Nonresidential

Nonresidential dispositions occur in telephone and in-person surveys of the general public when a case contacted or called by an interviewer turns out to be [Page 527]a business or other type of nonresidential location such as a hospital, government office, or library. The nonresidential disposition also usually includes institutions such as prisons, sanitariums, work camps, and group quarters such as military barracks and college dormitories. Nonresidential cases are considered ineligible for a survey of the general public survey because conducting an interview with these cases would violate critical assumptions of probability sampling. Although the proportion of nonresidential cases in a telephone or in-person sample varies based on the sampling area, the nonresidential survey disposition tends to be fairly common, with nonresidential numbers comprising up to a significant minority of the telephone numbers in a landline telephone random-digit dial (RDD) sample. More recent technologies that are capable of screening out numbers in an RDD sample without interviewers calling the number have substantially reduced the proportion of nonresidential numbers in many telephone survey samples.

One special challenge in telephone surveys is call forwarding technology, which allows one telephone number to be transferred to another number. For example, if a residential telephone number is transferred to another telephone number within the same household, the case may still remain eligible if the survey uses weighting techniques to adjust for unequal probabilities of selection (since the residential unit contacted effectively has an additional telephone line, and thus, a greater chance of being sampled). If a non-residential telephone number is forwarded to a residence (e.g. a business number being forwarded to the owner's home residence), the case should be considered ineligible. More detailed rules may be needed for special (but uncommon) cases in which the number for a residence outside the sampling area is forwarded to another residence inside the sampling area. Finally, a residence and a business occasionally may share the same telephone number; these cases should be treated as eligible for the interview. Only numbers that ring exclusively to a business or other nonresidential unit should be given the nonresidential disposition.

For in-person surveys, two special cases can make it more difficult to determine whether a case is truly nonresidential. For example, although the primary unit at an address might be a business or an institution, it is important to ensure that there is not a residential housing unit within the larger unit (such as an apartment above a business or a warden's house on the grounds of a prison). An additional challenge is posed by vacation and other seasonal homes, and special rules may need to be developed to properly determine the disposition of these housing units.

The challenges and special cases mentioned illustrate that at times it may be difficult to determine whether a telephone number or housing unit is residential. Although these instances are fairly uncommon, it important to ensure that there is definitive evidence that the case is nonresidential before applying the nonresidential disposition to a case. Obtaining this evidence may require additional investigation (such as talking to neighbors or documenting visible signs that the unit is uninhabited or not used as a residence). In surveys of the general public, a nonresidential outcome is treated as final and ineligible and thus is not used in response rate calculations.

...