This document describes a method of copying your module result nodes into a new node, and comparing your old and new module results. An important prerequisite to module submission is module verification. By verification, we mean that you should verify that

1. your module runs without error in its standalone shell.
2. your module results have been compared with those generated by the last version of your module, and you understand (i.e., can explain) the differences.

To compare module results, we suggest the following procedure:

1. Evaluate the node of interest in your original module within a standalone shell. (We suggest using the 'mean' option; if you use'statistics' you'll need to create a dummy statistics index ['Min','Median','Mean','Max','Std. Dev.'] in step 3.) Display the results in tabular form and write down the indexes (in order) of the result table. Use the Export... command from the File pulldown menu to save the result data to a file. Save the file on the desktop and name it descriptively, e.g., 'Visible Range result 1.0'
2. Open your updated module within a standalone shell. Create a new variable node to hold the contents of the data you exported (e.g., 'Original Visible Range result'). Create a table as the node's definition, and index it by the indexes you noted in step 1.
3. Open the table definition window for the new node, and order the indexes so that they coorespond to the order you wrote down in step 1. Then use the Import... command from the File pulldown menu to load the result data into the table. You now have a node in your new module which contains the old result data. You can repeat this process until all result nodes of interest are saved in your new module. The next step is to compare the old and new results.
4. Since your old and new results have identical indexes, it's easy to compare them in Analytica. Just create a new node that calculates a difference, percentage, or whatever between the two nodes. For example, a valid definition could be:
   old_result-new_result
   or
   (old_result-new_result)/old_result

You can then evaluate this node and visualize the difference in model results, across all dimensions, in both graphical and tabular form.