Effectivity of the weighted average method for analysis of respiratory movements in the concealed information test

Abstract

  Respiratory movement is one of the indices for the concealed information test (CIT). This study examined incremental validity of the weighted average method (Matsuda & Ogawa, 2011) to analyze respiration speed (RS), respiration amplitude (RA), and respiration rate (RR). A simple average method of RS is known to be biased by analysis interval. But using the weighted average method can avoid this problem. In addition, it indicated higher detection efficiency of memorized participants compared to the simple average method. With examining time-series data, RS and RR could distinguish two groups in all 20 seconds, but RA could distinguish only in the first 10 seconds. Comparing distinguishability between 10-second average and 20-second average, RS and RR were better in 20 seconds than in 10 seconds, but RA was better in 10 seconds than in 20 seconds. This study showed that the best analysis interval for the weighted average method is different from each of respiration indexes.