I love how Todd simplifies these tutorials and is to the point. This reflects a deep knowledge of the subject without complicating it and making it hard to understand. Good job!
I found this video to be helpful by demonstrating and describing the repeated measures ANOVA. It helped to clarify when to use a one way or repeated measures.
OMG I have just realised that you are the same RUclipsr who uploads the analyses of murderers - so your RUclips channel is both a great help and a great hindrance to me !! A great help with undertaking my Phd with videos like this and a great hindrance because I procrastinate from my PhD watching your murder videos....!! :D
I learned we utilize the Repeated Measures ANOVA when utilizing a data gathering tool that is utilized throughout the study repeatedly. The assumptions of a Repeated Measures ANOVA are: independent observations, data must be normally distributed, independent variables need to be measured by scale.
i have a number of variables that have been measured 3 times on intervention and control group. However, not all participants has taken part in all three measurements. for example, I am missing measurements on one or two of the test points, but they arent all missing from the same test. Can I still use this analysis? If not, how do i filter away these missing datas without "loosing" a great portion of my dataset/have very few participants left to analyze? Thank you for a very informative and easy to understand lecture
Thank you Dr Grande, this has been very helpful. Does the "Program * time" measure of Within Subjects Measures tell us whether the type of program affects the improvement in scores? I'm trying to figure out if I am using the correct method to analyse whether suicide risk (high, moderate, low) affects the level of improvement in depression scale scores at two time points (pre-intervention and post-intervention). Many thanks
Dear Dr. Grande. When I have a doubt in statistics, I go directly to your videos. You are a really amazing teacher. I have big doubt that I have not been able to explain. That is, there are no significant interactions between two groups, but, the partial eta square is medium. Would you explain this or just because the p is not significant I should say that there were no significant interactions and that's it?
Hi! For my dissertation I had to test the hydration difference between 2 hydration creams. I had 14 participants which had 3 testing sites - 1 without product (control) the other with a the first product and the last with the second product. The hydration measurements were taken at baseline before and after shaving their legs (so 2 baseline measurements), 30 min, 60 min and 120min. I think I should use Repeated Measures ANOVA however since there are 2 separate independent variables (time and product) and 1 dependent variable (hydration) I am not sure how to tabulate the data. Do you have any idea on whether this would be an appropirate test and how I should put it on a spreadsheet? Thanks a lot!
Same issue, just reviewing my dat and am now sure how to enter the data I have! Did you get an answer, if you have time to give me insight I would appreciate it.
I have no significant difference for time*group (test of within subjects effects) but then a significant effect for group ( tests of between-subjects effects) what does this mean? Pls help
Hello Dr Grande, thanks for your work. Your videos help a lot. I am using a Repeated Measures Mancova with 4 factors, each one with 8 groups. I also included 3 covariates in the analysis. I discovered there was a significant difference between at least two groups in one of my interactions. I would like to apply a Pairwise comparison to compare separately the groups of my factors, but taking into consideration the interaction. How can I do that in SPSS? Thanks in advance.
Thank you for your very clear video! I have a question tho: What do you put in the ' covariate' field? I have 2 between subjects variables (High vs low exposure and communication strategy) and multiple points in time where I have measured the output. Do I put one between-subject variable in the covariate field?
I wonder how do we do multiple comparisons in this case to know if there is any significant difference between different program at each time point. or between time point within the same program.
if I have four time points and an intervention at second time interval - how do I use spss to find difference between groups as well as individual subjects? thanks
Nice video Dr. Grande. However I tried to do repeated measures with my dataset which consists of 11 factors (Time) and three program. I dodnot got the multivariate results as I have insufficient residuals degrees of freedom, what does it mean and my levenes test only 6 factors are above 0.05.? Should i go for a non normality test.? Thanks
Hi thanks for your video I have a question how can I take a mediation and moderation analysis with within-subjects. I check articles and I found the method of Montoya & Hayes (2017) and Judd & al. (2001) but I don't know how manipulate my variables with SPSS. I have a 2*(condition : sham and active stimulation)*3(1-back, 2-back and 3-back)*2(stimOrder : shamFirst or activeFirst). Is there a videos for this problem ?
Dr. Grande, thank you for your work. Could I use this test to compare a particular statistic (sac hits %/plate appearance) baseball leagues and/or teams over a 5 year period? Thank you
Hi Todd, please I have two factors: drought (with four levels) and residue( with three levels), and say a variable Phosphorus content. how do i arrange the data in order to carry out the repeated measure anova? how can i actually do the RM-ANOVA? Thank you.
