Data set is following exponential pattern instead of showing linear response. Regression analysis is to be carried out with the data set.
Then log transformation is a possible solution. Check the residuals.
Sir, thank you very much or your valuable suggestion. After transforming the data set in to log and plotting them i did regression analysis of data to obtain values for slope and intercept. The values of intercept are high and regression line is not closure to origin. Is it fine to use the regression equation obtained for further assays?
Your regression of a straight line for the response Y on a predictor X with intercept b0 and slope b1 is
Y = b0 + b1*X
By log-transformation of Y you get
log(Y) = b0 + b1*X
b0 and b1 are estimated from the regression analysis. If you anti-log the equation you see what these values mean on the original scale of Y:
exp(log(Y)) = exp(b0+b1*X)
so
Y = exp(b0+b1*X) = exp(b0)*exp(b1*X) = exp(b0)*exp(b1)^X
It is a power curve like Y=A*B^X with A=exp(b0) and B=exp(b1).
Dear All,
Why am I not getting the linear relation ship between the concentration and the response (Area under curve) during HPTLC analysis of some natural products? What will be the best solution for it?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Regression