I have two waves of data, time 1 measure and time 2 measure. I not only want to determine if there is a mean difference between the two, but are there covariates such as age, BMI, presence of cardiac disease etc. which influence the results? I was told I could do a t-test, but the RM ANOVA would be able to account for the covariates. Do the covariates just reduce the noise, or can they explain the difference?
You should definitely do an RM ANOVA because those covariates have variance to be accounted for in the model. They may impact the main comparison and so you need to know that.
Hello, first, thank you very much for the tutorial, it is super helpful. I do have one question however: you are stating at minute 13:29 "if the p value is below your alpha you have a problem with sphericity". What is or where is the alpha in the given table, I do not manage to follow up in order to be able to compare the p value and alpha. Thank you!
Ah, good question. So you have options for the assumption alpha. You can use your basic alpha of .05, or you can do a more conservative value., like .01 or .001. Choosing a more conservative value means you are assuming the data is robust to a violation of the assumption, in this case sphericity. I usually teach a conservative .001 alpha, because I know I have really violated the assumption if the resulting p-value is more extreme than that. JASP uses a built-in ,05 and will flag assumption checks that are below that with a Note below the table.
Hello! Could you please tell me if i have to order my data like yours in this video in order to be able to put my variables in the cells of the Bayesian Repeated Measures Anova ? Because i'm studying the link between moral injury and sense of agency and i have 2 variables (or columns) ; one named "Memory" with two levels "Neutral" vs "Moral injury" and the variable "Condition" with two levels named "Voluntary" and "Unvoluntary". I also have a column named "Z-scores" with the z-scores of my participants because they had to give a number between 0 and 1000 to measure their sense of agency so I need to ensure that those results are normalized. However, when I go to the Bayesian Repeated Measures menu, nothing happens, I only see the variables "Condition" (categorical variable), "Memory" (also categorical), "Participants" (categorical) and "Z-scores" (numerical). Do I have to have only one lign per participant (I have 14 people in total) and then variables called "Moral Injury-Voluntary", "Moral Injury-Unvoluntary", "Neutral-Voluntary" and "Neutral-Unvoluntary" ? Thank you!
This video is a tutorial for the frequentist/classical analysis for ANOVA. Bayesian ANOVAs are different and set up differently. I don’t do Bayesian stats, so I can’t tell you how to set up your data, sorry!
Hi, thanks for your tutorial! I have a question: I have a continuous independent variable (individual scores, there’s only one score for each participant), and I want to know the effects of the scores and the interaction effects of the scores by other variables (two different within-subject variables, each has two levels), can I input the scores as the Covariate for the repeat measures ANOVA analysis?
Thank you for this great video! I have a question, if you had two different groups (experimental&control), with the same data, and Noise Level would have turned out significant (same as here), where can you see in the repeated ANOVA, which group performed better? Would I see that in the simple main effects? Thanks for your help!
You can't do post hocs in the Repeated measures modules in JASP for within subject variables. So I would just do a series of paired samples t-tests and adjust your alpha with a simple Bonferroni correction.
Hi, thanks for this video. It helps me a lot, but I still have a question (btw, i'm belgian so french native speaker and maybe I didn't understand everything). So, for my thesis, I have to do a repeated measures ANOVA. And my datas are randomised. I have to know if there are an order effect or not (Variable : Condition 1 (O then M) - 2 (M then O)). The name of "RM factor 1" is "nursing home types"; with "level 1" = "CCNH" and "level 2" = "CNH". My question is: to know if there are an order effect or not, do I have to put the variable "Condition" in "Between Subject Factors" ? Hope my question is clear and that you may help me. Have a nice day!
Hi--from what I can gather from your description, to see if you have any order effects, yes you'll input your Condition variable in the Between-subjects box. This will tell you whether your counterbalancing was effective or not. You're looking for a null effect for that one specifically; that way you can rule our any order effects. Cheers!
I wish you were my professor. You explain it perfectly . Thank you !!
You explained everything so well and clearly that I understood everything with my poor knowledge of English. Thank you!
Hey, the link to the data doesn't seem to work.
Not Found
The requested URL was not found on this server.
Aww bummer! Thanks for letting me know
I have two waves of data, time 1 measure and time 2 measure. I not only want to determine if there is a mean difference between the two, but are there covariates such as age, BMI, presence of cardiac disease etc. which influence the results? I was told I could do a t-test, but the RM ANOVA would be able to account for the covariates. Do the covariates just reduce the noise, or can they explain the difference?
You should definitely do an RM ANOVA because those covariates have variance to be accounted for in the model. They may impact the main comparison and so you need to know that.
