Automatic Content Analysis and Binary Characteristics Detection of Dreams
Abstract
For dream content analysis, automatic quantitative analysis techniques not only can be faster than traditional hand-coding, but also be lower in coding errors and bias caused by humans. Linguistic Inquiry and Word Count (LIWC, Pennebaker, Boyd, Jordan, & Blackburn, 2015) is an automatic technique possibly useful for dream research. We tested the suitability of LIWC for dream content analysis by comparing results by LIWC and Hall Van de Castle coding system (HVdC, Hall & Van de Castle, 1966) using canonical correlation analysis. Moreover, we analyzed the consistencies and inconsistencies between dreaming and waking by comparing the word frequencies in reports. Last, we introduced machine learning techniques to dream research and built support vector machines to achieve the binary characteristics detection of dreamers (e.g., female or male, blind or sighted, waking activity or dream) based on dream content. Our theoretical and methodological contributions to dream research would not only deepen people’s understanding about dreams but also introduce new methods for scientific research on dreams.
Degree
Ph.D.
Advisors
Schweickert, Purdue University.
Subject Area
Linguistics|Artificial intelligence|Cognitive psychology|Higher education|Psychology|Social psychology
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.