Description
"To consult the statistician after an experiment is finished, is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of." This quote is from the famous Statistician, Ronald Fisher. In order to avoid using a questionnaire or research instrument that is not valid or reliable and therefore not be able to do the statistical analyses, this course is developed. The lessons in this course is developed from 23 years of experience as a consulting Statistician at various Universities in South Africa. Please note that if you need a more comprehensive course on quantitative research design consider my other course on quantitative research design for a proposal, this is a short course focussing mostly on questionnaire validation and reliability. Testing perceptions with Likert scale statements in a questionnaire must be arranged as constructs(concepts) also called dimensions or factors in order to test some theory. The constructs of a scale that is used in any questionnaire for quantitative research, must be valid and reliable, before scores can be calculated which would be used as variables in the statistical analyses. These variables will then be used when answering research questions or testing hypotheses. For example, the Technology Acceptance Model(TAM) used in the Information science field, consists of the validated constructs: ease of use, usefulness, attitude and intention. These constructs forms the Technology Acceptance scale that is used in a questionnaire. For each of these constructs(ease of use, usefulness, attitude and intention) a few Likert scale statements will be used. This questionnaire with the TAM scale will then be distributed and the results analysed by using the TAM model to test the acceptance of new technology. This course will teach the participant to determine valid and reliable constructs to use in their own questionnaire. The following topics are discussed in this course: General questionnaire design and tipsThe use of Exploratory Factor Analysis for construct validation. The use of the Cronbach Alpha coefficient to test reliability of constructs in the questionnaire. How to choose a validated questionnaire when conducting the literature study. The pre-testing of the questionnaire. A practical example from a PhD.
