Dear All,
Could any body help me in finding the best statistical/model to see the un-measured titres of virus in a time points that extend beyond the ones I measured. For instance I am testing the effect of formalin on a virus after 5, 10, 15, 30 min. contact. Now, I would like to speculate whether an extended contact times would reduce or increase the titer, using a model/statistical approach. Could one use linear regression here ?
Thanks
You can use a lot of things the question is
how well it works. it's going to depend on lots of things besides statistics. For example are the virus particles treated in the presence of living host cells.? You can see how the answer to the question leads to a number of possible scenarios that will change the final titre. This will be difficult. BTW I forgot to say that extrapolation beyond a data set using regression is usually not recommended.
Dear Mohamed Samir first of all you should specify that if you are going to use same sample or a different one for all the timings you mentioned here(in example you mentioned above).
Because if you are using same sample you cannot go for linear regression as the sample wold be considered dependent.
I suggest to go for independent samples that is set four different setups for all four times which will eventually means conducting four experiments if you want to go for regression methlogies
Probit analysis is quite a common methodology when it comes to mortality percent studies
Also you can consider repeated measurement methodologies which usually comes handy when same experimental material is needed to be used again and again.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling