Wilk test doesn't yield a "degrees of freedom" value like some other statistical tests. Instead, the test involves computing the W statistic based on the observed data and expected values under a normal distribution. The critical value to which you compare the W statistic depends on the sample size, and tables or software often provide these critical values for various significance levels and sample sizes. Regards, Hannah
i have question. i have done 2 research hyphotheis and their normaality tests.For eg i have said ho the perceivness . bout social media follows normal and h1 follows not normal.My test shows both my objectives are not normally distributed .so how to interprete in disccussion chapter and conlunsion
wow mam well explaniation about normalilty test, but I have question that for analitical method we have a threshold of 0.05 i.e if results are less than 0.05 we can say that data is not normal and vice versa, but for graphical method what if the data is near to normal distribnution, that when we can cofidently say that our data is normal or not. just like a fix threshold value of p in analytical method.