Analysis of personality structure by new statistical techniques

2015 Fiscal Year Final Research Report

• Project/Area Number: 26590148
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Educational psychology
• Research Institution: Future University-Hakodate
• Principal Investigator: Hanada Mitsuhiko 公立はこだて未来大学, システム情報科学部, 准教授 (80323385)
• Project Period (FY): 2014-04-01 – 2016-03-31
• Keywords: 性格 / パーソナリティ / 独立成分分析 / 非負値行列因子分解 / 対応分析

Outline of Final Research Achievements

Personality structure was analyzed by statistical techniques in signal processing and machine learning. Analyses were conducted on data obtained by personality-trait scales or on descriptions about personality of famous people using independent component analysis． Factors or dimensions similar to big-five personality factors were obtained from these analyses. Analyses were also conducted on descriptions about familiar friends using nonnegative matrix factorization, and interpretable personality dimensions were obtained. These results show that independent component analysis and nonnegative matrix factorization are effective for analyzing personality structure.

Free Research Field

心理学