SURVEY RESEARCHgender, or ethnicity), use alternative interview meth-
ods (i.e., phone versus face to face), or accept alter-
native respondents in a household.Cooperation Rate.Cooperation among inner-city
residents, low-income persons, and racial-ethnic
minorities have increased as a result of using a
journalistic-style letter and a personal phone call
compared to using a standard academic letter.
Respondents who were pessimistic about govern-
ment and social service agencies and who felt mis-
understood were more likely to participate after
someone explained the nature of the survey to them
in terms to which they could easily relate.^47
As mentioned, small prepaid incentives increase
respondent cooperation in all types of surveys and
appear to have no negative effects on survey com-
position or future participation. For example, Brehm
(1994) found that without advance contact, 71 per-
cent of respondents cooperated, but the rate rose to
78 percent with advance contact (a letter) and an
incentive ($1) and the respondents were more talk-
ative. Moreover, respondents do not feel that differ-
ential payments for participation are unfair.^48
Instead of seeing respondents as already hav-
ing well-developed attitudes, beliefs, and opinions
that they are ready to retrieve and deliver when
asked in a survey, we see a survey as involving
several processes. The first is to win cooperation-
motivation so that people will participate fully in
the survey process. A second is assisting respon-
dents in correctly interpreting the survey question
and assembling an appropriate and accurate
response from memory or past experiences. A third
is helping respondents properly answer or deliver
the appropriate response (also see Example Box 1
earlier in this chapter).
Two related theories help explain the cooper-
ation-motivation process. Social exchange theory,
or the total design method (see Dillman, 1978,
2000), sees the formal survey as a special type of
social interaction. A respondent behaves based on
what he or she expects to receive in return for coop-
eration. To increase response rates and accuracy,
we need to minimize the burdens of cooperating by
making participation very easy and to maximize
rewards by providing benefits (i.e., feelings ofLocation Rate.Improving location means using
better sampling frames and maps or phone directo-
ries. Improving contact necessitates making many
repeat calls, varying the time of day for calls, and
lengthening the period to make contact. Several fac-
tors are associated with noncontact in the United
States: high population density, urban central city,
nonowner-occupied housing (i.e., rental), high
crime rate, high percentage of minority race popu-
lation, presence of physical barriers (i.e., fences,
bars on windows, beware of dog or no trespassing
signs), and a single adult living alone or households
without young children. Although they may be eas-
ier to locate and contact, people who have higher
income and more education may be less likely to
cooperate once contacted. As Groves and Couper
(1998:130) observed, “We find support in our data
for the notion that those in high SES [socio-
economic status] households cooperate less with
surveys than those in low SES groups.” Although
caller ID has increased, few respondents use caller
ID and telephone machine screening technologies
to block survey research in a significant way.^46
Contact Rate.A critical area of nonresponse or
refusal to participate occurs with the initial contact
between an interviewer and a respondent. Coopera-
tion increases when a respondent believes that
the survey topic or results will be salient to him or her
(i.e., are of great interest or will produce direct ben-
efits), or when interviewers use “tailoring” (discussed
later in this chapter) in their introductions to respon-
dents, or offer a small incentive (e.g., a few dollars).Eligibility Rate.We can improve eligibility rates
by creating careful respondent screening, using bet-
ter sample-frame definitions, and having multilin-
gual interviewers. We can decrease refusals by
sending letters in advance of an interview, offering
to reschedule interviews, using small incentives
(i.e., small gifts or amounts of money, as noted),
adjusting interviewer behavior and statements (i.e.,
making eye contact, expressing sincerity, explaining
the sampling or survey, emphasizing importance of
the interview, clarifying promises of confidential-
ity). We can also use alternative interviewers (i.e.,
different demographic characteristics, age, race,