LANYCIS is a rich source of information about immigrants, particularly low-income immigrants living in urban areas. More detailed information about the survey is contained in Appendix 1 of this report. This section describes some of the defining features of the new survey.
- LANYCIS has a large, statistically representative sample of 3,447 immigrant families (families with one or more foreign-born adult) in the two areas (1,893 in Los Angeles County and 1,554 in New York City). The sample includes detailed data about 7,843 "focal people" who were living in those immigrant families.
- Interviewing was conducted between August 1999 and July 2000, with most of it done between November 1999 and May 2000. The Survey Research Center of the University of California at Los Angeles administered the survey in both cities.
- The survey has a large sample of low-income immigrant families, and it oversamples those who were receiving or had received food stamps, as well as those with incomes below 200 percent of the poverty line. All analyses presented here are weighted to correct for sampling and to align with Census data. The overall response rate (the estimated share of targeted people who answered the survey) was 69 percent, which is high when compared to other random digit dialing (RDD) surveys, suggesting a strong response to the survey.(8)
- The survey collected a broad array of economic, employment, demographic, health, and program participation data about immigrant families.
- Unlike most surveys, LANYCIS collected self-reported information about immigrants' legal status, such as whether a person is a lawful permanent resident, refugee, naturalized citizen, or undocumented alien.
- Data were collected through both telephone and in-person interviews, using computer-assisted technologies.
- Interviews were conducted in five languages, including English, Spanish, Chinese (Mandarin), Russian, and Vietnamese. The majority of interviews were done in Spanish.(9)
Although LANYCIS did not interview native citizen families, in many cases we can compare the survey's immigrant families to native citizen families in California or New York, drawn from the Urban Institute's 1999 National Survey of America's Families (NSAF II) (Wigton and Weil 2000; Capps 2001). This was a survey of 42,000 households drawn to be statistically representative of nonelderly families in 13 states, including New York and California, and the rest of the nation. NSAF II was conducted from February to October 1999. Since LANYCIS and NSAF have comparable time frames and share many questions, we are able to compare LANYCIS data about immigrant families in Los Angeles County and New York City with NSAF II data for the native-born in California and New York State.
NSAF II was designed to represent states, not individual communities. While we would have preferred to compare LANYCIS data with NSAF II data about native citizens from Los Angeles County and New York City only, the NSAF data are not necessarily representative of those two areas. Based on the advice from the researchers who developed NSAF, we used state-level comparison groups in this report. Furthermore, since NSAF II did not sample elderly families, there were no good comparison groups for the elderly immigrants in LANYCIS. We acknowledge that the comparison groups are imperfect, and warn readers against drawing strong causal inferences from these comparisons; they should be viewed as simple, descriptive comparisons.
Both LANYCIS and NSAF II were designed as carefully as possible; nevertheless, there may be errors or omissions because respondents misunderstand the questions, do not know the answers, have imperfect memories, or do not want to reveal secrets. The risks of misreporting might be higher for immigrants than for the general population because of language problems or cultural misunderstanding. Moreover, a small share of the interviews were conducted face-to-face, while the bulk were done over the telephone. With sensitive issues such as legal status documentation discussed, the mode of the interview may have influenced responses. We used a variety of methods to reduce survey problems, such as pretesting the survey; having the survey conducted by a well-known, trusted organization (the University of California at Los Angeles); administering the survey in several languages; using computer-assisted interview techniques; and using multiethnic interviewers. We also conducted in-person, follow-back interviews with a sample (200) of families to both amplify and verify telephone survey responses. But we acknowledge the potential for reporting errors in this survey, as in any other.