Chosen Procedures within Multiple Logistic Regression Analysis

The paper is devoted to the practical issues within credit-scoring models' estimation (based on the multiple logistic regression analysis), particularly in case of the utilization of financial ratios as the independent variables. There are shown the differences in the resulting models and their predictive abilities, taking into account the economic interpretation of the particular ratios on their range of values and the different possibilities of sampling to estimate model, in the illustrative example in the paper.