What other forms of analysis can we do to account for the loss of power in the RM Anova when participants drop out? This is in context to my dataset which initially consisted of 21 participants, each assessed 4 times - however, by the 4th assessment there were only 12 participants...
Dr. Grande, my data seem so different from all samples you used. I have two independent variables z-scores of precipitation and temperature respectively, and dependent variable is growth response of plants (each column has about 40 values, surely all repeated measurements). Because I am interested in covariances of precipitation, temperature and precipitation x temperature, so thought spss linear mixed model is a right one to get information I need. However, it just did not work, looks like the model treats each value as one dimension variable. So frustrated. If you can help, I would appreciate very much!
Hi, thanks for giving a clear example! I was wondering. We now know that there's a significant difference between the groups, but is that based over all three data collections? In other words, what do you have to do to see the significant differences between groups per time-variable? Thanks in advance
What if you have 5-6 repeated measures within the same group of 30 players, measuring development of muscle strength? like pre- and post-test, and 24 hours, 48 hours and 72 hours after a match...which test to use? Just trying to learn SPSS.
Hey, thanks for this very usefull tutorial! I just have one additional question: How to do that, if the design was a latin square design? How to arrange the data then? Because time1 time2 and time3 would be different for each group. Any solutions for this issue? - Thanks in advance :)
Thanks for the tutorial. one question how to find the difference between two independent groups (different participants) measured by the same instrument at different times when both the independent and dependent variables are continuous in nature.
Hi todd, this video really helped clear up most of my confusion however i am still confused on the post hoc results. In an example, from a repeated two way ANOVA there is a significant difference in the TIME variable . i am confused as in to which independent variable it is for. For example your example There are 3 time periods where data was collected pre, 6 and 12 weeks and two independents experimental and treatment usual. With the post hoc there was a sig difference between baseline and 6 and 12 weeks, however i am not sure how i know whether this for the experimental or the treatment as usual. TIME, there is a significant diff in baseline and 6 weeks for Experimental or is it significant diff in baseline and 6 weeks for treatment as usual? I cant quite understand what this mean? what is this TIME variable is it that experimental and usual combined? I would appreciate any help as i am stuck on how to report my results or how to make out the post hoc results?
Sir, Thank you for you class, my doubt is when there is a significant difference "Between subjects effects", then how do we say which one is better? We can get an Idea from Profile plots, don't we have numerical value to interpret it ?
Dear Dr. Grande, I'm a huge fan of your RUclips channel! I'm now looking for an equivalent of the repeated measures anova with a between subjects factor (3 different groups) and looking into Friedman However, I cannot find if it is possible to integrate the between-subjects factor. Can you help?
It seems that the plot will be easier for readers to understand and interprete if the time is at the horizontal axis, and the value of the two groups is at the vertical axis.
How time taken to reach a specific test score for example 40 in the above example? How is it calculated how many have reached the specific test score i.e 40?
thank you dr so much, it was a useful video for me because it is same design of my project. I have a question.. in some other videos it is called mixed design ANOVA (split-plot). is same statistic test or different??
Repeated measures will show variance in the time frame of tests administered to conclude if they are statistically significant showing the experimental treatment as successful or not. I think repeated measures are great but have more room for threats and errors. How much error is ok to continue using the experimental treatment?
+April Brooks I have that same question April. The homogeneity of the variance test failed for a measure but it is still okay, when do we call it and say the data is no good?
What if you added sex as a categorical variable, in which both male and females receive each treatment and wanted to test that as well? What would you do to still be able to test the repeated measures?
I understand we need to look at significance on Mauchly's Test of Sphericity but since the value is greater than the alpha what does that tell us? What does it mean to meet the assumption?
Hello, I have followed the same tutorial for my data however I have Male and Female for my Independent variable and Dash scores in time (base, 3 month, 6 month, and 1 year). However when I follow the steps, I only get a N of 58 and only male, no female. It only counts the first 58 cells or so. Please help if anyone can.
Inquiry: If I had one group of males and females, and I applied treatment to them and took three measurements If the conditions for a t-test are present, use a two-sample t-test to test the treatment difference between males and females in each measurement Secondly, it is possible to use a repeated-measures test to compare the effect of treatment on males in repeated measurements I repeat this again for females
Hi, Todd, great video, hope you can help me or someone here in the comments. I have results from a experiments, i took measure of mouse weight for several days during a anticancer treatment. I want to know if there is significative diference between treatment each day, for ex Day 1: Tx vs Control, Day 2, Tx vs Control etc...
