Dr D. K. Nyahunzvi Midlands State University 2012
If the research methodology is faulty The conclusions are invalid The methodology is therefore a critical aspect
A procedural plan that is adopted by the researcher to answer questions, validly, objectively and economically The plan has to be revealed and defended/justified The construct of ‘intellectual honesty’
Your study will most likely be a combination of designs Specify the dominant first e.g. an interpretive paradigm Progressively move into others e.g. case study
Based on number of contacts with study population Cross-sectional Aimed at finding prevalence of a particular phenomenon/ situation or problem/attitude Useful in obtaining an overall picture as it stands at time of study Cheap to undertake but cannot measure change
Measures change Suitable for measuring effectiveness of a programme Two sets of data need to be collected, more expensive and time-consuming Attrition-some who participated in the first phase may move out or withdraw for various reasons
Data is gathered over an extended period of time There is a need to collect information on a continuing basis Measures pattern of change
Demerits: conditioning effect Respondents may respond without thought or may lose interest Expensive to conduct in terms of time and money
Cohort study A cohort is a group of people with the same characteristic Cohort study is based on the existence of a common characteristic Selective sampling of the sample takes place unlike in panel studies
Panel Study collect data repeatedly from the same respondents Other designs Experiments? Quasi-experiments? The historical method? Ethnography?
Sampling strategy Study Population Sample selection, if census Instrumentation used Negotiating data access Pilot study
Detail the weaknesses and strengths Justify why they were used Merits and demerits must relate to your study Avoid over-quoting texts here
Avoid jargon e.g. Is anyone in your family a dipsomaniac? Avoid ambiguous questions: Is your work more difficult because you are expecting a baby? Yes/no
Are you satisfied with your canteen?” Avoid double-barreled questions Avoid leading questions Smoking is bad, isn't it?
Your questions should relate to your objectives Need a logical progression of questions Move from simple themes (general) to complex (sensitive)
Validity Validity refers to the appropriateness and accuracy of your data collection procedures The validity of the study design, sampling strategy, statistical procedures used and conclusions drawn Is the research providing answers for the research questions for which it was undertaken? If so, are the methods and procedures appropriate and correct
Face validity-is instrument measuring what it is supposed to measure? Compare questions and objectives, there must be a logical link Content validity-do the questions cover the full range of issues being measured? Predictive validity-how well can instrument forecast an outcome?
Do repeat measurements under constant conditions give the same result? Ability of an instrument to produce consistent measurements
Instead of validity Trustworthiness is determined by credibility Credibility is judged by the extent of respondent concordance
Transferability Thoroughly describing the context and methods So that others can follow and replicate Dependability- Do we obtain the same results if we observe the same thing twice?
Confirmability The degree to which the results could be confirmed/corroborated by others Researchers need to follow the same processes and procedures
Remember to be faithful to your chosen paradigm e.g. If you choose interpretivism, the need for representativeness and generalising falls away The paradigm affects all the stages (cascading influence) Learn from the methodology section of peer- reviewed papers