edf40wrjww2CF_PaperMaster:Desc
Executive Summary
The settlement from the insurance company was more than reasonable. The mean estimate excess cost of operations was roughly $254,000 but that is just the mean estimate. Unfortunately for Krog’s MetalFab Inc. we have zero confidence in this model’s exact mean estimate of $254,000. We are 95% confident that the estimated excess costs of operation was between roughly $9,600 using the lower 95% and almost $500,000 using the upper 95%. Because of this, the insurance company should have not immediately settled for $254,000 and should have only offered $9,600 as their starting settlement offer. Krog’s got a very good deal here.
The way that I figured out these numbers is by creating a regression to estimate the materials, labor, overhead, and I also created a model for overall cost estimation. The materials regression is a mixed cost model meaning that there are some fixed costs that we have but also have some variables costs that we are able to control based upon output. The R-square for this model is over 99% which is the goodness of fit which explains how well the model is able to explain the sum of the squared errors. The labor estimate is also a mixed cost because there is always a cost even at zero output. We still have some control over this variable as well depending on our output so it is mixed. This also has an R-square of over 99%. Finally, I used a mixed cost model as well for the overhead cost and the R-square was only 85% but I could have tightened this model up by dropping the Feb. outlier but I chose not to so that we could keep the dataset complete. The F-stat for all of the models were high which means that the overall regression model ...