Research Concepts
Research Instruments
Your research instrument is how you record and measure the data that you collect.
Research instruments in the physical sciences can include things like thermometers, pressure guages, voltage meters and so on.
In social science research, your research instrument serves a similar role. Your instrument is the tool you use to collect and measure the data needed for your study. Sometimes you can use a pre-designed tool and sometimes you will need to design your own specialized tool.
In dealing with human subjects, your instruments will often consist of a list of questions to which the subjects respond. Other times the researcher might observe the behaviors of subjects in a particular situation, recording specific behaviors on a checklist or other form that facilitates and classifies the observations.
When your instrument is a questionaire the responses might take any number of forms:
- open-ended
- ordinal scales
- multiple choice
- fill-in-the-blank
Forms used to record observational studies often have similar options.
Open-ended responses are the most difficult to quantify and measure since there may be little consistency in the responses. The other kids of responses all involve some degree of pre-determination of what the responses might be and hence involve an inherent measurement. The danger with this pre-determination is that the researcher's expectations about what responses are possible might introduce bias.
An ordinal scale -- sometimes called a Likert scale -- asks the subject to rate their response along some continuum.Examples might be "on a scale from one to ten" or from "agree strongly to disagree strongly"or some other scale appropriate to what is being measured.
The ordering of the questions, how the subjects is exposed to the questions, the manner in which subject can respond to the questions, how the researcher interacts with the subjects before and after the questions are all part of the protocol for the research. Since all of these can influence the answers that the subjects give to the questions, the protocol can influence both the reliability and validity of the instrument.