SPSS provides the output. It is in the table "Test of Within Subjects Effects", but you have to select the type of the calculation you are going to use. If the epsilon value(Greenhouse-Geisser value) is less than .75, you should go with Greenhouse-Geisser correction itself, however if epsilon is bigger than .75, then it is appropriate to use Huynh-Feldt correction. Both have their own effect sizes in the very same table.
I don't think we can say that since both of these numbers should refer to the levels of the factors (independent variables). Maybe we can say 3x3, since both of the independent variables has three levels (pretest, 6 weeks, 12 weeks, but I'm not so sure:) I've never seen such a symbol as AxA in repeated measures ANOVA, but I've seen they use it for Two-Way ANOVA where interaction effect is an issue.
I love how Todd simplifies these tutorials and is to the point. This reflects a deep knowledge of the subject without complicating it and making it hard to understand. Good job!
The best description and explanation of the repeated measure test. Thank you!
I found this video to be helpful by demonstrating and describing the repeated measures ANOVA. It helped to clarify when to use a one way or repeated measures.
Thank you very much for this comprehensive tutorial, Todd. It has already helped and encouraged a lot of students here at TU Berlin!
You're welcome!
Your SPSS videos have been very valuable in preparing me for my doctoral comprehensive exams. You are an excellent teacher, Dr. Grande. Thank you!
The clearest explanation during my searching time.
OMG I have just realised that you are the same RUclipsr who uploads the analyses of murderers - so your RUclips channel is both a great help and a great hindrance to me !! A great help with undertaking my Phd with videos like this and a great hindrance because I procrastinate from my PhD watching your murder videos....!! :D
This was excellent! Thank you for providing a clear example and explaining so much of the data analysis.
Thanks a lot! your explanation is so much helpful especially your slow speech.
Thank you, Dr Grande! This video has been really useful!
I found the graph at the end of the video to be especially helpful in examining the data and understanding it better.
Thank you, needed some freshing up about interpreting the output of RM ANOVA. Now i'm al set to interpret my output 🙂!
I learned we utilize the Repeated Measures ANOVA when utilizing a data gathering tool that is utilized throughout the study repeatedly. The assumptions of a Repeated Measures ANOVA are: independent observations, data must be normally distributed, independent variables need to be measured by scale.
hi dear Todd...your video was great...thank you so much...it helped me so much..
Todd this is great but in future would you maximise screen so we can see clearly and also do the write up pls. Great vid apart from tweaks x
Great video, thank you! Which post-hoc tests to run if our independent variable has more than 2 levels?
i have a number of variables that have been measured 3 times on intervention and control group. However, not all participants has taken part in all three measurements. for example, I am missing measurements on one or two of the test points, but they arent all missing from the same test. Can I still use this analysis? If not, how do i filter away these missing datas without "loosing" a great portion of my dataset/have very few participants left to analyze? Thank you for a very informative and easy to understand lecture
Thank you Dr Grande, this has been very helpful.
Does the "Program * time" measure of Within Subjects Measures tell us whether the type of program affects the improvement in scores? I'm trying to figure out if I am using the correct method to analyse whether suicide risk (high, moderate, low) affects the level of improvement in depression scale scores at two time points (pre-intervention and post-intervention). Many thanks
Why Wilk's Lambda is used as reference and not the other 3? Thanks for answering.
Thank you somuch for the enerprising video.It helped my research analysis a lot.
Thank you, Dr. Grande!
very helpful learning from you clips!!
You're welcome - thanks for watching.
Hi Dr. Todd, I was wondering, what would you do if you have missing values when conducting a repeated measures ANOVA in SPSS ?
Dear Dr. Grande. When I have a doubt in statistics, I go directly to your videos. You are a really amazing teacher. I have big doubt that I have not been able to explain. That is, there are no significant interactions between two groups, but, the partial eta square is medium. Would you explain this or just because the p is not significant I should say that there were no significant interactions and that's it?
❤❤❤thank you so much sir.. excellent teaching 😅
dr todd you are grt teacher.plz enhance voice quality
Thank you for an excellent explanation!
Hi! For my dissertation I had to test the hydration difference between 2 hydration creams. I had 14 participants which had 3 testing sites - 1 without product (control) the other with a the first product and the last with the second product. The hydration measurements were taken at baseline before and after shaving their legs (so 2 baseline measurements), 30 min, 60 min and 120min. I think I should use Repeated Measures ANOVA however since there are 2 separate independent variables (time and product) and 1 dependent variable (hydration) I am not sure how to tabulate the data. Do you have any idea on whether this would be an appropirate test and how I should put it on a spreadsheet? Thanks a lot!
Same issue, just reviewing my dat and am now sure how to enter the data I have! Did you get an answer, if you have time to give me insight I would appreciate it.
Thank you so much! It was really helpful!
