Repeated Measures Anova In Graphpad Prism

[Alexia Borson]

Considerthe following setup: two mouse strains ("KO" and "WT") have been compared in three independent experiments ("E1", "E2" and "E3"). In each experiment, there were two groups corresponding to the two mouse strains compared, and each group consisted of four different mice. There were all in all 8 mice per experiment, and three experiments (total of 4 x 2 x 3 = 24 mice).

However, I have been asked this question by persons who work on a regular basis with GraphPad Prism. It does feature "repeated measures ANOVA" which is (in the help file) said to be equivalent to random block design, but I find the explanations somewhat confusing:

Some experiments involve matching but not repeated measurements. The term randomized-block describes these kinds of experiments. For example, imagine that the three rows were three different cell lines. All the Y1 data came from one experiment, and all the Y2 data came from another experiment performed a month later. The value at row 1, column A, Y1 (23) and the value at row 1, column B, Y1 (28) came from the same experiment (same cell passage, same reagents). The matching is by row.

Randomized block data are analyzed identically to repeated-measures data. Prism always uses the term repeated measures, so you should choose repeated measures analyses when your experiment follows a randomized block design.

Repeated measures anova is an old technique that assumes the correlation between any two time points are the same (compound symmetric covariance structure). It also requires a correction to be applied to get correct P-values that account for non-independence of repeated observations within subject. Other approaches work better such as the full likelihood methods of mixed effect models and generalized least squares. R provides many approaches to modeling repeated/longitudinal data and to using realistic correlation structures such as exponential declines over longer time spans. Full likelihood methods are also more robust to missing values that are not missing completely at random. My course notes at have a case study worked out using generalized least squares.

I'd be inclined to analyze each of your three experiments separately, with an unpaired t test or nonparametric Mann-Whitney test. Then you can report the results of all three experiments (confidence intervals and perhaps p values) and look for consistency. I suspect that reporting three consistent experiments will be more informative than trying to combine all the analysis into one pooled result.

If you wanted to be fancier, you could put all the data into a two-way ANOVA, where genotype is one factor, and experiment is the other. Showing that the interaction P value is high would demonstrate that the experiments are consistent (the difference between genotypes is consistent for all three experiments). But, from your description, there would be no repeated measures (because no mouse was measured more than once).

What are they? The t-test is a method that determines whether two populations are statistically different from each other, whereas ANOVA determines whether three or more populations are statistically different from each other. Both of them look at the difference in means and the spread of the distributions (i.e., variance) across groups; however, the ways that they determine the statistical significance are different.

Proteins 1 & 2 have the same difference in protein concentration means but different group variances. Alternatively, Proteins 3 & 4 have similar variances but Protein 4 has a larger difference in protein concentration means between the patient groups.

i have 4 treatment and 4 replications. the treatment is temperature which I used 0,2,4,8 degree celcius. the observed data is on lettuce weight loss for every 48 hours after being stored in different temperature. i want to ask how do i insert this data using SPSS. I really need your help

We have got serum dosages of several metabolites and we want to determine if their serum levels differ in a group of patients as compared to a group of healthy controls.

Usually, having just 2 groups, I would have probably imagined using an independent t-test or a Mann-Whitney test, depending on the variable distribution.

However, in our case, the 2 groups (patients and controls) significantly differ in age and male-to-female ratio, as well as in the number of subjects in each group.

Because of the significant mismatches in age and gender, is there a way to compare the various dependent variables (serum metabolites) between the 2 groups (patients vs controls), taking also into account the possible effect of age and gender?

Moreover, do I have to consider any corrections for multiple testing?

Considering there are covariates (age & gender) with difference across patients/controls, it is necessary to control the possible confounding. A generalized linear model analysis with age and gender as covariates (data transformation may be warranted if the measurements do not follow Gaussian distribution) is preferable.

Hi! I have a study with a drug supplementation at different time points (different samples) and a control group (with no drug supplementation). I was thinking a two way anova for my analysis (factors: drug- yes/no, time points as a factor) but the control group has no value in the time point. What should I do in this situation?

Hello. I am conducting a study on a knowledge score with three groups which will be taken a pre-test, then I did an intervention for each groups, and then they will be taken a post-test again. May I ask, what is the best test to use when we are dealing with a before and after data where there is three groups?

hi, my research is about pretest and posttest performance of two experimental groups. should i use t-test or anova?

or can i use ancova, i was planning on making the pretest as covariate for the post test performance

Hi, I hope you can help me out. I have done a experiment with two groups with control variables. To compare the groups I have choose to use regression analysis. But to do the randomization check I have choosen to do a Anova, To be able to find out if an average value of a (control)variable differs between two groups. In this case, could the t-test worked just as good and why? ( my control variables: education level, age, gender). Thanks a lot!!

Hi,

In my study, my alternative hypothesis is that exercise coupled with diet modifications will result in lower monthly weights. The null hypothesis is that exercise coupled with diet modifications will not result in any change. What test should I use and why? I am having difficulty determining this. Thanks.

If the weights are measured only two times (e.g., baseline and after intervention), the t-test can compare the effect of the intervention against control (i.e., exercise coupled with diet vs. exercise only, or exercise coupled with diet vs. diet only). The weight change would be the dependent variable/outcome of interest.

If there are more groups/interventions, such as diet only, exercise only, diet with exercise, and no exercise and no diet, try two-way ANOVA (with or without repeated measurements depending on whether there are repeated measurements or not).

Hi. Good day. In my study, I have a group of 5 people who took a pre-test, then I did an intervention, and then I took a post-test. I wanted to see, considering that I have 1 independent variable and 3 dependent variables. Attention and focus variables are my annoying variables that I measured before and after the intervention. Should I use T-test or ANOVA?

You should perform a paired t-test on the pre- and post-test changes per dependent variable. That is, a paired t-test would be performed for attention, another paired t-test would be performed for focus, etc. Just as a side note, it appears that controls (no intervention) were not included in this study.

I have three different types of cells and each of them was split in exposed to treatment and unexposed one (in triplicates). do I use a t test between each treated and untreated of one sample for each triplicates and then use Anova in order to compare the three cell types?

thanks for the help!

You should use two-way ANOVA with two factors: drug concentration and time points. The data denoting cell viability (e.g., percentage) may need some transformation (e.g., square-root or log) to approximate a Gaussian/Normal distribution.

Hello there! I would like to combine two datasets collected from two different sources (the particpants were not the same). I used the same survey (including all the measurements) for both data collections, how can the t-test result justify my decision of combining the two datasets?

From the perspective of statistics, using the t-test on each data measurement in the survey could work. Make sure to set the significant level (alpha) at a higher level like 0.10. Another analysis method that you could use is principal component analysis (PCA). PCA will help show if there are batch effects. That is, if the two different sources produce very different datasets, the datasets will cluster separately from each other when the first and second principal components are plotted. Learn more about PCA here: -center/pca-analysis/. If there are no batch effects, you will not be able to discern the two datasets from one another based on the the PCA plot.

Hi! Good day. I would like to ask if i should use T-test or ANOVA in determining the significant level of difference of the TEACHERS assessment as to use of Marungko approach as to:

a.age of teacher

b. sex of teacher

c. years of service of teacher

You can use correlation analysis or linear regression model with assessment of self-efficacy/problem solving capacity as a dependent variable, and the mathematics performance as an independent variable.

ANOVA is an appropriate analysis option to compare bacterial growth among three groups. The post-hoc analysis of ANOVA will calculate the significant difference between each pair of groups (for example, boric acid vs. control, boric acid vs. citric acid, and citric acid vs. control).

3a8082e126