Age, Income, Gender Phone Samples

Exact Age, Hard-coded There are two basic types of age-selectable samples. The first involves hard-coded data on specific ages within individual households. This approach offers the clear advantage of maximizing your dialing hit rate and minimizing wasted screening calls. Sources of hard-coded age data are diverse, drawing upon voter registration lists, drivers license records, survey data, county records, school registration data and others. Samples are systematically nth selected within your designated geographic area. Age Ranges, Predicted Because the availability of household-specific age data is uneven and many individuals are not recorded, representation is less than 100%. Therefore, we offer age range selections based on predictive techniques. With these, the researcher has a somewhat lower probability of reaching households containing a specified age, but will have greater confidence that the sample is more broadly representative. The predicted age algorithm involves the use of a multiple regression model. The entire consumer database of listed telephone households is analyzed and each household is assigned a score which indicates its probability of containing a specified age group. The multiple independent variables in the model include length of residence, geography, car value and a variety of census demographic variables. The scoring and age assignments are done at the lowest level of census geography, using 1990 census data at the tract and block group level. Samples are systematically nth selected within the your specified geographic area. Income, Net Worth, Income Producing Assets Using a sophisticated income prediction technique, ASI offers high quality income samples designed specifically to meet the needs of survey researchers. Income ranges may be selected in combination with other targeted selections, including age and our extensive list of low-incidence consumer categories. The income prediction technique is much the same as that which is used to predict age. A multiple regression model is evaluated against the consumer file of all listed households in the country. The model is driven by a variety of variables including length of residence, geography, car value and other census demographic variables. The end result is an income value, within $1,000 increments, assigned to each individual household. For reaching affluent households with low salary income - the retired, for example - ASI also offers Net Worth or Income Producing Assets selections. Gender For filling quotas of males and/or females, you can either 'take them as they fall' or sample systematically by gender. Often combined with an exact age selection, our gender samples allow you to reach the exact proportion of males or females that is called for in your sample design. Gender is determined by matching first names against a gender coded file of names, together with other sources of gender data. Other Demographically Targeted Groups Looking for some other demographic group? Call Affordable Samples, Inc. We can advise you on its availability and work with you to identify the best sampling alternative to meet your needs.

Low Incidence

Our Low-Incidence Consumer Samples provide an easy and cost-effective way for researchers to identify narrow consumer subgroups. We make it our business to bring to the survey research community a large selection of targeted consumer segments, and to be constantly seeking out new categories and data sources as the need arises among our customers. For example, you may have the need to interview: Internet Users Do-It-Yourselfers Occupational categories . . . and so on. Our low-incidence samples offer the advantages of targeting, proper sample selection, balanced geographic representation and rapid turnaround. Look though the list on the following pages and find the selection that matches your needs. If it is not included, give us a call and we will help you with an alternative sampling solution. What Incidence Rate Will You Get? Generally, you can expect to achieve an incidence rate on your selection criteria in the 40% - 60% range, though the actual percentage will depend on the specific selection and the survey particulars. Incidence rates will not be 100% due to the errors which typically are found in self-reported mail survey data and differences among respondents reporting within individual households. When the true incidence in the population is very low, a 40% - 60% achievable incidence rate represents a substantial improvement in dialing efficiency and cost savings. How Are Samples Created? Samples are drawn from a national database of household information gathered via mail questionnaires. The questionnaires are distributed in two ways: via cooperative mail packages and product registration cards. File compilation is continuous as questionnaires are completed and returned. All parts of the country are represented and the responses are distributed generally according to population densities. In order to maximize response quantities, several million questionnaires are distributed each wave. The great advantage of this approach is that resulting response quantities are high, thereby uncovering the very low-incidence groups researchers often need. Our low-incidence categories are based on actual survey responses provided by a household member and not on predictive techniques. Samples are additionally selectable by age, income, presence of children, age of children, gender and length of residence. Are These True Probability Samples? No. Like many mail surveys, they are self-selected samples of consumers who volunteered to respond to a detailed questionnaire which they receive by mail or to a product registration card. Often a response incentive is offered. However, because of the size, breadth and consistency of distribution, all parts of the country are well-represented. What about bias? Respondents included in these files skew somewhat older, more affluent and more female than general population. If desired, sample selections may be stratified by age, income or gender.