Mann Whitney U Test in SPSS - Concept, Interpretation, and Reporting Mann Whitney U Test

[@Eudaimonic_life_23]

My PhD Data analysis..all credits to you… from these tests to SmartPLS and reporting! Thank you sir 🙏

[@researchwithfawad]

Most welcome!

[@incantation1321]

Thank you for this video, Sir. You talk about the stuff that matters for research results. Really helped me with my dissertation.

[@researchwithfawad]

Glad it was helpful!

[@incantation1321]

@@researchwithfawad One question - if there is a difference between two groups in measuring a dependent variable say 'I like apples', then, which statistic is used to determine which group likes apples more? Is it the mean, median, or rank sums? Since you said in the video to report the median of the two groups while reporting the U test results, is the median the correct indicator? If Group A has a higher median than Group B, do we say, "Group A with the higher median value like apples more"? So, is it the median? Or do we use the mean, the rank sums, all of them, or some other statistic? Please advise.

[@kiranlakhlan7810]

Interpretation is just fabulous that I needed thanku sir

[@researchwithfawad]

Thanks and welcome

[@seemiTariq]

sir , You made my life easier . its really a great job. keep it up👏

[@researchwithfawad]

Thanks. I am glad you liked it.

[@vishnuprabha4202]

I found class in you ,sir...salute

[@researchwithfawad]

Thanks. I am glad you liked it.

[@aryobimo7327]

Well done sir, great video!

[@researchwithfawad]

Thank you kindly!

[@shawolss1701]

Thank you sir for the clear explanation! Just a quick question, when the p is significant, do I need to report the + or - sign of z value? Thank you in advance sir.

[@researchwithfawad]

Pleasure. I am glad you liked it. Yes please

[@shawolss1701]

@@researchwithfawad Thank you sir.

[@thezatminecraftchannel5187]

What does negative sign of z value mean.kindly update will be if G8 help

[@mahaasal7313]

Hello, I conducted a pre test and post test, I used a catagorical ordinal scale of 7 points, the pre test scored as 6 or 7 for all the samples (study and control) the follow up tests scored as 3,4,5,6,7 on time 2 and 3. Within a group comparison, I will compare each point to the baseline. How I can compare within and between the groups?

[@researchwithfawad]

Please read about Repeated Measures ANOVA.

[@mahaasal7313]

Thanks for your reply, but should I consider the scale as continuous data. I tried to use the marginal homogeneity test but it did not work.

[@Alhamzah_F_Abbas]

Well explained. thank you

[@researchwithfawad]

Glad it was helpful!

[@gopikaozhukayil5920]

How can we do that when we want to do the test with a mean value rather than the median value? Is it possible to do like with mean and one more doubt? Here the data independent and dependent are separately coded. So median identification is simple. How do you identify the median of 2 values in a single variable?

[@researchwithfawad]

For mean values use Independent Sample T-Test.

[@soklyeav6929]

Hello sir, I would like to ask you related to P-value in your videos, why the p-value is 0.09. why u could conclude that they were significant difference ?

[@researchwithfawad]

You can say it is partially significant as it is less than 0.10. (10%)

[@fatmaabdallah1112]

Dear sir, I got this sentence below test statistic table (Not corrected for ties),, Would you pls explain what does mean.

[@researchwithfawad]

In the context of a Mann-Whitney U test, the phrase "Not corrected for ties" refers to a specific situation that may arise when calculating the test statistic. The Mann-Whitney U test is a non-parametric test used to compare two independent groups to determine if there are any significant differences between their distributions. When dealing with tied ranks, it means that there are identical values in both of the samples being compared. For example, if you have the following two sets of data: Sample 1: 10, 15, 20, 20, 30 Sample 2: 12, 18, 20, 20, 25 In both samples, the value "20" appears twice, creating a tie. Ties can occur in non-parametric tests when dealing with ranked data because the original values are converted to ranks to perform the analysis. The Mann-Whitney U test has two different methods for handling ties: 1. Not corrected for ties: This means that when calculating the test statistic (U-value), no adjustments are made for tied ranks. In other words, the ties are treated as unique ranks, and the standard formula for the test statistic is used without modification. 2. Corrected for ties: In this approach, adjustments are made to the test statistic to account for ties in the data. The correction takes into consideration the number of tied ranks and their frequencies to get a more accurate representation of the test statistic. Choosing between the corrected and not corrected methods depends on the specific software or statistical package being used. The corrected method is generally preferred as it provides a more accurate result when ties are present, but some implementations may default to the not corrected method for simplicity or historical reasons. When interpreting the results of the Mann-Whitney U test, it's essential to consider whether the test was corrected for ties or not, as it can impact the accuracy of the conclusions drawn from the analysis.

[@svts_diamonddd572]

when reporting we use the median value? what does the median value shows and what does the mean rank shows?

