Hi, my dependent variable is the number of questions answered in a privacy sensitive context (0-19). Can I also use the ANOVA for this count data ?

Thank you so much for this video!!!

Thank you

I tried to run your codes from my R-studio, but the following errors were faced as a problem. Error: unexpected symbol in "model1=1m." ### Testing Normal distribution for each group, no major outlines Treat1 <-subset(Diversity_Index_Table,Treat="CON") Treat2 <-subset(Diversity_Index_Table,Treat="ANT") Treat1 <-subset(Diversity_Index_Table,Treat=="D800") qqnorm(Treat1$ace) ########Homogeneity of variances bartlett.test(ace~Treat,data = Diversity_Index_Table) ##One way ANOVA model1=1m(ace~Treat,data=Diversity_Index_Table) Especially is there any package install for...to run "model1.1m"?

Thank you so much. Let me ask you one question. 1. Here you have only one dependent variable, but how can we do if we have more than 2 or 3 dependent variables? 2. How can we test using Duncan or Tukey test in addition to LSD if there is significant variations? 3. How can we give subscription letters to those significant treatments or groups? and draw a box plot by providing subscription letters (like a, ab, c....) in a grouped box plot and adding SEM for each treatment? Thanks again.

You did a fantastic job explaining this. Thanks for helping with my r-stat homework! I have proven that cougars found in boreal forests have a different environmental features in comparison to cougars found in prairie habitat. Yes, it is very obvious which makes it excellent practice data.

In case anyone else is trying the updated script but still getting the same error as me for > datascores <- as.data.frame(scores(NMDS)) #extract the site scores Error in (function (..., row.names = NULL, check.rows = FALSE, check.names = TRUE, : arguments imply differing number of rows: 30, 9 use "datascores = as.data.frame(scores(NMDS)$sites)" instead and it should work :) this is because of an update in the vegan package

Thank you for this video. Your explanation was really useful!

Sir, Thank you so much for this well-explained tutorial on NMDS. I tried to do it with my data and successfully did up to NMDS calculation. when I ran the code datascores <- as.data.frame(scores(NMDS)) it gives me an error given below, Error in (function (..., row.names = NULL, check.rows = FALSE, check.names = TRUE, : arguments imply differing number of rows: 233, 29 Could you please help me to sort out this problem?

Hi. I need help. Which Anova test do I use for a dataset of 3 species of beetles? It would be great if you helped me out a little. Thank you.

you keep saying that its okay that the data are not normally distributed, but the majority on internet say that the data HAS to be. Im so freakin confused, I wouldve have used the sign test but alot dont recommended it either so what is left/which test can be applied for a one sample ordinal data?

when I'm gonna apply qqnorm and qqline for normal distribution analysis R returns this words: Error in (1 - h) * qs[i] : non-numeric argument to binary operator.

Thanks man that really helped!!

Can you provide a peer reviewed publication that says this test requires homogeneity of variance, please?

Great stuff

So understandable

VERY HELPFUL

Explained the best!!!! Thank you!!!

One thing I don't follow - how is the value of 18 arrived at? Is it a nice integer chosen somewhat arbitrarily for teaching purposes? Or is the use of an integer biologically or statistically useful in a way that this video does not cover? Thanks y'all.

I can use Euclidean distance instead of Bray Curtis... I told you this because I was trying to do with an abundance base but at the time of running NMDS I was not able to do it. (mnds <- metaMDS(mnds_1, distance = "bray", k = 2) I got error unlike (mnds <- metaMDS(mnds_1, distance = "euc", k = 2) you can change it to Bray Curtis...

This was exactly what I was needing. Thank you so much!

Hello, I am following along but am getting the error - in if (autotransform && xam > 50) { : missing value where TRUE/FALSE needed. I have no NAs or blank cells in my dataset. Can anyone help me with this?

How to choose the number of permutations if n=24 ?

Easy to follow. Thanks

The plots/samples with 0 species, shall they be part of the matrix?

I have watched many videos but yours is very simplified and easy to understand. Keep making tutorials.

Thank you very much for this video, it's really interesting. Can I have your email?

This was super helpful! Thank you so much!

I got the visual using my own data. thanks! will you also do for other multivariate analysis like PCA, CA, PCO etc?

thank you!! was having huge trouble on windows, this helped a lot

Very nice and clearly explained

Would you please explain (with example) type l, type-II and type=III anova

How do you report it in APA style¿

Thanks for your instructions. May you help me with adding name of centroids in the new version of your code... Once more, so much Thanks...

Why the R have two calls foi dun test ? dunnTest and dunn.test, and this two calls give me different results

Where is the link to access the codes and data

Great video, and great channel in general! Congratulations! One question though: how can we calculate the ES for this test? Everything I search shows how to do it using parametric tests. Thank you in advance!

Such a great video thank-you!! do you know what i'd do to change the points to three colours instead of two?

This is absolutely amazing! You have clarified a problem that I have been struggling to understand for months. I was so close to scrapping this analysis from my thesis. You are a hero!

how would you plot the data if the explanatory variable is discrete?

Hi, thanks for the video! I was hoping to use ANOSIM and SIMPER but have trouble finding any tutorials. Is there any reason why you chose adonis instead? I tried adonis2 once on my data, and it was absolutely terrible, while anosim was significant (but I'm not sure if I ran either of them correctly. I suspect that I didn't meet the assumptions for the permanova). But I also am having trouble figuring out how to interpret the ANOSIM results. :(

You have only two parameters "Forest-size & Highest-tree" in your data and this code (cor.test(Forest_size,Highest_tree,method="pearson", data=datacorrelation) can analyze only two. What should we do if we have more then two parameters in our data?

Hi Great Video I have subscribed. Do you mind if I ask a question please

Thank you!! Help me a lot ! Anticipate more video about PCA or CCA interpretation

My p value in bartlett test come <0.05 what to do next ?

As for the error we take using cld function: 1) install package multcomp: install.packages("multcomp") 2) load the library multcomp: library("multcomp") 3) change the syntax of the cld function: cld(posthoc[[1]], alpha=.05, Letters=letters). Hope it works for you.

The viedo is very good, it crealry explain a important statistical method. I am novice to R and I would say that a lesson that takes 3 or more package to be installed tells us that probably R is not for everyone: every package has its logic, its syntax, and so on, for example when a version change sometimes the syntax is non loger valid, that is probably the case with the error we take in using cld function.

Thank you very much...Very helpful

Thank you VERY much

Thanks. Very useful set of lessons you run here! Liked and subbed