INITIAL STEPS 29The cost of the new equipment for taking the measurements and recording the data
needs to be investigated.
3.1.2 Deciding How to Collect the Data
Data concerning patients or people in general can be collected directly using question-
aires, interviews, direct observation, medical devices, or indirectly from electronic
or written records (Bourque and Fielder [ 19991, and Bourque and Clark [ 19921). In-
formation that is unavailable in records or cannot be observed can often be obtained
from questionnaries. For example, information about past activities or conditions, at-
titudes or anticipated behavior can be collected from questionnaires. On the negative
side, the reliability and validity of this type of data are uncertain since they depend
on the respondents’ memories and willingness to be truthful. Also, questionnaires
are a good option only when the responders are available and willing to answer the
questions. Questionnaires can be administered face to face, by telephone, using the
Internet, mailed, or handed out. The data can be entered either by the respondent
or by the person asking the questions. In medical studies, when the questionnare is
administered by a medical person who is entering the responses from the patient as
part of the treatment for an illness, high numbers and accurate response rates can
usually be obtained. Subsequent questionnaires that are mailed out a year or more
after the treatment commonly have lower response rates and less certain accuracy.
There are major advantages in having the person asking the questions also record
the responses. More complex questions can be asked since the questioner can give
needed information to the respondent. Also, a more complex questionnaire design can
be used. For example, if parts of the questionnaire need to be answered only by certain
responders, such as patients with special risk factors, such as smoking or previous
medical conditions, the questions asked can be tailored to the answers obtained in
earlier questions. The downside to having a professional asking the questions is the
increased cost.
Questionnaires can also be administered in a group setting. Here the supervisory
person can give introductory information, answer questions, and monitor completion
of the questionnaire. This type of administration can be helpful in obtaining answers
to all the items being questioned. Sometimes this is done in a physician’s waiting
room by a receptionist, who distributes the questionnaire to the patients in the study.
Completion of the questionnaire can be checked by the receptionist.
Mail questionaires are the most commonly used that do not involve an interviewer.
Self-administered mail questionnaires have to be short, easy to fill out, and be of some
interest to the respondent. Sometimes a reward is offered for filling them out. Mailed
questionnaires are much less expensive to use. However, there is no certainty as to
who actually answered the questionnaire and their understanding of the questions.
Often, questionnaires are not filled out completely or returned.
Mailed questionnaires can be mailed virtually anywhere except to the homeless,
as long as an address is known. With lower costs, a larger sample can be taken and all
respondents will receive their questionnaires at the same time. If there is a seasonal
effect to what is under study, this might be important. The disadvantages are that