comparison of single and multiple hypotheses test of aggregation the “pulses” and “sugar and sugar cubes” in urban areas of iran

The main purpose of this study was to investigate the possibility of aggregation different types of pulses as well as sugar, using the single and multiple hypotheses test. the former hypothesis tests include composite commodity theorem (leontief and hicks) and generalized composite commodity theorem (gcct) and the latter hypothesis tests include the bonferroni, simes, holm, and hochberg procedures and the results of mentioned methods were compared. data of the period 20062018 for this study were obtained from the statistics center of iran..the results of multiple tests of bonferroni, simes and hochberg for different types of pulses showed that with the exception of “mixed pea and bean”, other products can be aggregated into the group of pulses. also, based on the results of bonferroni, simes, holm and hochberg, different types of sugar can be aggregated into the group of sugar. the results of the individual hypothesis test are not the same for different types of pulses and different types of sugar. in other words, according to leontief method, the hypothesis of aggregate the different types of beans together was not confirmed, while according to hicks method, this hypothesis was confirmed. similarly, according to the leontief method, the hypothesis of aggregate the different types of sugar together was rejected, while according to the hicks method, this hypothesis was confirmed. the result of the gcct showed that all types of pulses (except “other beans”) can be aggregated into the pulses group. the types of sugar can also be aggregated into the sugar group according to the generalized composite method. based on the results, when the number of observations is low, the use of single tests and specifically the gcct will not show the exactly same results, which confirms davis (2003) finding that the gcct does not guarantee proper aggregation of goods. in these cases, multiple tests would be recommended.

Comparison of Single and Multiple Hypotheses Test of Aggregation the “Pulses” and “Sugar and Sugar Cubes” in Urban Areas of Iran

The main purpose of this study was to investigate the possibility of aggregation different types of pulses as well as sugar, using the single and multiple hypotheses test. The former hypothesis tests include Composite Commodity Theorem (Leontief and Hicks) and Generalized Composite Commodity Theorem (GCCT) and the latter hypothesis tests include the Bonferroni, Simes, Holm, and Hochberg procedures and the results of mentioned methods were compared. Data of the period 20062018 for this study were obtained from the Statistics Center of Iran..The results of multiple tests of Bonferroni, Simes and Hochberg for different types of pulses showed that with the exception of “mixed pea and bean”, other products can be aggregated into the group of Pulses. Also, based on the results of Bonferroni, Simes, Holm and Hochberg, different types of sugar can be aggregated into the group of Sugar. The results of the individual hypothesis test are not the same for different types of pulses and different types of sugar. In other words, according to Leontief method, the hypothesis of aggregate the different types of beans together was not confirmed, while according to Hicks method, this hypothesis was confirmed. Similarly, according to the Leontief method, the hypothesis of aggregate the different types of sugar together was rejected, while according to the Hicks method, this hypothesis was confirmed. The result of the GCCT showed that all types of pulses (except “other beans”) can be aggregated into the Pulses group. The types of sugar can also be aggregated into the Sugar group according to the generalized composite method. Based on the results, when the number of observations is low, the use of single tests and specifically the GCCT will not show the exactly same results, which confirms Davis (2003) finding that the GCCT does not guarantee proper aggregation of goods. In these cases, multiple tests would be recommended.