Output after run
When the calibration process is complete, PEST provides some additional statistics. These and other information are shown in additional panels after PEST has finished the optimization.

Run Details
PEST journalizes all important information about setup and results of a PEST run into a run record file, which is shown in full in the Run details panel. 
Parameter Sensitivities
For all iterations, PEST journalizes composite parameter sensitivities in the sensitivities file, which is shown in full in the Run Details panel. This information can be used to check for example for hypersensitive parameters. The composite sensitivity of a parameter is a measure of the sensitivity of all model outputs for which there are corresponding observations to this parameter. By inference, it is a measure of the information content of the calibration dataset with respect to this parameter. 
Observation Sensitivities
PEST writes the observation sensitivities of the last iteration to an observation sensitivities file, which is shown in observation sensitivities. 
Residuals
A comparison of the residual (departure of simulated value from the observation) of each observation with the observed and measured value is shown in the Residuals tab. This information is useful for the identification of poorly calibrated observations. The simultaneous display of weighted residuals allows checking if the choice of observation weights is appropriate.
The content of this tab is identical to the residuals file created by PEST. 
Covariance and Correlation matrix
These tables show the covariance and correlation between the parameters, respectively, the latter being normalized to values ranging from 0 (no correlation) to 1 (strong correlation). A colour gradient facilitates interpreting the values.
Covariance and Correlation matrices are not shown if regularization is activated. 
Eigenvectors and Eigenvalues
Each observation contains a certain amount of information that contributes to the identification of the calibrated parameters. Because observations and parameters are often correlated, respectively, the information contributed from a particular observation might overlap with the information that was already provided from a different observation. As a result the number of observations is not necessarily proportional to the combined information of these observations. Consequently, the solution of the calibration (the inverse problem) might be nonunique even if the number of observations exceeds the number of parameters.
If the covariance matrix undergoes principle component analysis, orthogonal combinations of parameters can be identified, together with the extent to which these combinations have been informed by the calibration process. These combinations are the Eigenvectors of the covariance matrix. If a low eigenvalue is associated with this Eigenvector, then its postcalibration variability is low. Conversely, if a high eigenvalue is associated with an Eigenvector, then its postcalibration variability is high. The ratio of highest to lowest eigenvalue is a measure of the extent to which the inverse problem approaches illposedness. If this ratio is more than about 5e7 then the problem can be considered to be illposed. Furthermore, unless some kind of regularization is being employed (Tikhonov or SVD) PEST will not be able to solve this inverse problem. Instead its behaviour will be numerically unstable, and the objective function may not fall at all.
The contents of the text panels can be easily transferred to a spreadsheet program for further processing and visualization. 