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.

66 ASKING QUESTIONS<br />

calendar year, even if they are unable to remember the details of<br />

how or why the money was spent or obtained. <strong>The</strong> best explanation<br />

of this is that they obtained summary information for another<br />

purpose, such as tax records, or because they budgeted a specified<br />

amount of money for a vacation.<br />

If summary information is likely to be available from records<br />

and is all that is required, you should use that information instead<br />

of taking data for a much shorter time period and calculating the<br />

yearly amount. Ordinarily, however, you will be interested in both<br />

the summary data and the details of individual events. In this case,<br />

both summary questions and detailed questions for a short time<br />

period should be asked. Comparing the summary results with those<br />

obtained from extrapolating the data from the shorter period allows<br />

you to check the reliability of responses.<br />

How to Minimize Telescoping. An appropriate time period is also<br />

important if you are to minimize backward telescoping, or remembering<br />

events as happening more recently than they did. Suppose<br />

that a national sample of households are asked to report the amount<br />

of coffee they purchased in the past seven days and that this total is<br />

then compared with shipments of all coffee manufacturers or observed<br />

sales in retail outlets. <strong>The</strong>se comparisons usually show that<br />

the amount reported is more than 50 percent higher than the<br />

amount manufactured and sold. What is happening is a process<br />

called telescoping.<br />

Telescoping results when the respondent remembers that the<br />

event occurred but forgets the exact date. In the past, most researchers<br />

were not concerned about telescoping because they<br />

believed that errors in the dates would be randomly distributed<br />

around the true date. However, recent research indicates that, as<br />

time passes, respondents are more uncertain about dates. As a<br />

result, respondents typically round their answers to conventional<br />

time periods, such as ten days ago, one month ago, or three months<br />

ago. <strong>The</strong> result of these two processes is to produce an overstatement<br />

of the reported events. Thus, an overstatement of coffee pur-

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

Saved successfully!

Ooh no, something went wrong!