How can I ask you my question?
I have no significant difference for time*group (test of within subjects effects) but then a significant effect for group ( tests of between-subjects effects) what does this mean? Pls help
Hello Dr Grande, thanks for your work. Your videos help a lot.
I am using a Repeated Measures Mancova with 4 factors, each one with 8 groups. I also included 3 covariates in the analysis. I discovered there was a significant difference between at least two groups in one of my interactions. I would like to apply a Pairwise comparison to compare separately the groups of my factors, but taking into consideration the interaction. How can I do that in SPSS? Thanks in advance.
Thank you for this video it was tremendously helpful!
Thank you for your very clear video! I have a question tho: What do you put in the ' covariate' field?
I have 2 between subjects variables (High vs low exposure and communication strategy) and multiple points in time where I have measured the output. Do I put one between-subject variable in the covariate field?
I wonder how do we do multiple comparisons in this case to know if there is any significant difference between different program at each time point. or between time point within the same program.
if I have four time points and an intervention at second time interval - how do I use spss to find difference between groups as well as individual subjects? thanks
thanks for the helpful video. I will appreciate if you add videos on how to analyze repeated data with missing values. THANKS!
Nice video Dr. Grande. However I tried to do repeated measures with my dataset which consists of 11 factors (Time) and three program. I dodnot got the multivariate results as I have insufficient residuals degrees of freedom, what does it mean and my levenes test only 6 factors are above 0.05.? Should i go for a non normality test.? Thanks
Hello, did you find a solution to this? I have the same problem
Hi thanks for your video
I have a question how can I take a mediation and moderation analysis with within-subjects. I check articles and I found the method of Montoya & Hayes (2017) and Judd & al. (2001) but I don't know how manipulate my variables with SPSS. I have a 2*(condition : sham and active stimulation)*3(1-back, 2-back and 3-back)*2(stimOrder : shamFirst or activeFirst). Is there a videos for this problem ?
Dr. Grande, thank you for your work. Could I use this test to compare a particular statistic (sac hits %/plate appearance) baseball leagues and/or teams over a 5 year period? Thank you
Hi Todd, please I have two factors: drought (with four levels) and residue( with three levels), and say a variable Phosphorus content. how do i arrange the data in order to carry out the repeated measure anova? how can i actually do the RM-ANOVA? Thank you.
What other forms of analysis can we do to account for the loss of power in the RM Anova when participants drop out? This is in context to my dataset which initially consisted of 21 participants, each assessed 4 times - however, by the 4th assessment there were only 12 participants...
thank you so much, what if the normality of distribution is not met?
Dr. Grande, my data seem so different from all samples you used. I have two independent variables z-scores of precipitation and temperature respectively, and dependent variable is growth response of plants (each column has about 40 values, surely all repeated measurements). Because I am interested in covariances of precipitation, temperature and precipitation x temperature, so thought spss linear mixed model is a right one to get information I need. However, it just did not work, looks like the model treats each value as one dimension variable. So frustrated. If you can help, I would appreciate very much!
Hi, thanks for giving a clear example!
I was wondering. We now know that there's a significant difference between the groups, but is that based over all three data collections? In other words, what do you have to do to see the significant differences between groups per time-variable?
Thanks in advance
what do "time" and "time*program" in the test of within subjects effects mean? how are they interpreted?
I was wondering this too
so ultimately in this example, which treatment had better performance - experiemntal or as usual? what is the significance level?
What if you have 5-6 repeated measures within the same group of 30 players, measuring development of muscle strength? like pre- and post-test, and 24 hours, 48 hours and 72 hours after a match...which test to use? Just trying to learn SPSS.
Hey, thanks for this very usefull tutorial! I just have one additional question: How to do that, if the design was a latin square design? How to arrange the data then? Because time1 time2 and time3 would be different for each group. Any solutions for this issue? - Thanks in advance :)
Thanks for the tutorial. one question how to find the difference between two independent groups (different participants) measured by the same instrument at different times when both the independent and dependent variables are continuous in nature.
Can I use ANOVA with repeated measure if the sample size of the independent variables are not equal ?
Can it be used with only pre-test and posttest?
I do not have the third column.
Hi todd, this video really helped clear up most of my confusion however i am still confused on the post hoc results. In an example, from a repeated two way ANOVA there is a significant difference in the TIME variable . i am confused as in to which independent variable it is for.
For example your example There are 3 time periods where data was collected pre, 6 and 12 weeks and two independents experimental and treatment usual. With the post hoc there was a sig difference between baseline and 6 and 12 weeks, however i am not sure how i know whether this for the experimental or the treatment as usual.
