Cluster sampling techniques are used when a complete sampling frame for a population is unavailable. It is typically a two-staged process. Initially, the total population is divided into clusters or groups, and then a random sample of clusters is selected. In the second stage, a random sample is selected from within each of these clusters. A common example of this approach is geographical sampling where clusters are based on geographical areas (e.g., neighborhoods). For example, assume that the development of an intervention that is targeting teenagers with substance abuse problems involves conducting interviews with social workers in New York City, the targeted area. The interviews are designed to collect initial input as to the need for, and format/content of, the intervention. However, a complete list of social workers in the targeted area may not be available. Thus, one strategy would be to select agencies in the area that are likely to employ social workers. In addition, the city is large and the agencies are dispersed across neighborhoods (e.g., SoHo, Upper East Side, Upper West Side). Using two-stage clustering techniques would entail initially selecting a random sample of neighborhoods (large clusters) and then a smaller random sample of agencies (small clusters) within these select neighborhoods to include in the sample. One would then recruit social workers from these clusters. An advantage of cluster sampling is that it can be more economical and reduce costs such as, in this case, travel.