With regard to the numerical comparison of estimators in the three packages examined, fixed-effects
algorithms are numerically the same to the available decimal places. Random-effects algorithms yield slightly
different results because of the different methods used to compute the between-regression variance. In this
respect, we have achieved a more confident outcome than that envisaged by McCullogh and Vinod.
In addition, we compared the relative efficiency of the random-effects algorithms provided by the
three packages. This was done by means of a set of suitably designed Monte Carlo experiments, varying
the time span and the number of provinces considered. The outcome of this exercise is asymptotic
equivalence between STATA and TSP, whereas LIMDEP shows a persistent bias up to about one
thousand Monte Carlo replications.
The three econometric software packages are equivalent as regards both usability and numerical
accuracy. TSP seems to be the fastest and the simplest to work with; LIMDEP displays immediately a
more thorough range of statistical tests; STATA probably offers the greatest variety of estimation
commands. The crucial factor in deciding which package to use thus appears to be the experience
researchers have already gained.
18
More intriguing information
1. The Social Context as a Determinant of Teacher Motivational Strategies in Physical Education2. Should Local Public Employment Services be Merged with the Local Social Benefit Administrations?
3. The name is absent
4. The Demand for Specialty-Crop Insurance: Adverse Selection and Moral Hazard
5. The name is absent
6. The name is absent
7. Tissue Tracking Imaging for Identifying the Origin of Idiopathic Ventricular Arrhythmias: A New Role of Cardiac Ultrasound in Electrophysiology
8. Confusion and Reinforcement Learning in Experimental Public Goods Games
9. Banking Supervision in Integrated Financial Markets: Implications for the EU
10. Subduing High Inflation in Romania. How to Better Monetary and Exchange Rate Mechanisms?