Editing Holland Codes (section)

=== Tracey and Rounds's octagonal model ===
In the United States, the energetic trial is being made with the aim of the new model which surpasses Holland hexagon model in 1990's. Tracey & Rounds's octagonal model is one such example.<ref>{{Cite journal|last1=Tracey|first1=Terence J. G.|last2=Rounds|first2=James|date=1995|title=The arbitrary nature of Holland's RIASEC types: A concentric-circles structure.|journal=Journal of Counseling Psychology|volume=42|issue=4|pages=431–439|doi=10.1037/0022-0167.42.4.431|issn=0022-0167}}</ref> Based on the empirical data, they argue that occupational interests can be placed circularly in a two-dimensional plane consisting of People/Things and Data/ldeas axes, and the number of regions can be arbitrarily determined. According to their model, only Holland's hexagonal model does not adequately represent the structure of occupational interest, and it is possible to retain validity as an octagonal or 16 square model if necessary.

Tracey, Watanabe, & Schneider conducted an international comparative study of job interests among Japanese and U.S. university students, and the results suggest that the Tracey & Rounds's octagonal model is more fitted to Japanese students than Holland's hexagonal model.