Examining the opinions of potential consumers about plant‐derived cosmetics: An approach combining word association, co‐occurrence network, and multivariate probit analysis

Abstract

Abstract This study explores the opinions of potential consumers of cosmetic products containing plant‐derived ingredients and examines the impact of personal characteristics on those opinions. A word association approach was used to collect opinions about plant‐derived cosmetics from 901 Japanese females. Word co‐occurrence network analysis was used to identify the main themes in responses, including sensory properties and efficacy of plant‐derived cosmetics. Multivariate probit analysis was performed using the themes identified as dependent variables and personal attributes of respondents as independent variables. The results indicate that while respondents viewed plant‐derived cosmetics to be gentle to the skin and safe to use, they also viewed such products to be expensive and to have a slow effect. Some respondents, especially older women, were concerned about potential negative effects. Older respondents, full‐time workers, and respondents with high household income were less likely to view plant‐based products as expensive. Practical applications This study demonstrates that word co‐occurrence network analysis using a community detection method can be a useful tool for systematically identifying the main themes in responses to open‐ended survey questions. By combining this method with discrete choice analysis it is possible to examine the influence of characteristics of respondents on the occurrence of the themes identified. In the current study, the use of these two approaches provided valuable insights into the opinions of potential Japanese consumers about plant‐derived cosmetics and the impact of demographic and socioeconomic characteristics of individuals on those opinions.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 17 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.