The use of Chi-squared Automatic Interaction Detector (CHAID) analysis to identify characteristics of agricultural households at risk of financial self-exclusions

Abstract

The aim of the research was to identify the characteristics of agricultural households at risk of financial self-exclusion using the Chi-squared Automatic Interaction Detector (CHAID) analysis. The source of data was the results of a survey conducted among 348 agricultural households of Central Pomerania (Poland). The obtained results prove that nearly 40% of the analyzed households did not use financial services and products of their own choice. Using the CHAID analysis, it was found that the strongest predictor differentiating the studied population in terms of financial self-exclusion was the level of average monthly household income (INC). Significant predictors were also those relating to the size and composition of the household (SHW and SHCH), the propensity to save (SAV) and features related to the production potential of a farm (AREA) and the effectiveness of using this potential, i.e. the value of produced output (PROD). The Statistica 13.3 software was used to build the algorithm and create a graphic form of the decision tree.