Bacterial death rate statistics - I need help!

My masters dissertation focusses on bacterial survival time on paper surfaces, using E.coli and S.aureus

When plotting log cell count against time it can be seen that E.coli dies quicker.

The gradient of the line shows the one log reduction in cell numbers and is 0.86 for E.coli and 0.42 forS.aureus.

Each of the data points is the average from six identical repeats for each time interval.

I'm at a loss as to how I can prove that these gradients are significantly different.

Can anyone suggest an appropriate statistical test that will prove this and advice on carying it out?

Many thanks,



TS Contributor
My vote is for generalized linear model. Logit link, normal error.

Logit( E( fraction dead ith trial | species,time) ) = Beta(Species) + Beta(time) Beta(Species X time).

You should look for the test of Species X Time interaction.

Thanks for the quick reply.

Do you know of any online resources where there are exampples of this (preferably similar situations to my own) being put into practice? Stats isn't exactly my strong point and i'm not entirely sure where to start with what you suggested!

Thanks again,