Characteristics of successful gazelles - problems in approaches and methods of analysing the data

This paper focuses on the problems of analysing growth and success of a firm. The data consists of 348 growing SMEs. 75 of them were growing rapidly and highly successful (HGS). The methods include discriminant, regression and logistic regression analysis. In regression models, growth and success seem to be inversely related. Regression analysis is not an appropriate way to analyse HGS firms. Logistic regression and discriminant analysis should consider a priori probabilities to produce reliable results. The use of different methods depends on the design of the study, the characteristics of the data and the validity of the research questions. A robust analysis presupposes refinement of the data, accurate model specification, and competent interpretation of the results. Future research should focus on periods longer than 3-5 years and take into account uncertainty and discontinuity of variables. A qualitative approach where behavioural and management issues will be included is needed.