Comparison of Predicted Results vs Actual Observations

[#1]

A data review was conducted to evaluate the impact of animal-diet nutrient density (x) and subsequent final body weight (y).  Data represented a true response of the input across the entire population to be studied.  The regression lines developed were then input into an empirical model to predict body weight response if one or more of the nutrients were changed.  A subsequent study was conducted, and the data review yielded a predicted body weight to be compared to the actual observation of the replicate. 

How do I compare the fit of the observed data to the predicted data from the model?  Model is quadratic with a fixed intercept.

[#2]

There are multiple data visualization methods to '...compare the fit of the observed data to the predicted data...". One place to start is the Fit Y by X platform. The very first graph produced in the default report does exactly what you are after.

[#3]

I'm not sure I understand...I'll re-state what I think you did and perhaps what you want to do.  First, you created a model by using regression to fit y(final body weight) using x(diet nutrient density). It is unclear how large the population is? I don't understand what "final body weight" is and how it is calculated.  Is it when the animal is dead? I assume weight will constantly fluctuate. You have one variable in the model nutrient density.  It is unclear how many levels and how much variation there was in x, but apparently you developed a model from that data.  Did you do any diagnostics for that model?  Did you perform any residuals analysis?

Assuming you developed an appropriate model, you should be able to create a JMP data table with a column for x, a column containing the prediction formula and a column of y(measured weights).  Once you have this data table, there are many graphical and quantitative options to compare the actual to the predicted data (see residuals plots).

[#4]

Thanks for the advice, I appreciate it.