Hello, first, thank you very much for the tutorial, it is super helpful. I do have one question however: you are stating at minute 13:29 "if the p value is below your alpha you have a problem with sphericity". What is or where is the alpha in the given table, I do not manage to follow up in order to be able to compare the p value and alpha. Thank you!
Ah, good question. So you have options for the assumption alpha. You can use your basic alpha of .05, or you can do a more conservative value., like .01 or .001. Choosing a more conservative value means you are assuming the data is robust to a violation of the assumption, in this case sphericity. I usually teach a conservative .001 alpha, because I know I have really violated the assumption if the resulting p-value is more extreme than that. JASP uses a built-in ,05 and will flag assumption checks that are below that with a Note below the table.
Hello! Could you please tell me if i have to order my data like yours in this video in order to be able to put my variables in the cells of the Bayesian Repeated Measures Anova ? Because i'm studying the link between moral injury and sense of agency and i have 2 variables (or columns) ; one named "Memory" with two levels "Neutral" vs "Moral injury" and the variable "Condition" with two levels named "Voluntary" and "Unvoluntary". I also have a column named "Z-scores" with the z-scores of my participants because they had to give a number between 0 and 1000 to measure their sense of agency so I need to ensure that those results are normalized. However, when I go to the Bayesian Repeated Measures menu, nothing happens, I only see the variables "Condition" (categorical variable), "Memory" (also categorical), "Participants" (categorical) and "Z-scores" (numerical). Do I have to have only one lign per participant (I have 14 people in total) and then variables called "Moral Injury-Voluntary", "Moral Injury-Unvoluntary", "Neutral-Voluntary" and "Neutral-Unvoluntary" ? Thank you!
This video is a tutorial for the frequentist/classical analysis for ANOVA. Bayesian ANOVAs are different and set up differently. I don’t do Bayesian stats, so I can’t tell you how to set up your data, sorry!
Hi, thanks for your tutorial! I have a question: I have a continuous independent variable (individual scores, there’s only one score for each participant), and I want to know the effects of the scores and the interaction effects of the scores by other variables (two different within-subject variables, each has two levels), can I input the scores as the Covariate for the repeat measures ANOVA analysis?
From your description, yes, that sounds like it would be good and possible
@@AlexanderSwan okay, thank you!
Can i ask for the problem and dataset?
I found this dataset online and I don’t know if I still have it, sorry
can you show us the video with covariates? thank you
I have a JASP ANCOVA video here: JASP 0.14 Tutorial: Analysis of Covariance (ANCOVA) (Episode 26)
ua-cam.com/video/kCOHz9WgLBQ/v-deo.html
Can you please help me? you only use "scale" variables??
Scale in this context means "continuous" variables
Please any assistance on how to use JASP RCBD?
I'm sorry, I do not know what RCBD means?
@@AlexanderSwan Randomized Complete Block Design in ANOVA
@@AwuduAbubakari ok, I need more information. Please send me an email -- you can find that on my About page
@@AlexanderSwan I have been trying to get your email but to no avail.
@@AwuduAbubakari if you go to the Contacts page on my website, it's there
Thank you for this great video! I have a question, if you had two different groups (experimental&control), with the same data, and Noise Level would have turned out significant (same as here), where can you see in the repeated ANOVA, which group performed better? Would I see that in the simple main effects? Thanks for your help!
You can't do post hocs in the Repeated measures modules in JASP for within subject variables. So I would just do a series of paired samples t-tests and adjust your alpha with a simple Bonferroni correction.
Hi, thanks for this video. It helps me a lot, but I still have a question (btw, i'm belgian so french native speaker and maybe I didn't understand everything).
So, for my thesis, I have to do a repeated measures ANOVA. And my datas are randomised. I have to know if there are an order effect or not (Variable : Condition 1 (O then M) - 2 (M then O)).
The name of "RM factor 1" is "nursing home types"; with "level 1" = "CCNH" and "level 2" = "CNH".
My question is: to know if there are an order effect or not, do I have to put the variable "Condition" in "Between Subject Factors" ?
Hope my question is clear and that you may help me. Have a nice day!
Hi--from what I can gather from your description, to see if you have any order effects, yes you'll input your Condition variable in the Between-subjects box. This will tell you whether your counterbalancing was effective or not. You're looking for a null effect for that one specifically; that way you can rule our any order effects.
Cheers!
@@AlexanderSwan Great, thanks a lot!
came here to study, why do I hear Childe Snezhnaya talking 😭. Thank you so much for this!