Arnel Aguinaldo

Пікірлер

@bleunoir12 9 ай бұрын

Thank you Dr A! Very clear!

@yogavibes9371 Жыл бұрын

Hi. Can I get the pdf of the commands used? Thank you for the step-by-step explanation.

@devanshibansal1575 2 жыл бұрын

I wanted to know which version of R you're using. Since I'm unable to get these packages on RStudio 4.1.3.

@jamescastle4227 2 жыл бұрын

What if normality is violated and transforming doesn’t fix this? What is the non-parametric model?

@ElowenFaye 2 жыл бұрын

Thank you very much for the video! May I ask something: in my study, I have a computer categorisation task where I want to analyse the incluence of 3 factors (each of them has 2 levels) on the outcome variable (reaction time). it's a within-subjects design so throughout the computer task, each participant will experience every condition. The task consists of 6 blocks; each block has 32 trials (1 trial is categorising one word). Between the blocks, there is just a few seconds break. So my question: do I have to calculate a rmANOVA? I am confused because one the one hand, I have repeated measures on each condition of each participant; on the other hand, the task is completed in a single session (and not with a distance of 2 months or so) and I don't analyse the change of their performance over the 6 blocks but only the influence of the different conditions on the RT. I'd be very very grateful for an answer! :)

@arnelaguinaldo1231 2 жыл бұрын

The design depends on your overall research question and hypothesis. Are you also interested in examining the differences in reaction time across trials within each block? If so, a mixed effects model may be more appropriate to take into account the fixed (3 conditions) and random effects (trials) on RT. If you are only analyzing RT after each of the 6 blocks (or perhaps the average of all the trials within each block), then it sounds like a traditional RM-ANOVA would suffice.

@ElowenFaye 2 жыл бұрын

@@arnelaguinaldo1231 Thank you really much Dr. Aguinaldo, that helped a lot !! :)

@BANGTIDY555 2 жыл бұрын

does X1 refer to participant ID number?

@arnelaguinaldo1231 2 жыл бұрын

Yes, the column should have been labeled "ID" after I imported the data from the CSV file. A lazy oversight on my part!

@jaimecaplette4611 2 жыл бұрын

Thanks alot for this video, its very clear and easy to follow. I was wondering if you wanted to run a 3 way mixed ANOVA can one just include a third variable where you have group*time?

@arnelaguinaldo1231 2 жыл бұрын

In this case, I would recommend using a mixed effects model, which is the general form of the between-within ANOVA. The lme4 package is a very useful tool to perform this in R.

@mohammedal-bader2323 2 жыл бұрын

That was a great lecture..thank you!

@craysoup2086 2 жыл бұрын

Dear Dr. A, Thank you very much for the video! Have a wonderful day.

@shantanutamuly1787 3 жыл бұрын

Dear Dr. Arnel Aguinaldo Thanks for the excellent video. Please make one video on two way mixed design using non-parametric methods

@arnelaguinaldo1231 2 жыл бұрын

There are a few options for performing a non-parametric analog of a two-way mixed ANOVA but there are also many discussions on which one and whether one is needed in the first place. I suggest double checking the normality of your data and if applicable, try data transformation or bootstrapping. You may find the mixed ANOVA would suffice.

@chrislusk3497 3 жыл бұрын

Thank you Arnel, very clear and totally analogous to my own research problem

@mylifeisinhishandsamen4167 3 жыл бұрын

Hi, thank you for the video. Where can I find more explanation that the error code is actually for within subject design? I used the same code but someone said it seems the error is a clustered-adjusted standard error.

@arnelaguinaldo1231 3 жыл бұрын

The error term "X1/time" represents the error variance due to time nested within each subject. If there are unequal number of repeated observations (p) per subject, this nested term takes into account the variance due to random chance per trial per subject. In the example I discuss in this video, however, all subjects in the dataset were tested twice so the 2-way ANOVA with the nested error (X1/time) will output the same results as the within-subject error (X1) term. I just have the ANOVA function "default" with a nested error to account for both scenarios (equal and unequal p).

@andrisuheri2503 3 жыл бұрын

Dear Sir.. I see no sampel dataset and the R-Script has presented.. I hope you could be share in here and many thank..

@rayray0313 3 жыл бұрын

You can also use the “emmeans” package for post hoc analysis. Emmeans(modelname, pairwise~iv, adjust =“Bonferroni”)

@xiaochenliu2326 3 жыл бұрын

Thank you for the vid. When I try to convert my data frame from wide to long, I keep getting this error: Error: Must subset columns with a valid subscript vector. x Subscript has the wrong type `data.frame<. This is frustrating me to no end, I don't know what I'm doing wrong or how to fix it. My data frame is very similar to yours where ID is the first column (I have 6 rows) and after ID I have 90 columns which are my 90 time points. Could you point me to the right direction? Thank you

@arnelaguinaldo1231 3 жыл бұрын

What function are you using that returns the error? What type of data are in the 90 columns? Make sure the column names match when piping the column data into rows for the given key using the `gather` function.

@xiaochenliu2326 3 жыл бұрын

@@arnelaguinaldo1231 Thank you for the reply, I've managed to solve it. Apparently using -c() in the gather(key = Time , value = Temp , -c(ID)) bit. In the columns I had numerical values with one decimal place (these are temperatures in celcius) i.e. 36.7, 39.8 etc for 6 individuals taken over a course of 90mins and my first column is their ID (1, 2, 3, 4, 5, 6).

@amnaghani1352 3 жыл бұрын

Thanks for the detailed and easy to follow video. Could you please make a video/tutorial for non-parametric alternative of two-way mixed ANOVA. Thanks!

Ең жақсы KZbin

Пікірлер