24.10.2014 Views

Asking Questions - The Definitive Guide To Questionnaire Design ...

Asking Questions - The Definitive Guide To Questionnaire Design ...

Asking Questions - The Definitive Guide To Questionnaire Design ...

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

114 ASKING QUESTIONS<br />

carefully for such measures before concluding that no validation is<br />

possible. If the behavior involves a product or service, you can<br />

compare consumer reports with those of manufacturers, retailers,<br />

or suppliers of the service, as with purchases of beer, wine, and<br />

liquor. For socially desirable behavior such as giving to charity, you<br />

can compare the amounts reported with the total amounts<br />

received; reported play and concert attendance can be compared<br />

with figures on total tickets sold. But be careful to avoid comparing<br />

apples and oranges. In many cases there will be a nonhousehold<br />

component in the validation data. Thus, business firms also<br />

contribute to charity and purchase goods from retailers. On the<br />

other hand, validation data may be useful, even if the comparisons<br />

are not perfect.<br />

Summary<br />

Threatening behavior questions are intrinsically more difficult to<br />

ask than are nonthreatening questions. As the questions become<br />

very threatening, substantial response biases should be expected,<br />

regardless of the survey techniques or question wordings used. For<br />

less threatening questions, carefully designed question formats and<br />

wording can substantially improve response accuracy.<br />

<strong>The</strong> procedures suggested in this chapter for obtaining more<br />

accurate reports of threatening topics include: (1) using selfadministered<br />

methods to increase perceptions of confidentiality,<br />

(2) using card sorting and randomized response, (3) using openended<br />

questions, (4) using long questions with familiar words,<br />

(5) using informants, (6) using diaries and panels, (7) embedding<br />

the question, (8) choosing the appropriate time frame, (9) making<br />

the questions less threatening. For socially desirable behavior,<br />

ask respondents about their most recent behavior rather than about<br />

their usual behavior. For socially undesirable behavior, ask respondents<br />

whether they have ever engaged in a particular behavior<br />

before asking about their current behavior.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!