Hi there,
What can be the suitable statistical test when, by Langendorff technique, electrocardiogram measurements of hearts isolated from WT and KO mice are performed?
The experiment set up was as follow:
First measurements without treatment were performed for defined time (control condition)! After that mediator with lower and then with higher concentration were perfused and measurements were generated! Indeed for each heart we have 3 conditions (Control, Mediator lower, Mediator higher concentration).
Question is what statistical test has to be performed in order to check difference between:
WT vs KO (under all conditions) and
WT Control vs WT Mediator low concentration vs WT Mediator high concentration, same for KO?
I was thinking unpaired T-test when WT vs KO is compared!
But how about when the 3 conditions are analyzed within WT and /or KO?
Hope I was clear in my description! Thank you very much for your suggestions! Any help is more than welcome!
best wishes
Hello Mariya
Until an Expert can add an answer :-
1- Statistical Tests of Normality
2- Sample Size to choose parametric test or non-parametric test
3- you said that you will use unpaired T test that means >> Outcome is quantitative variable so :-
A- comparing between 2 means : Independent T test if parametric normal distribution.
B- If more than 2 means , then you will compare by ANOVA ( Analysis of Variance) .. Variance is the square of SD (Standard Deviation) if parametric normal distribution. .
And the table shows to you parametric and non-parametric tests simply.
N.B.:-
-- Sample size has cut-off point as : 25 , 30 , 60 (above this number >> parametric = Normal distribution).
-- These are 2 common types of normality test as :-
1. Shapiro-Wilk
2. Kolmogorov-Smirnov
3. Lilliefors
4. Anderson-Darling
5. D'Agostino's K-Squared
6. Chen-Shapiro
============================
How to test for Normality :
1. When mean = median (whatever there is mode or not)
2. Normal distribution range : mean + SD , mean – SD
3. Histogram shows Normal Bell Symmetrical Shape.
======================
Interpretation of The Shapiro Wilk Test
For the Shapiro–Wilk statistic:
◦ If p is more than .05, we can be 95% certain that the data are normally distributed. (In other words, the null hypothesis is probably true.)
◦ If p is less than .05, we can be 95% certain that the data are not normally distributed. (In other words, the null hypothesis is probably false.)
-- Comparing between qualitative variable you will have to use : Chi Square Test whatever 2 or more ( except in condition that fits with Fisher Exact test).
Please refer to:
1. Landau, S. and Everitt, B. S.(2004). A Handbook of Statistical analyses using SPSS.
2.Howitt, D. and Cramer, D.(2008). Introduction to SPSS.
Regards,
Zuhair
You have repeated measurements taken on individual hearts. This should be modelled as a random effect. The other variables are fixed (genotype: WT/KO and concentration: none/low/high). The scientifically most intersting part ist the interaction of genotype and concentration. Thus, your data should be analysed using a mixed model with the heart (or mouse) ID as random factor and the genotype and concentrations as well as their interaction as fixed factors.
The error model needs to be chosen according to your dependent variable. I don't know what the measure is you want to analyze. You can try a normal distribution and check the residual diagnostic plots. If there are problems appearing, you might think harder what error model or transformation of your data might be appropriate.
- Badges
- Science topic
- Similar topics
- Biological Science
- Molecular Cell Biology
- Culture Cells
- Cell Line