[@researchwithfawad]

In the Mann-Whitney U test, the median value is not typically reported. Instead, the test statistic U is reported. The median value represents the middle value of a dataset and is less affected by outliers compared to the mean. It gives an idea of the typical value in each sample. The mean rank represents the average of the ranks assigned to each value in the combined sample.

[@emmajocelyn8144]

Does it mean anything when the Medians in both groups are the same ? thankyou

[@researchwithfawad]

It is a sum of rank tests, not a median test. Median may be same. For more detail, please visit stats.idre.ucla.edu/other/mult-pkg/faq/general/faq-why-is-the-mann-whitney-significant-when-the-medians-are-equal/

[@shakilakhan2207]

A/a sir. What if result shows insignificant. P=. 347. Should I report Z score or not. Please guide. Thank you sir

[@researchwithfawad]

No need to report Z. Since there are no differences.

[@shakilakhan2207]

@@researchwithfawad thank you sir

[@drrahila9302]

Informative sir...thank you...any easy way to calculate iqr in SPSS easily like how you calculate median ...apart from getting in descriptive statistics... thanks in advance

[@researchwithfawad]

I am glad you liked it. You can get quartiles through Frequencies in SPSS.

[@drrahila9302]

@@researchwithfawad thank you very much sir

[@deekshayadav8456]

Hello sir, i have my data in which I have four experimental groups and one control group please suggest how I can use this test to show each experimental group is equal to the control group on the baseline measures

[@researchwithfawad]

If you are looking to compare each experimental group to the control group, you will need a categorical variable that against the value to be compared and add in the type of group it is. Then you can run this test for multiple comparisons

[@azizahmad2723]

Dear Sir, I am comparing two groups in terms of different research aspects using the Mann-Whitney test. But I want to add a categorical covariate (discipline) How do I take it into account with this test as a covariate? thank you

[@researchwithfawad]

Thanks for your comment. I recommend you to follow this thread. www.researchgate.net/post/Adjusting_for_Covariates_for_Mann_Whitney_U_Kruskal_Wallis_Test

[@Heedwell1991]

If z value is negative the effect size(r) is negative, how to interpreate r value

[@fatmaabdallah1112]

How to check the confidence level weather its 95% or 99%?

[@researchwithfawad]

By default it is 95%

[@nailabhurgri1893]

I have to compare data between two departments, sample is same but population is different. which test is more suitable. can you please explain

[@researchwithfawad]

If there are two different groups, use either independent sample t-test (Continuous variable is on interval scale/ no significant deviation from normality) otherwise use Mann Whitney U Test.

[@Heedwell1991]

If z value is negative the effect size(r) is negative, how to interpreate r value

[@researchwithfawad]

In your case, if z value is negative, the effect size (r) is also negative, it means that the groups have a negative correlation, meaning that scores in one group tend to decrease as scores in the other group increase.

[@komalzahra3895]

As in non parametric test we say p< .05 is insignificant and p> .05 is significant. (As we see in Chi square test) But here you say .009 is less than .05 and it significant. Please explain

[@researchwithfawad]

We never say greater than .05 is significant. That is a wrong interpretation.

[@komalzahra3895]

@@researchwithfawad In chi square test we interpret this way as it is non parametric test

[@komalzahra3895]

In parametric and non parametric tests . Interpretation for p value is opposite?

[@researchwithfawad]

No. Who told you thay

[@komalzahra3895]

@@researchwithfawad my teacher as she taught us Chi square goodness of test

[@SylviaJoyLagunda]

Where did you get the value of the r?

[@SylviaJoyLagunda]

Thank u so much for this great video btw!

[@researchwithfawad]

Please watch the full video.

[@izzahameerah6430]

Hi is the write up in APA format?

[@researchwithfawad]

Not entirely. You can watch this video for guidance on APA style kzbin.info/www/bejne/o32uh4OOrNaHb9E

[@dishachoudhary9261]

SIR I HAVE TAKEN ONLINE AND OFFLINE RESPONSES AND WANT TO TEST IS THERE ANY SIGNIFICANT DIFFERENCE ..CAN I APPLIED SAME TEST

[@researchwithfawad]

Yes you can.

[@dishachoudhary9261]

@@researchwithfawad, sir, i don't get how to run the test could you please share the process sir.....AS I have two sample 50 response from online and 50 from offline i have made two excel sheets.then how to run

[@researchwithfawad]

Merge the sheets and to differentiate create a variable Mode in which you mention 1 as online and 2 as offline. Add respective values of other variables next to it.

[@dishachoudhary9261]

@@researchwithfawad A big thanks for the response.......I m trying to run sir

[@dishachoudhary9261]

@@researchwithfawad sir could you please refer any video or book where this process is mentioned if possible

[@gokulkrishna792]

👍👍👍

[@researchwithfawad]

Thanks for watching.

[@Heedwell1991]

How ro get p value

[@researchwithfawad]

There is p value in the results.