Language of interview: relevance for research of southwest Hispanics

Am J Public Health. 1991 Nov;81(11):1399-404. doi: 10.2105/ajph.81.11.1399.


Background: This paper reports the results of a survey investigating health status, access, satisfaction with care, and barriers to care in Arizona. The major focus is on the association between language of interview and the dependent measures; interviews were conducted in English and Spanish.

Methods: The differences between groups were tested using chi-square statistics for each independent categorical variable; the significance of all the independent variables on each of the dependent variables was tested simultaneously using maximum likelihood logistical regression models.

Results: Language of interview for Hispanic children was a significant variable, more important than ethnicity itself, in determining health status, access, satisfaction with care, and barriers to care; language of interview for Hispanic adults was not a significant measure, but neither was ethnicity. Instead, income affected access to care for adults.

Conclusions: This pattern of results suggests that in the southwestern United States, studies on health status and access to care that use only ethnicity and do not include language of interview may fail to identify populations of Hispanic children who are remarkably more vulnerable. Public health research of Hispanic populations can be more instrumental toward policy improvement if it increases its specificity with this heterogeneous group. Analysis of language of interview has a low cost and a high benefit toward this specification.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Arizona
  • Child
  • Child, Preschool
  • Data Collection
  • Health Services Accessibility / statistics & numerical data*
  • Health Services Research / methods*
  • Health Status*
  • Hispanic Americans / statistics & numerical data*
  • Humans
  • Infant
  • Interviews as Topic / methods*
  • Language*
  • Logistic Models
  • Patient Satisfaction / statistics & numerical data
  • Refusal to Treat