Non parametric Kruskal Wallis test + POST HOC TEST in SPSS TUTORIAL

Kruskal Wallis test, SPSS, non-parametric test, post hoc test, parametric Nova test, statistically significant difference , medians of three or more independent groups, ranks, samples , not normally distributed, groups are independent, multiple comparison combination of the group, P-value, pair-wise, null hypothesis, alternative hypothesis, mean rank, degree of freedom, significance level, box plot,
COMPARING 3 GROUPS OR MORE
GROUPS ARE NOR NORAMALLY DISCTRIBUTED
DATA ARE NON PARAMETRIC
See also
Kruskal Wallis H test assumptions
• Kruskal Wallis H test ...
Whether you are an undergraduate or postgraduate student enrolled in Bachelor, Master or PhD study or if you are employed staff, researcher and whether you are in the field of Engineering, Medical, Biomedical, Health sciences, Economics, Business, Marketing or Social study or any other filed and have qualitative (questionnaire, survey) or quantitative (numeric) data to analyze and you want to become professional in statistical analysis using Statistical Package for the Social Sciences (SPSS) software. Follow my TH-cam channel link and subscribe to get the latest released videos that are uploaded frequently. The channel will cover all aspects of data analysis gradually. If you are interested, then follow the link below and subscribe to view and select the topic you want to learn in SPSS.
/ @statisticstutor

[@StatisticsTutor]

You are welcome, subscribe to get more as there will be more videos on similar to this video with more description and how to write it up in a document

[@chamanthawickramarathne2029]

Thank you sir for your precious explanation on the test. Really appreciate!

[@shijahaneef8945]

thankyou sir..you explained it very clearly without making any unwanted explanation to make the class longer

[@startupeco2257]

Thanks a lot Dr! I would be thankful if you also made a video on how exactly we can find out whether our data are normally distributed because this will tell us whether we need ro run a parametric or non-parametric test.

[@StatisticsTutor]

I have 220 videos on the channel. There are dozens on how to check normality of data. Subscribe to get latest one.

[@ujjwalsharma7283]

You can perform a Shapiro-Wilks Test for checking the normality of the data. The null hypothesis for this test is that the data is normally distributed. So you want the p value to be more than 0.05 at 95% confidence interval to accept the null hypothesis.
Go to analyze => Descriptive Stattistics => Explore => in the dependent list put your variable (age, height, income, or whatever it is) => options => click and check on 'Histogram' and 'normality plot with tests' => continue => ok, and you are done.
Note down the value for Shapiro-Wilks and you are ready to assess whether your data is normally distributed or not.

[@zwelonkeericbeato7637]

Thank you Doc, anyone know how to compute compact letter display?

[@asz9608]

I assume the Dunn test was used for the post hoc test?

[@penelitiiseng2338]

Welldone, sir..! Thank you so much!!

[@_Ahmed__Hesham_]

شكرا يا دكتور

[@drasifMuhammed]

Thanks a lot, doctor. Very much usual video.

[@StatisticsTutor]

Subscribe to get new releases

[@shreedevihalawar2112]

Thank you so much sir for valuable video

[@gamalnasser5575]

Can the Bonferroni test be used after the Karuskal-Wallis test? Does the Bonferroni test require that the data depend on a normal distribution or not?

[@agnesberesinski3429]

Thank a lot, but I have a bit special case, in about 90% of my cases after the rejection of the null hypothesis one sees thanks to pairwise comparisons (post-hoc with Bonferroni corecture) which groups differ significantly. However, I have a few cases where KH rejects the null hypothesis (somewhat close but already 0.044) but pairwise comparisons show no significant differences. I thought I might look at the post-hoc values without Bonferorroni correctors and base the significance of the differences on them, because some of the significant values are already below 0.05. Is it possible to understand such an approach? Are the corrected significances according to Bonferroni possibly too imprecise if KH is just below 0.05?

[@izzatulain8692]

thank you sir for the video. however, I have one question. the significant level after post hoc test is still 0.05?

[@martynastasiak9959]

Can the Kruskal Wallis test statistic (H) be negative? What does it mean? You have H=-9.500 p=0,012 (Control- 1 week)

[@aprilroseveniegas5539]

I am having a difficulty on what letter/symbol I should use in my APA Format report in Post Hoc. Should I use H or F?

[@aprilroseveniegas5539]

And oh, how do I do it?

[@harlem3344]

Can the grouping variable be ordinal? Or only nominal accepted

[@StatisticsTutor]

It is for both

[@StatisticsTutor]

Ordinal and nominal variables

[@ilannmagalhaesdemorais5150]

😀 !