Scheduled for Free Communication: Assessment and Validation of Instrumentation and Theoretical Models, Wednesday, April 9, 2008, 10:30 AM - 11:45 AM, Convention Center: 202C


Validation of Bioelectrical Impedance Analyzer for Measuring Percentage of Body Fat

Sang-Jo Kang, Youngeun Song, Dae-Yeon Kim, Sae-Hyung Kim and Jae-Hyeon Park, Korea National Sport University, Seoul, South Korea

Bio-electrical Impedance Analyzer (BIA) has been widely used to measure person's body fat because it is a safe, rapid, and accessible method. Limited evidence of validity, however, was provided for this technique. The purpose of this study was to establish validity evidence of BIA for measuring body fat. According to the Standards for Educational and Psychological Testing (AERA, APA, & NCME, 1999), validation is an ongoing process of accumulating various types of evidence to support the intended use and interpretation of test scores. For this study, the validity evidences based on internal structure and based on relations to other variables were established. The BIA instrument, Bio Space Inbody 4.0, was used to obtain percent body fat on 51 women and 49 men in Seoul, Korea (age ranged from 22 to 39). To establish validity evidence based on internal structure through known-difference group design, percent body fat from BIA was compared by gender. To establish the validity evidence based on relations to other variables, percent body fat from BIA was compared to that of Dual-Energy X-ray Absorption (DEXA). Independent t tests, Pearson correlations, and other agreement indexes (i.e., TEM, %TEM, and ICC) were used to analyze the data. Women usually have higher percent body fat than men, so higher body fat percentages were expected from women. The independent t test results showed a significant difference in percent body fat between men (Mean ± SD; 19.6 ± 5.4) and women (28.6 ± 4.7), t = -3.49, p < .001, supporting the validity evidence based on internal structure using known-difference evidence approach. Percent body fat obtained from BIA was positively correlated with those of DEXA (r = .91). There was no statistically significant difference in percent body fat between two measures, t = 1.22, p>.05. Three agreement indexes (TEM = 2.4, %TEM = 10.1, and ICC = 89) also supported the validity evidence of BIA based on relations to other variables. Based on the study results, the BIA method appears to be a valid method of measuring percent body fat for adult population.
Keyword(s): measurement/evaluation, obesity issues

Back to the 2008 AAHPERD National Convention and Exposition (April 8 - 12, 2008)