Dear Researchers,
I need to analyse biochemical experiments and i am confused which parameters will i use.
1. I performed in vivo and in vitro antioxidant enzyme activity of different treatment of drugs against time.
2. for group, i have untreated (control), treated with various doses of extract and standard drugs treatment (positive control).
3. i have to observe toxicity and antioxidant potential of extract.
What statistical parameters like parametric, non parametric, anova, t test, etc? what will be most appropriate and suitable for above mentioned queries statistically?
Thank you,
ANOVA for comparison within and between the groups and perform mean by Duncan multiple range test with p
You must answer to some qustions before:
the number of observations
Is the variance equal between groups
dépenadance or indépendace of observation
Linear regression gives you the effect on the dependent variable
Anova gives you the comparaison between variances
post hoc test is used when anova result is significant
Dunette post hoc test is used to compare controle withe other groupes
if the conditions are not respected you have to use Kruskal Walis test (non parametric)
and sorry for language if you anderstand french I will explain more good luke.
You want to compare results in study group & control group, is it? Simple descriptive statistics you can apply to get the significance. MS Excel or ANOVA will do.
I guess, this is an exploratory experimental design aiming at identifying the relationship between dose and antioxidant activity.
If the aim was at getting dose-effect information, it would be better to have many different doses (only one experiment per dose). If there are too few doses, you can only perform an ANOVA and cannot find the dose-effect relationship.
If you have enough doses, you can perform a scatter plot to get a look the relationship and use a Generalized Additive Model to describe the relationship.
I guess, your nature of the response variable is categorical. For this kind of data it’s better to use Generalized linear model with binomial distribution this will allow you to determine the relationship between the variables and calculate LD50 and LD90 or LT50 and LT90 or LC50 and LC90 for the toxicity of the drug
Now this is a very fundamental question asked by many researchers involved in bio assays. The statistical calculations (as simple as standard Deviations) speaks a lot about the validity of the results. If you are in the field of research, then I am sure you have some tolerance limits and the data after processing by statistical methods will tell you whether to accept it or not or some time might even suggest you to rethink of setting new tolerance limits.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis