Sample design is a key factor of survey quality since it directly affects the magnitude of sampling error. In other words, how a sample is designed determines how reliable survey results will be. At Sistemas Integrales we strive for efficient sample designs that maximize the precision of survey estimates considering both budget constraints and operational feasibility in real field conditions.

We apply only rigorous probability sampling methods, which entails strict random selection in all sampling stages and control over inclusion probabilities. This principle applies to household, business and agricultural surveys and to all modes: face-to-face, phone and web surveys.

In household surveys, GIS tools can also help field workers locate the sample households when standard addresses do not exist. This is done by using the GPS coordinates recorded during the listing operation.

Sampling for phone surveys is based on Random Digit Dialing (RDD), respondent data from a past face-to-face survey that collected phone numbers, or a combination of both.

Rare and special populations require specific sampling strategies based on screening, multiple frames, or network and adaptive sampling. At Sistemas Integrales we have helped our clients survey special populations such as:

• Refugees and immigrants
• Ethnic and linguistic minorities
• Victims of violence
• Disabled persons