A diverse Chinese sample of 2567 high school and college students was utilized to examine the structural validity of the PGI-S (Tracey, 2010) with respect to the fit to the circumplex structure, the theoretical model underlying the RIASEC types and the eight PGI (Tracey, 2002) interest types. The randomization test of hypothesized order relations (Hubert & Arabie, 1987) was conducted separately for eight basic interest scales and the six RIASEC scales for both the high school and college samples and by gender. Results indicated strong fit for the eight type and RIASEC models in each examination. Concurrent validity of the RIASEC scores from the PGI-S was examined as they covaried with the Self-Directed Search (SDS, Holland, 1994), however the SDS demonstrated poor structural validity and thus was not an appropriate criterion. Mean differences in the eight type scores were examined as they varied across age (high school vs. college) and gender. Men demonstrated greater Thing scale scores but there were no differences on People scales. People scales did manifest age differences with college students having higher scores. Low prestige was higher for males and college students. Results are discussed with respect to interests in China.
While it is becoming increasingly common to translate and export U.S. interest measures to other countries, it is far less common to examine the validity of such applications. It is not tenable to assume that because interest measures have demonstrated validity in the U.S. that this will generalize to other countries. Thus, it is imperative to examine the validity of any U.S. interest measure in the foreign contexts where it may be used. The focus of the current study is on evaluating the validity of an adaptation of the Personal Globe Inventory-Short (PGI-S) in China.
The Personal Globe Inventory (Tracey, 2002) is a broad interest measure that incorporates many aspects of vocational assessment. It incorporates Holland's (1997) hexagonal model of RIASEC types but also expands these into eight circularly arranged types (Social Facilitating, Managing, Business Detail, Data Processing, Mechanical, Nature/Outdoors, Artistic, and Helping). The spatial arrangement of these basic interest types is presented in Fig. 1. Further, the PGI incorporates a separate interest dimension of prestige along two dimensions (People/Things, and Data/Ideas; Prediger & Vansickle, 1992) underlying Holland's RIASEC and the Personal Globe's eight types. Finally, the measure also includes separate scales to assess interests and competence perceptions. Tracey (2002) has demonstrated that the PGI has strong structural and concurrent validity in U.S. samples. The cross-cultural structural validity of this model has gained support in Japan (Long et al., 2006 and Tracey et al., 1997), Croatia (Šverko, 2008), Serbia (Hedrih, 2008), Ireland (Darcy, 2005), China (Long, Adams, & Tracey, 2005), Iran (Akbarzadeh, 2010), Singapore (Caulum, Tracey, Gresham, & McCarty, 2011), and the Caribbean (Wilkins & Tracey, 2012).
Full-size image (23 K)
Fig. 1.
Graphical representation of the PGI-SC 8 interest types with Holland's RIASEC types.
Figure options
Tracey (2010) utilized item response theory to create the PGI-Short (PGI-S), an abbreviated PGI measure that consisted of 40 items. Reducing the number of PGI items to 40 did not result in any change to the structure of interests in diverse U.S. samples. The PGI-S consists of 10 scales, including the same eight basic interest scales as the PGI and the high prestige and low prestige scales. This measure also provides data to calculate RIASEC scale scores for each respondent. However, this revision of the PGI has not been examined in any non-U.S. context and thus requires evaluation prior to any application. Long and Tracey (2006) demonstrated, in a meta-structural analysis across a number of samples and measures that the structural validity of RIASEC measures in China is moderate and not as high as that obtained in the U.S.
The focus of the present study was to examine the validity of the PGI-S in China using a broad sample of high school and college students. Specifically, the structural validity of the PGI-S scores was examined with respect to how well they fit the circumplex structure, which is the theoretical cornerstone of the RIASEC types and the eight Personal Globe interest types. Further, the concurrent validity of the RIASEC scores from the PGI-S was examined with respect to their covariation with the Self-Directed Search (SDS, Holland, 1994) scales. The SDS was selected as a comparison measure because of its prominence and that it has received moderate structural support when applied to Chinese contexts (Long & Tracey, 2006).
