[Gretl-users] Panel data fixed effects model differing diagnostic statistics results

I have estimated a fixed effects model for 4 entities 15 observations on each. In one model (Model 2), the dependent variable has been subject to a prior logistic transform. In model 4, I used the Panel logistic regression fixed effect model in which the dependent variable is transformed within the estimation process. Both re-assuringly produce the same coefficient estimates. However the diagnostics on the test of common intercepts are different. In model 4 there is a significant difference at the 7% level, in model 2 it is clearly non-significant. Why might these conflicting results arise.? *Model 2: Fixed-effects, using 60 observations* Included 4 cross-sectional units Time-series length = 15 Dependent variable: logit_RER01 coefficient std. error t-ratio p-value --------------------------------------------------------- const −2.34361 0.168446 −13.91 1.17e-019 *** EFFORT 0.0121251 0.00258107 4.698 1.80e-05 *** Mean dependent var −1.585677 S.D. dependent var 0.522564 Sum squared resid 7.732874 S.E. of regression 0.374964 LSDV R-squared 0.520034 Within R-squared 0.286351 LSDV F(4, 55) 14.89787 P-value(F) 2.62e-08 Log-likelihood −23.67039 Akaike criterion 57.34079 Schwarz criterion 67.81251 Hannan-Quinn 61.43685 rho 0.184053 Durbin-Watson 1.345259 Joint test on named regressors - Test statistic: F(1, 55) = 22.0687 with p-value = P(F(1, 55) > 22.0687) = 1.79828e-005 Test for differing group intercepts - Null hypothesis: The groups have a common intercept Test statistic: F(3, 55) = 1.76308 with p-value = P(F(3, 55) > 1.76308) = 0.164948 *Model 4: Fixed-effects logistic, using 60 observations* *Included 4 cross-sectional units* Time-series length = 15 Dependent variable: RER01 Robust (HAC) standard errors yhat =~ E(1 / (1 + exp(-X*b))) coefficient std. error t-ratio p-value ------------------------------------------------------- const −2.34361 0.112702 −20.79 0.0002 *** EFFORT 0.0121251 0.00180297 6.725 0.0067 *** Statistics based on the transformed data: Sum squared resid 7.732874 S.E. of regression 0.374964 LSDV R-squared 0.520034 Within R-squared 0.286351 Log-likelihood −23.67039 Akaike criterion 57.34079 Schwarz criterion 67.81251 Hannan-Quinn 61.43685 rho 0.184053 Durbin-Watson 1.345259 Statistics based on the original data: Mean dependent var 0.181867 S.D. dependent var 0.078366 Sum squared resid 0.195932 S.E. of regression 0.255559 Log-likelihood 96.15616 Akaike criterion −188.3123 Joint test on named regressors - Test statistic: F(1, 3) = 45.2268 with p-value = P(F(1, 3) > 45.2268) = 0.00671188 Robust test for differing group intercepts - Null hypothesis: The groups have a common intercept Test statistic: Welch F(3, 28.2) = 2.51034 with p-value = P(F(3, 28.2) > 2.51034) = 0.0789321 Brian J Revell Professor Emeritus Harper Adams University , Shropshire Current Chair of Defra Economic Advisory Panel Tel 01952 815237 Tel: +44 1952 728153 Mbl +44 7976 538712 University: +44 1952 815235 alt: email: bjrevell(a)harper-adams.ac.uk