The General Ethnicity Questionnaire was adapted from four different pre-existing measures of acculturation: Cultural Life Styles Inventory, Acculturation Rating Scale for Mexican Americans, Suinn-Lew Asian Self-Identity Acculturation Scale, and Behavioral Acculturation Scale.

Mendoza's Cultural Life Styles Inventory

Items were constructed by: (1) asking a sample of Mexican American and Anglo-American professionals of different generations to generate lists of cultural items and practices that differentiate Mexican Americans and Anglo Americans, (2) creating a pilot inventory that included the most commonly-listed items from the above lists, (3) asking a second group of judges to rate the degree to which each item measured acculturation on a 5-point scale, and (4) empirically-testing those items that received the highest mean ratings.

The inventory from (4) was administered to 97 first generation Mexican Americans and 82 Anglo Americans (17-55 y.o.). Those items that discriminated between the two groups were kept in final inventory.

The final inventory was factor analyzed; five distinct factors emerged: (1) intrafamilial language use, (2) extrafamilial use, (3) social affiliation, (4) cultural familiarity, and (5) cultural identification and pride.

Internal consistency (Cronbach's alpha) was determined with a sample of 68 first-generation, 76 second-generation, and 41 third generation Mexican Americans (16-52 y.o.) and were as follows: (1) intrafamilial language use (.87), (2) extrafamilial use (.91), (3) social affiliation (.89), (4) cultural familiarity (.84), and (5) cultural identification and pride (.89). Test-retest reliability was determined by administering the inventory to a sample of 88 Mexican Americans and 59 Anglo Americans (16-49 y.o.), with an interval of two weeks. The reliabilities were as follows: for Mexican Americans who took the inventory in English (r=.91), for Mexican Americans who took the inventory in Spanish (r=.88), and for Anglo Americans (r=.95). Parallel forms were administered to two groups of bilingual Mexican Americans (16 to 36 y.o.), one that took the Spanish version first, and the other that completed the English version first. The correlations were as follows: r=.80 for the first group; r=.77 for the second group.

To establish validity, the investigators tested hypotheses regarding relationships between acculturation and generation level, acculturation and other measures of exposure to mainstream culture, and acculturation and temporary vs. permanent residence. In addition, reports made by relatives of subjects were obtained to test concordance. In each of these cases, the correlations were significant and in the hypothesized direction.

Cuellar's Acculturation Rating Scale for Mexican Americans

The scale was administered to 222 subjects (92 males, 129 females) comprised of hospitalized Mexican American, Spanish-speaking psychiatric patients with a psychotic diagnosis and students or staff participating in in-service training activities. The sample included subjects of different generation levels.

Internal reliability (coefficient alpha) was statistically significant: .88 for normal subjects, and .81 for hospitalized subjects. Test-retest reliability was measured on 16 psychotic patients (.72, p<.01) and on 26 staff members (.80, p<.01). Rater reliability was measured between two raters for 26 psychotic Mexican American patients (.89, p,.01).

Validity was assessed through staff ratings of the same patients. In addition, acculturation scores were found to differentiate Mexican, Mexican American, and Anglo subjects. Acculturation scores were found to be different from language fluency, highly correlated with generation level, and highly correlated with other measures of acculturation.

The factor analysis yielded four factors: (1) language familiarity, usage and preferences, (2) ethnic identity and generation, (3) reading, writing and general heritage and exposure, and (4) ethnic interaction.

Suinn-Lew Asian Self-Identity Acculturation Scale

The measure was administered to 82 Asian subjects (28 male, 54 female), who were students from Colorado State and UCLA.

Internal reliability (coefficient alpha) was .88.

To establish validity, acculturation scores were found to be highly correlated with generation level, length of residence, and answers to the question, "How would you rate yourself?"

Behavioral Acculturation Scale

Items were adapted from research and previous work on acculturation. The items were administered to two samples: (1) 265 Cuban Americans of diverse ages and SES (105 males, 150 females) and (2) 201 White individuals of diverse ages and SES. Those items that loaded highly on the factors yielded from a factor analysis and that discriminated between Cubans and Whites as well as discriminated between high and low acculturated Cubans were used in the Behavioral Acculturation Scale.

Internal reliability of the Behavioral Acculturation Scale was determined with a sample of 69 Cuban high school students and of 50 White American high school students (.97). The correlation between parallel (Spanish and English) forms of the scale was high (.88, p<.001) when administered to 27 bilingual subjects. The test-retest reliability was obtained by administering the scale four weeks apart to 30 subjects (.96, p<.001).

Validity was determined by testing three hypotheses related to acculturation differences in length of residence and differences in rates of acculturation by age and sex. Findings were in the hypothesized direction.