Paper
Tourism and Hospitality Research (2005) 5, 295–305; doi:10.1057/palgrave.thr.6040029
Academic Papers: How to present results from logistic regression analysis in hospitality and tourism research
Christer Thrane
Department of Social Science, Lillehammer University College, 2626, Lillehammer, Norway, Tel: +41 61 28 82 47; Fax: +47 61 28 81 70; Email: Thrane@hil.no
Revised 18 October 2004.
Abstract
Many papers in hospitality and tourism research use logistic regression as the multivariate estimation strategy. When the results from these models are presented, only the logistic regression coefficients are usually reported. This is unfortunate since these coefficients have no intuitive metric. Hence, it is very difficult both for other researchers and for laymen to evaluate if the effects of the independent variables in question have important effects on the dependent variable or not. By using logistic regression coefficients from two real datasets as input in simple calculations, this paper shows how the question of the independent variables' net and relative importance can be more satisfactorily assessed.
Keywords:
logistic regression, presentation of results, predicted probabilities



