Date

2013-11-28

Author

Toker, Yonca

Metadata

Show full item record

Item Usage Stats

198
views

0
downloads

Cite This

Aim of the present investigation was to delineate gender differences on one specific domain (i.e. symbolic interests) of the vocational interest measure focusing on interests toward the Science, Technology, Engineering, and Math (STEM) areas; STEM Interest Complexity Measure (Toker & Ackerman, 2012). The measure was developed to incorporate STEM-related tasks that vary in their complexity levels to represent a more comprehensive assessment of interests that pertain to Holland’s Realistic and Investigative themes along the span of occupational complexity levels (Gottfredson, 1986). The measure covers tasks dealing with numeric, symbolic, spatial information, and STEM-related ideas. Each of these domains include items that cover the “data” complexity levels identified in the Dictionary of Occupational Titles; which are copying, comparing, computing, compiling, analyzing, synthesizing, and generating in order of ascending complexity. Furthermore, items were developed with a focus on the work activities outlined by the Occupational Information Network (O*NET) for STEM-related jobs. The 121-item measure was shown to have good construct and criterion-related validity (Toker & Ackerman, 2012). In the prediction of vocational criteria (i.e. STEM major satisfaction, intentions to further pursue the field) the relative contribution of measures revealed that the relevant Holland themes explained from 0% to 3%; math-and-science self-concept explained from 1% to 7%; and STEM Interest Complexity explained from 5% and 12% of the variance in criteria. STEM Interest Complexity added 5% variance on the above-mentioned interest measures in the prediction of STEM-GPA. Gender differences have been documented for Holland’s (1997) interest themes and STEM-subject interests with a recent meta-analysis (Su, Rounds, & Armstrong, 2009). Effect sizes of observed differences were large for realistic and engineering interests, and small for investigative, science, and math interests, with men reporting higher preferences than women. I hypothesized that men would have higher interests on the STEM Interest Complexity-Symbolic domain. The sample was 273 undergraduate students in an American college (126 men and 147 women). Latent Mean Differences was tested using EQS6.1 for the moderate- and high-complexity symbolic-interest factors, after omitting the items with differential slope and thresholds across the genders based on Differential Item 16 104 Functioning. As a result 21 items, 14 as indicators of the moderate-complexity latent factor and 7 of the highcomplexity latent symbolic-interests factor, were included in the analysis. The women sample was designated the reference group and latent factor loadings were set to zero. For the sample of men, latent factor loadings were freely estimated to yield this group’s latent factor means relative to the reference group. According to the results (S-Bc2(329) = 468.41, p < .001, CFI = .97, RMSEA = .058, CI .046, .069, MACS RMSEA = .056, CI. .044, .067), men reported higher preferences on symbolic-interests than did women. Latent mean differences were significant for moderate-complexity symbolic-interests (B = .643, t = 4.89, p < .05) and also for highcomplexity symbolic-interests (B = .734, t = 6.38, p < .05). Results are discussed in relation to the desire of increasing the representation of women in STEM-areas and related vocational counseling and intereststimulating activities.

Subject Keywords

Vocational interests,, Interest complexity,, Symbolic interests,, Gender differences

URI

https://hdl.handle.net/11511/88332

Collections

Department of Psychology, Conference / Seminar