30 COLLECTING AND ENTERING DATA
correct addresses are needed, response rates tend to be low, and problems of literacy
and language may exist.
Especially with mailed questionnaires, it is essential to make them easy to fill out.
One suggestion is to start out with the easiest questions so that the respondent will be
comfortable answering the questions. Some researchers will also put the questions
that they are most interested in early in the questionnaire so that if a responder does
not complete the questionnaire, they at least get the response to that question. Using
short questions with a specific choice of answers and a place to put the answers makes
it simpler for the responder. Sometimes answers such as “don’t know” or “does not
apply” are included to encourage getting some answer from the respondent. Only
medical terminology that is widely known or known to the respondent should be used.
Making plans for increasing the response rate are essential. Response rates can be
increased by having the reasons for needing the information made clear to the respon-
dent. If possible, some type of reward can be given for answering the questionnaire.
Note that in medical or public health studies, this does not have to be money. Plans
for conducting a second mailing should be considered. For example, a postcard could
first be sent out, reminding respondents to mail in their responses to the questionnaire
and why their responses are important. After another waiting period, the questionaire
could be sent out a second time.
Telepnone interviews are also less costly than face-to-face interviews. The in-
creased use of computers in computer-assisted-personnel-interviewing (CAPI) or
audio-computer-self-interviewing (ACASI), and use of the Internet have provided
additional options. Cell phone use has made some telephone interviews more diffi-
cult since cell phone numbers are not in local telephone directories or in older records.
Another disadvantage in telephone surveys is nonresponse or early hang-ups. Surveys
can also be done online (see SurveyMonkey.com).
In medical studies direct observation is often used to collect data. Here the major
advantage is that the reseacher obtains the information directly and does not have to
count on the memory or honesty of the respondent. The results can be entered directly
into a computer. The disadvantages are the expense and the fact that it can only be
used to collect data that can be observed directly. Also, if there are several observers,
they may have to train together so that their observations are done in a similar manner.
Use of medical records or other documents are common in medical studies. Med-
ical records, birth and death certificates, and coroner’s reports can be used. Their
advantages are that they are readily available data, are often less expensive, and have
few missing cases. Information on people who are no longer available can still be
found in records. The major disadvantages are that the records are often not collected
for research purposes, and the data may have been collected by persons with varying
levels of accuracy and completeness. It may also be difficult to access the desired
data.
3.1.3 Testing the Collection ProcessThe next step is to test the collection process. In collecting data from medical records,
the researchers should first examine a sample of the medical records they plan to use