Method for detecting significant patterns in panel data analysis

Data Mining in Institutional Economics Tasks

“…It was shown [4] that the RPT tests fulfill the null hypothesis that the target y is independent on the corresponding explanatory variable. In the case of two-dimensional regularities the combination of two null hypothesis is tested.…”

“…The Bonferroni correction is too pessimistic when it is applied to two-dimensional regularities search. An alternative approach was discussed in [4]. Two dimensional regularities found in a true data set were compared with two-dimensional regularities found in random samples that were obtained from the true data set by target permutation.…”

“…Usually a Bonferroni correction [4] is used to receive more realistic validity estimates: the calculated p-values are multiplied by the number of tested explanatory variables. The multiple testing problem is more difficult when two-dimensional regularities are searched because it is necessary to multiply the calculated p-values by a very large number of tested pairs of variables.…”

“…In [3] the relationship between the dominating type of institutional matrices and the climate or infrastructure variables was studied. In [4] the optimal valid partitioning (OVP) method was applied to compare joint dynamics of macroeconomic variables in groups of X and Y countries. A specially developed technique for evaluating multiple testing effects was used.…”