|
4.2.2 Data collection techniques
Besides the research design, research can be classified according to the technique used in
the study to collect data (McMillan & Schumacher 1993:40). There are six ways to
collect data. There are questionnaires, interviews, documents, tests and unobtrusive
measures. All research uses a variation of one or more of these, depending on their
applicability to the research being undertaken. Like research designs, the techniques can
be classified as either quantitative or qualitative. As McMillan and Schumacher (p40)
point out, the fundamental difference is that quantitative approaches use numbers to
describe phenomena, while qualitative data collection techniques use narrative
descriptions.
While most of the techniques can be used with any of the research designs, research
design is closely related to technique. The quantitative technique emphasises a priori
categories to collect data in the form of numbers. The goal is to collect data to provide
statistical descriptions, relationships and explanations. Quantitative techniques
(McMillan & Schumacher 1993: 41) are used with experimental, descriptive and
correlational designs as a way to summarise a large number of observations and to
indicate numerically the amount of error in collecting and reporting data. Quantitative
research uses different types of data collection techniques such as structured
observations; standardised interviews; tests; questionnaires and unobtrusive measures.
101
The questionnaire and interview of the study under discussion – the Video Conference
Focus group interview (VConf-FGI) - are discussed in the sections which follow. The
chief data collection methods are quantitative and qualitative. They serve to complement
rather than rival each other in the study. Flick (2002:265-266) distinguishes two
alternative ways of concretising the use of the two methods discussed above. The one is
to focus the single case where the same people are interviewed and fill in a questionnaire.
Their answers in both are compared with each other, put together and referred to each
other in the analysis. Sampling decisions are taken in two steps. The same people are
included in both parts of the study, but in a second step it has to be decided which
participants of the survey study are selected for the interviews. The other alternative is to
establish the link between quantitative and qualitative research on the level of the data set
(p226). The answers to the questionnaires are analysed for their frequency and
distribution across the whole sample. The answers to the interviews are also analysed and
compared and a typology is developed. The questionnaire is then the distributed and
answers and typology are linked and compared.
4.2.2.1 The questionnaire
Questionnaires (McMillan & Schumacher1993:42) encompass a variety of instruments in
which the subject responds to written questions to obtain reactions, beliefs and attitudes.
The researcher chooses or constructs a set of appropriate questions and asks the subjects
to answer them, usually, in a form that requests the subject to check the response. Most
survey research uses questionnaires because of their advantages over other data collection
methods. A questionnaire (McMillan & Schumacher 1993:238) is relatively economical,
has standardised questions, can ensure anonymity, and can be written for specific
purposes. Questionnaires can use statements or questions, but in all cases the subject is
responding to something written.
The first step when developing new questionnaires is to consider justification. McMillan
and Schumacher (1993:239) advocate the use or adaptation of an existing instrument
instead of preparing a new one. Doing so would save time and money and may ensure
that an instrument with established reliability and validity is used. Defining and listing
102
specific objectives is the second step in using a questionnaire. The objectives are
formulated according to the research problems/questions and they show how each piece
of information will be used. The objectives need not be strict behavioural objectives but
they must be specific enough to indicate how the responses from each item will meet
them. Babbie (1989), in McMillan and Schumacher (1993:240), suggests guidelines for
writing effective questions and statements. These guidelines are:
•
items should be written clearly and ambiguous words, jargon and complex
phrases should be avoided
•
questions should be limited to a single idea or concept, therefore double barrelled
questions, that is, questions that contain two or more ideas, should be avoided
•
respondents must be competent to answer, that is, the respondents should be able
to provide reliable information.
•
questions should be relevant; in other words, respondents should b asked to
answer questions that are important to them and to respond to things about which
they care and think
•
items in the questionnaire should be written in a simple and easy to understand
format
•
negatively stated items should be avoided as should biased items or terms.
Neuman (1997:251& 2000:271-272) enumerates the following advantages of mail and
self-administered questionnaires. This type of survey is cheap and can be conducted by a
single researcher. Questionnaires can be sent to a wide geographical area and the
respondent can complete the questionnaire when it is convenient. Anonymity is provided
by mail questionnaires and interviewer bias is avoided. The disadvantages are that the
response rate is often low and there is lack of control over the conditions under which the
questionnaire is completed. Moreover, some respondents may give incomplete answers.
It is also not possible to observe the respondent’s reactions to questions, physical
characteristics or the setting. In addition the kinds of questions a researcher can use are
limited.
103
Do'stlaringiz bilan baham: |
