This study shows that a large, computerized survey instrument (the National Survey of America's Families) can serve as an excellent vehicle for identifying a target sample for further research (very poor families with children not currently working or receiving cash government assistance) and that qualitative information can be obtained effectively through telephone interviews. The interviews reveal that 56 percent of the families interviewed currently fit all study income criteria; 13 percent had a change in status since their NSAF interview; and 31 percent provided different information during the qualitative interview than they provided during the NSAF interview. The rate of survey discrepancies is actually lower than that found in previous studies that have attempted to verify income data in government surveys. The interviews suggest ways to improve the collection of income data in quantitative survey instruments.