Coefficients (linear models)

1. Click the Coefficients view thumbnail.

   This is a visualization of the traditional parameter estimates table. The chart displays the intercept first, and then sorts effects from top to bottom by decreasing predictor importance. Within effects containing factors, coefficients are sorted by ascending order of data values. Connecting lines in the diagram are colored based on the sign of the coefficient (see the diagram key) and weighted based on coefficient significance, with greater line width corresponding to more significant coefficients (smaller p-values).

   

2. From the Style dropdown of the Coefficients view, select Table.
3. In the parameter estimates table, click the Coefficient cell. This displays the standard error, t statistic, and confidence interval.

   This is the traditional parameter estimates table for the overall model and individual effects. The coefficients show the relationship of each model parameter to Cost of claim in thousands. For example, the value of 0.261 for coverage_transformed means that, all other things being equal, we would expect the cost of a claim where the policy has 200 thousand in coverage to cost (200-100)*0.261 = 26.1 thousand more than a policy with 100 thousand in coverage.

   Similarly, the value of 137.226 for claim_type_transformed=2 (Type of claim=Contamination) means that, all other things being equal, we would expect the cost of a claim where there is contamination to cost 137.226 thousand more than a claim for Wind/Hail or Theft/Vandalism, represented by claim_type_transformed=3; you can see which categories of the untransformed field it represents by hovering over the model term cell to reveal the tooltip.

   

4. Click the Estimated Means view thumbnail.

This gives an alternate visualization of the coefficients so that you can quickly see that, all other things being equal, fire/smoke and contamination are associated with higher claims costs than water damage, wind/hail, and theft/vandalism.

Next