Gary King Homepage Previous: What do I need Up: Frequently Asked Questions Next: How do I get

Why do I get slightly different results every time?

$ {\mathfrak{E}I}$ uses random simulation to approximate various statistics. This entails no compromise, and it makes computations easy that would otherwise be impossible. The only issue is ensuring that you have sufficient random draws for the level of precision (number of significant digits) you desire. For example, if you need 2 digits to the right of the decimal point for a table you plan to present, you can check that the number of simulations is set high enough by running the analysis twice and verifying that the first two digits do not change. If they do change, increase _Esims and try again until the two runs give the same answer to two digits. (If you really want exactly the same numbers for each replication, set the Gauss random number seed with the Gauss command rndseed before running $ {\mathfrak{E}I}$, although this would only be useful for testing.) See Chapter 8.



Gary King 2006-09-13