Test Review: The Intercultural Development Inventory manual

Introduction

The Intercultural Development Inventory v.3 (IDI; Hammer, 2011) was constructed to measure individuals’ orientations toward cultural differences. The theoretical framework for this instrument was Bennett’s (1993) developmental model of intercultural sensitivity (DMIS). The DMIS identifies three ethnocentric orientations (i.e., denial, defense, minimization) and three ethnorelative orientations (i.e., acceptance, adaptation, integration; Bennett, 1993).

Ethnocentric orientation is defined as an attitude that regards one’s own culture as the center of the world. The first stage of ethnocentric thought is denial, in which people who are isolated from other cultures consciously ignore those other cultures’ values. Defense is the next stage of ethnocentrism, in which one begins to acknowledge the existence of other cultures but believes that one’s own culture is superior to other cultures. The last stage of ethnocentric orientations is minimization. People in the minimization stage focus only on personal similarities (e.g., physical, biological, philosophical), but do not legitimize other societies’ broad cultural frameworks.

The second HALF of the DMIS continuum includes three stages of ethnorelative orientations, each of which recognizes to some extent the influence of other cultures on one’s own culture. The first ethnorelative orientation is acceptance in which people accept the existence of various cultures in the world and respect their values. The second stage is adaptation, in which people adjust their attitudes and behaviors to accommodate and exist within a specific cultural context. According to Bennett (1993), to achieve this latter level of ethnorelative orientation, pure exposure to foreign culture is insufficient. More critical to adaptation, people must consciously construct their cultural experiences and views to adapt fully. The last stage of ethnorelative orientation is integration, in which one is comfortable switching identities within one culture versus another. Bennett (1993) emphasized that integration is not necessarily a better orientation than adaptation; it is simply an orientation that some people assume after many years of experience in or interacting with another culture, or when a person may have inherited and been raised within multiple cultures.

The IDI was based on the DMIS model and is purported to be appropriate for any cultural group (Hammer, 2011). As such, the IDI includes seven subscales, each measuring one of the six dimensions of the DMIS, plus an additional scale (i.e., Cultural Disengagement [CD]). Hammer (2011) pointed out that CD is not a dimension along the DMIS orientation continuum; rather it is a separate dimension that measures how people relate to their own cultural group and other cultures.

The entire IDI is composed of 50 items, with10 additional demographic items. The scale is self-report and can be administered either in paper and pencil format or in an online administration in 15 languages (i.e., Arabic, Bahasa Indonesian, Chinese, Czech, English, Finnish, French, German, Italian, Japanese, Korean, Norwegian, Portuguese, Russian, and Spanish). The IDI is appropriate for individuals who are ages 15 or older and who are reading at least at a 10th-grade level (Hammer, 2011). Also, the instrument can be administered in either individual or group settings.

One significant drawback of the instrument is that to use the IDI, examiners are required to attend an IDI Qualifying Seminar and consent to a licensing agreement. The training seminar requires a 2.5-day commitment, and as of 2013, the training costs US$1,800. Moreover, participants must complete the entire seminar to receive their IDI Qualified Administrator Certification.

Purpose

According to the IDI official website, the instrument can be used for a wide variety of purposes, including individual assessment, group analysis, and organization-wide needs assessment, program evaluation, and research. The IDI has been used for assessing intercultural sensitivity in short-term (Anderson, Lawton, Rexeisen, & Hubbard, 2006) and year-long (Pedersen, 2010) study abroad programs. The IDI was also used in assessing the long-term impact of study abroad programs on intercultural development in longitudinal studies (Hansel, 2009; Meyer-Lee & Evans, 2006; Rexeisen, Anderson, Lawton, & Hubbard, 2008).

Scoring

Administration of the IDI generates a diagnostic profile, which represents the following six cultural orientations: (a) Perceived Orientation (PO), computed using an unweighted formula, reflects one’s perception of intercultural development; (b) Developmental Orientation (DO), using a weighted formula, that is defined by the IDI assessment; (c) Orientation Gap (OG) is the difference between clients’ perception and the assessment, and is calculated by subtracting PO from DO; (d) Trailing Orientation (TO) presents a score that is intended to reflect unresolved orientations from lower DMIS stages, and is determined by a score less than 4 in each subscale (e.g., a score of 3.2 in the subscale of Minimization indicates a TO); (e) Leading Orientation (LO) is the stage above the examinee’s DO (e.g., Adaptation is the LO to Acceptance); and (f) CD, which addresses how disconnected examinees feel toward their own respective cultural group; a score less than 4.00 indicates an unresolved feeling toward one’s own culture (Hammer, 2011). PO and DO scores are represented as standardized IQ metric scores (i.e., M = 100, SD = 15), ranging from 55 to 145. A score below 85 demonstrates a cultural orientation in Polarization (i.e., defense and reversal), scores between 85 and 114.99 indicate an orientation in Minimization, and scores of 115 and above represent degrees of Acceptance and Adaptation (Hammer, 2009b; Sample, 2009). The gap between PO and DO is the OG score, and “a gap score of 7 points or higher can be considered a meaningful difference between where you perceive you are on the developmental continuum and where the IDI places your level of intercultural competence” (Hammer, 2009b, p. 6). TO scores range from 1 to 5: any scores below 4 suggest an unresolved TO. Scores of 4 and higher suggest that a person has resolved that particular TO (Sample, 2009).

Development

The development of the IDI consisted of three phases. During the beginning phase (226 respondents), the researchers developed an initial (60-item) version of the IDI (Hammer, 1999). The items represented all six DMIS orientations. In Phase 2 (591 respondents), the researchers conducted a confirmatory factor analysis, eliminated 10 items and determined the final 50-item IDI, with its five dimensions (Hammer, Bennett, & Wiseman, 2003). In addition, in 2007, a sample of 4,763 respondents from a wide range of ages and professions completed the IDI in their native languages (six languages were used). The translation process included a rigorous back-translated methodology. The participants came from the profit sector, international organizations, nonprofit organizations, and high schools and colleges (Hammer, 2011). Phase 3 in 2007 yielded a new IDI version (v.3) with seven subscales (Hammer, 2009a). The authors also created two composite measures: PO (α = .82) and DO (α = .83; Hammer, 2009b).

Reliability

The IDI researchers provided detailed information on scale reliability estimates during each stage of the instrument’s development, which is described below.

Validity

There are various forms of validity associated with an instrument such as the IDI. Among them, face validity, content-related validity, and internal validity can be determined through self- and expert review, pilot studies, and data analysis (Cohen & Swerdlik, 2010; Gall, Gall, & Borg, 2007); additional criterion-related validation is needed to demonstrate the predictive value of the construct assessed.

Test Bias

The researchers also addressed the possible effects of gender, age, and education on IDI scale scores. No significant differences were found between groups, except the male mean Denial/Defense score was significantly higher than female mean (t = 4.84, df = 553, p < .01). As there was no significant difference on other scales, the researchers concluded that the IDI scores in general were not significantly influenced by gender differences.

To investigate test bias, Hammer et al. (2003) and Paige et al. (2003) tested for social desirability bias for the IDI using the (10- or 20-item) Marlowe–Crowne Social Desirability Scale short form (Strahan & Gerbasi, 1972). Neither group produced significant correlations between the Social Desirability scale and the IDI subscales, suggesting that the IDI scale scores were not influenced by a tendency for participants to provide socially desirable responses.