If you have a sufficiently large sample size and the regressors of
interest are significant predictors, then it is best to leave in the
controls; they do not harm but help consistency (even if only a tad).
What suffers is efficiency (standard errors) and this is amplified in
small sample size conditions. I would (mostly) always err on the side of
caution and include the controls. A couple of things to do before
considering dropping are: (a) to do a Wald test to test whether the
controls are simultaneously different from zero; (b) to do a Hausman
test comparing the consistent and efficient estimator or a Chow test for
the common regressors. See the first series of discussions (eq. 2-5c to
see that consistency is never harmed even if a omitted regressor is not
a significant predictor: Antonakis, J., Bendahan, S., Jacquart, P., &
Lalive, R. 2010. On making causal claims: A review and recommendations.
The Leadership Quarterly, 21(6): 1086-1120.
http://www.hec.unil.ch/jantonakis/Causal_Claims.pdf
Best,
J.
__________________________________________
John Antonakis
Professor of Organizational Behavior
Director, Ph.D. Program in Management
Faculty of Business and Economics
University of Lausanne
Internef #618
CH-1015 Lausanne-Dorigny
Switzerland
Tel ++41 (0)21 692-3438
Fax ++41 (0)21 692-3305
http://www.hec.unil.ch/people/jantonakis
Associate Editor
The Leadership Quarterly
__________________________________________
On 07.03.2013 08:07, James Bernard wrote:
