[R] Confidence Intervals on Standard Curve

[Ben Ward]

Hi David,

I had use log(x)inside the lm call and used predict, although I didn't 
know about logs of data making a multiplacative model  
exp(log(x)+log(y)) = x*y.
I'll have a look at the poisson model. and see what I manage to produce. 
Looking at the internet the Cumulative distribution function looks like 
a sort of S with pleataus at the top and bottom, which I suppose my 
standard curve would be a part/chunk of? And if I predicted outside of 
what I have, it might become like the full S. I apologise for not 
putting it very elequently in mathematical terms, I'm a biology student 
so stats and hypothesis testing is taught, but doing this standard curve 
is unprecedented in my work so far, normally lm() and anova and such is 
enough.

Thanks,
Ben.

On 19/02/2011 19:06, David Winsemius wrote:
>
> On Feb 19, 2011, at 1:08 PM, Ben Ward wrote:
>
>> Hi Graham,
>>
>> Thanks, that does explain lots. I've been playing with making log's of
>> data in models to make the relationship linear, which it does, which
>> suggests to me that lm() is the right way to go, however, after if  try
>> to predict after y values after about 60% on the x axis for light
>> transmission, the y value, for bacterial numbers, crosses the axis and
>> gives me negative values for y, which on a practical level isn't
>> possible, as one can't have less than no bacteria in a culture.
>
> Once you have your estimated parameter in the transformed "analysis" 
> scale, you need to apply the inverse transformation, in this case 
> exp,   to return the estimate to the measured scale. You also need to 
> consider that in the process of transforming the values, performing an 
> additive estiamtion, and transforming back to the "natural" scale, you 
> will have estimated a multiplicative model, since exp(log(x)+log(y)) = 
> x*y.
>
> Had you used log(x) inside the lm call and then used predict, the 
> predictions would be correct.
>
> You might want to consider glm models with family="poisson".
>