CHAMPAIGN, Ill. -- Every year, millions of Americans take vocational tests, almost all of them based on the Holland theory of personality types and vocational interests.
But what if the Holland-based map to the world of work was causing some people to get lost? What if certain groups of people just thought about the world of work in a different way? Other tests, after all, such as college-entrance and IQ exams, have come under fire as biased toward some groups.
No need to worry, says University of Illinois researcher Susan X Day, the author of a recent paper on the topic. When it comes to how Americans perceive different job-related activities and how they relate to one another, people apparently think remarkably alike, Day said. Race, ethnicity and gender don't appear to play a role.
Day, a doctoral student in the department of educational psychology, draws her conclusions from an analysis of the answers provided by 49,450 college-bound persons who took the ACT Interest Inventory in October 1989. Her sample, supplied on her request by ACT Inc., consisted of one-twelfth of the Caucasians and all of the African-Americans, Asian-Americans, Mexican-Americans and Native Americans who took the test at that time. For her analysis, Day also divided each of those groups by gender, to create a total of 10 groups.
Her paper on the topic, co-written by James Rounds, a U. of I. professor of educational psychology, was published in the July issue of the journal American Psychologist.
In presenting her findings at the National Career Development Association conference in July, Day noted "nobody was surprised, but everybody was very relieved it's been a concern, of course, for anyone who counsels people about careers, and it's been a concern for the people who develop and sell the tests."
Although most of those in the field of vocational testing had had confidence in the use of Holland-based tests across U.S. ethnic and racial groups, "our study is probably the first big empirical justification for thinking that," Day said.
Day said she was surprised by the "very remarkable" degree to which the 10 different groups matched up. The Holland theory lays out six occupational types in a circular order, and each type is represented in the ACT Interest Inventory by 15 related work activities, for a total of 90 items. Day used a statistical method called multidimensional scaling to plot on a graph the relationships among the 90 activities, derived from the 10 groups' responses. Not only did the activities array themselves in the expected Holland order, but all 10 groups' responses fit the same pattern with outstanding similarity.
Looking at the results, Day said: "It was just 'Eureka!' when I got the first printout. It was one of those really exciting research experiences."
[Editor's note: Ms. Day's middle initial is X, without a following period.]
Journal
American Psychologist