Linearity of response- which koeficient? - (Aug/19/2006 )

Hi guyz-me again...
Now I am dealing with question, how to calculate, whether the response I got from my measurements is linear with respect to concentrations of toxicant. I found there exist 2 coefficients-regression and corelation. From calculating correlation coefficient and comparison to values in the table I can say directly, if response and tox. concentrations are linearly correlated, right? Is that OK? Or should I use regression (but I have troubles determining linearity in this case-hmmm). Yeah, allways problems with statistics

---

Who else...
correlation analysis: You want to detect a association between two "equal" random variables (eg. body size and weight). In linear regression you measure the linear relation between dependent (Yn) and independent (Xn) quantitative variables (e.g. reaction time and blood alcohol content). You can predict (if you have a strong relation) y via x. Use coeffcients R^2) for mesurement of power of relation. simple correlation anaysis = linear one-way regression. There are also multiple regressions and non-linear ones...
Hope there are not too much errors, long time ago that I used it .

---

Oh, hi Hob -you seem to be my big saver .Thank you again.
Yesterday I found a great, easy- to- understand website about statistics, so I guess I won't bother you so moch anymore in the future .

---