MarketBasket & Other Statistical Surveys

Sample survey methodology is used to gather information about a "population" by selecting and measuring a "sample" from that population. The sample is a fraction of the population studied. The population could consist of sampling units, which could be people, animals, plants, foods, etc. The sample should be chosen on a statistically viable basis, such that each sampling unit has a measurable probability of being selected. Characteristics of the sampled items are measured and used to extrapolate to and represent the population. Standardized procedures are used to collect the information. Because sampling units have varying characteristics, scientific sample designs are used to reduce the risk of a distorted view of the population. The size of the sample depends on the purpose of the study, the variability among the sampling units, the desired precision, and the sample design. Sample designs include simple random sampling, which gives all units in the population an equal chance of being selected and randomly selects a subset of these units; stratified random sampling, which divides the population into homogeneous subgroups and then takes a simple random sample from each subgroup; systematic random sampling, which numbers the units in the population, selects a unit at random, and then systematically samples every nth unit; cluster random sampling, which divides the population into clusters, randomly selects a subset of these clusters, and measures all units in the selected clusters; and multistage sampling, which combines any of these four sampling methods.

How Can Exponent Help You With Survey Design, Conduct, and Analysis?

Market-Basket Survey: Design and conduct nationally or regionally representative surveys to collect samples or information from various sources along the production line. For example, Exponent scientists have designed several statistically based surveys to collect samples of fresh fruits and vegetables and processed commodities (e.g., milk, juices, potato chips) from supermarkets, and prepared foods from fast-food restaurants across the country. These surveys were designed for several purposes, including acute and chronic exposure assessment for pesticides and contaminants, analyses of nutrient content for food labeling, compliance assessment under California’s Proposition 65, and market research. Similar surveys can be designed for other non-food products. 

Food Consumption and Consumer "Habits" Survey: Design and conduct surveys to collect information on food consumption habits and consumer practice. For instance, Exponent has teamed up with a market research company to design and conduct a survey to collect data on water consumption patterns in the U.S., and has designed and conducted an Internet-based survey to collect information on the typical consumption pattern of a seasonal food. These surveys can be designed for several purposes, including providing risk assessors with refined data to estimate exposures and to provide serving-size data for nutrition labeling.

Analysis of Survey Data: Analysis of data from nationwide food-consumption surveys: Nationwide surveys such as USDA’s Continuing Survey of Food Intakes by Individuals (CSFII) and CDC’s National Health and Nutrition Examination Survey (NHANES) provide a wealth of information about food consumption patterns in the U.S. However, proper analysis of the data from these surveys requires the use of specialized methods to adjust for the complex sample designs. Exponent has the tools and experience to help you make sense of these data, and we routinely use these data in nutrition and contaminant risk assessments designed for regulatory submissions, market research, product development, and litigation support.