PSPP-BUG: problems with regression calculations

[Martin S Edwards]

Greetings:

I’m using PSPP in a grad stats class.

A number of my students have found that the more recent versions (and by this I mean since March of this year) don’t give correct estimates in Bivariate regression.

In other words, the estimates PSPP spits out are not the same as those computed by Stata or SPSS on the same data.

This includes the constant, coefficients, and levels of statistical significance.

 

Thought you might like to know….

 

MSE

 

_________________________________

Martin S. Edwards

John C. Whitehead School of Diplomacy

Seton Hall University

106 McQuaid Hall

South Orange NJ 07079

Tel: (973) 275-2507

Fax: (973) 275-2519

http://pirate.shu.edu/~edwardmb/