An Investigation of the Criterion and Concurrent Validity of the Inventory of Dimensions of Emerging Adulthood (IDEA)

Background and Purpose: This study investigates the concurrent and criterion validity of the Inventory of Dimensions of Emerging Adulthood (IDEA). Although this measure has been used with college students, it has yet to be validated with lower income emerging adults (ages 18-25). This research gap is important to address, as some have questioned the application of Arnett's theory to the emerging adults with whom social workers are most likely to interact. Furthermore, Arnett's (2005) hypotheses about how developmental factors may influence substance use during emerging adulthood (EA) remain largely untested. However, past studies using other measures have shown that identity formation is associated with substance use. Thus, the purpose of this study is to examine the associations between self-reported substance use and five proposed dimensions of EA measured by the IDEA(i.e., criterion validity), as well as the associations between these dimensions and ego identity status (i.e., concurrent validity).

Methods: Racially diverse, low income, substance-abusing young adults ages 18-25 (n=105) were recruited from a private nonprofit substance abuse treatment center. Participants completed the Global Appraisal of Individual Needs (GAIN; Dennis, Titus, White, Unsicker, & Hodgkins, 2002), the Inventory of Dimensions of Emerging Adulthood (IDEA; Reifman, Arnett, & Colwell, 2007) and the Extended Version of Objective Measure of Ego Identity Status (EOM_EIS; Adams, 1998). We examined the correlations between the Substance Frequency Scale (SFS) and the Substance Problem Scale (SPS) of the GAIN, the five subscales of the IDEA, and the subscales of EOM-EIS.

Results: Results: We found significant positive correlations between the IDEA's insecurity/negativity subscale and both the substance use frequency (r = .25, p<.05) and substance problem scales (r = .21, p<.05). Most correlations between IDEA subscales and substance abuse were in the expected direction, but were modest in magnitude and not statistically significant. Thus, there was little evidence for Arnett's (2005) criterion validity hypotheses in this study. We replicated past findings that EOM-EIS subscales correlate with substance abuse. Specifically, significant positive correlations were found between substance use frequency and the tendency to hold beliefs in line with their parents (r = .26, p<.05) and between problems from use and holding vague beliefs due to being in identity crisis (r = .22, p<.05). Some support was found for the concurrent validity of the IDEA. Negative correlations were found between the IDEA's Feeling-in-Between subscale and both the EOM-EIS's openness to changing identity (r = -.33, p01) and identity achievement scales (r = -.27, p<.05).

Conclusions and implications: This study was the first study to test the criterion validity of the IDEA, and the general lack of support for Arnett's (2005) hypotheses implies that unobserved variables may better account for differences in substance use severity among lower income substance-abusing young adults. Although more studies are needed, such findings may also indicate the presence of a middle class bias in the theory of emerging adulthood (Hendry and Kloep, 2007).