TIME, there is a significant diff in baseline and 6 weeks for Experimental or is it significant diff in baseline and 6 weeks for treatment as usual?
I cant quite understand what this mean? what is this TIME variable is it that experimental and usual combined?
I would appreciate any help as i am stuck on how to report my results or how to make out the post hoc results?
Sir, Thank you for you class, my doubt is when there is a significant difference "Between subjects effects", then how do we say which one is better? We can get an Idea from Profile plots, don't we have numerical value to interpret it ?
I guess you can find numerical value from the descriptives (i.e., the means of different groups)
Hello Dr. Grane. So how it would be with Variable "Group" if it would had 3 Factors?
Thank you Dr. Grande.
if n
This is a *Mixed ANOVA* since there is a between subjects categorical IV along with the within subjects repeated measures.
Dear Dr. Grande, I'm a huge fan of your RUclips channel! I'm now looking for an equivalent of the repeated measures anova with a between subjects factor (3 different groups) and looking into Friedman However, I cannot find if it is possible to integrate the between-subjects factor. Can you help?
It seems that the plot will be easier for readers to understand and interprete if the time is at the horizontal axis, and the value of the two groups is at the vertical axis.
Very well explained, thanks...
You're welcome!
How time taken to reach a specific test score for example 40 in the above example? How is it calculated how many have reached the specific test score i.e 40?
thank you dr so much, it was a useful video for me because it is same design of my project.
I have a question.. in some other videos it is called mixed design ANOVA (split-plot).
is same statistic test or different??
Repeated measures will show variance in the time frame of tests administered to conclude if they are statistically significant showing the experimental treatment as successful or not. I think repeated measures are great but have more room for threats and errors. How much error is ok to continue using the experimental treatment?
+April Brooks I have that same question April. The homogeneity of the variance test failed for a measure but it is still okay, when do we call it and say the data is no good?
What if you added sex as a categorical variable, in which both male and females receive each treatment and wanted to test that as well? What would you do to still be able to test the repeated measures?
I understand we need to look at significance on Mauchly's Test of Sphericity but since the value is greater than the alpha what does that tell us? What does it mean to meet the assumption?
+Amanda Sutton I am curious as well, what it means to meet the assumption!
Hello, I have followed the same tutorial for my data however I have Male and Female for my Independent variable and Dash scores in time (base, 3 month, 6 month, and 1 year). However when I follow the steps, I only get a N of 58 and only male, no female. It only counts the first 58 cells or so. Please help if anyone can.
Inquiry: If I had one group of males and females, and I applied treatment to them and took three measurements
If the conditions for a t-test are present, use a two-sample t-test to test the treatment difference between males and females in each measurement
Secondly, it is possible to use a repeated-measures test to compare the effect of treatment on males in repeated measurements
I repeat this again for females
whats the research qustion of this video ??
Thanks a lot for this great videos. Can please explain how to use dummy variables in RM ANOVA
Great lesson, helped a lot! Sweet thanks!
You're welcome.
It's shocking that SPSS doesn't give the MOST important thing by default, which is the significance test of the group differences by time.
Hi, Todd, great video, hope you can help me or someone here in the comments. I have results from a experiments, i took measure of mouse weight for several days during a anticancer treatment. I want to know if there is significative diference between treatment each day, for ex Day 1: Tx vs Control, Day 2, Tx vs Control etc...
how do we calculate effect size in these type of measures?
SPSS provides the output. It is in the table "Test of Within Subjects Effects", but you have to select the type of the calculation you are going to use. If the epsilon value(Greenhouse-Geisser value) is less than .75, you should go with Greenhouse-Geisser correction itself, however if epsilon is bigger than .75, then it is appropriate to use Huynh-Feldt correction. Both have their own effect sizes in the very same table.
what if we have 3 groups?
Is this a 2 x 3 repeated measures ANOVA?
I don't think we can say that since both of these numbers should refer to the levels of the factors (independent variables). Maybe we can say 3x3, since both of the independent variables has three levels (pretest, 6 weeks, 12 weeks, but I'm not so sure:) I've never seen such a symbol as AxA in repeated measures ANOVA, but I've seen they use it for Two-Way ANOVA where interaction effect is an issue.
Yes, thank you,a very useful video.
What about the normality assumption???
Box's M p value should be greater than .01 to meet assumption.
Thanks for this!
great. Help me a lot. thanks
I'm glad you found the video useful. Thanks for watching.
Patidermos
Hello, Dr.
May I have your help to review my data please?
statistics confuse the hell out of me.. i need help analysing my data XD
You can NOT interpret the results, you just said what appear on screen.