4. lecture 4 research design
Download as PPT, PDF0 likes130 views
The document discusses research design and provides definitions and examples of different types of research designs. It covers the key elements of research design including purpose, type of study, measurement, unit of analysis, and time horizon. It also distinguishes between quantitative, qualitative, and mixed methods approaches. Specific research designs like experiments, surveys, case studies, and action research are described in terms of their focus, characteristics, and strengths/limitations.
![Research designs
Definition:
• A research design provides the framework for
the collection & analysis of data.
• It is the general plan of how you [the researcher]
intend to conduct your research to answer your
research questions and in a way that ensures
research quality.](https://image.slidesharecdn.com/4-210628142318/85/4-lecture-4-research-design-2-320.jpg)
4. lecture 4 research design
- 1. Lecture 4 – Research Design
- 2. Research designs Definition: • A research design provides the framework for the collection & analysis of data. • It is the general plan of how you [the researcher] intend to conduct your research to answer your research questions and in a way that ensures research quality.
- 3. Research designs The research design reflects the decisions around the following elements: a) Purpose of the study (exploration, descriptive, explanatory) b) Type of study (causal or correlational study; observational study or case study) c) Researcher involvement (minimal or control through manipulation/simulation)
- 4. Research designs d) Study setting (natural / contrived) e) Measurement (operational definitions, measures) f) Unit of analysis (study population) • Sampling design g) Time horizon (cross-sectional/longitudinal) h) Quality of data (minimal bias & maximum reliability)
- 5. Research designs Research Purpose: The purpose of the study may either be: 1. Exploratory 2. Descriptive 3. Explanatory – It could be a combination of the above e.g. descriptive and explanatory; exploratory and descriptive.
- 6. Research designs Research Purpose: • Exploratory study: It is undertaken when not much is known about the phenomenon or limited information is available. – The aim is to enhance understanding of a particular phenomenon or to develop a viable theoretical framework. – An exploratory study is flexible and adaptable to change.
- 7. Research designs Research Purpose: • Descriptive study: It is undertaken to gain an accurate profile of events, persons or situations. – The ‘who, what, when, where and how’ questions. – Description is a means to an end and not an end in itself. Graduate research cannot be purely descriptive. – Qualitative research is descriptive in nature.
- 8. Research designs Research Purpose: • Explanatory study: It is undertaken to establish causal relationships between variables. – It can also explain the nature of certain relationships between variables through hypothesis testing. – Hypothesis testing is used to explain the variance in the dependent variable or to predict policy/ organizational outcomes.
- 9. Research designs Research Purpose: • Case study analysis: It involves an in-depth, contextual analysis of a phenomenon. – Case studies are useful in understanding certain phenomena, and generating theories for empirical testing. – They can also be useful in applying solutions to current problems based on past experiences. – Case studies are qualitative in nature but can be part of a mixed methods research design.
- 10. Research designs Research design types: • Experiment. • Survey. • Cross-sectional or Longitudinal study • History • Case study. • Ethnography. • Action research. • Narrative inquiry.
- 11. Research designs Causal vs. Correlational studies: • A causal study is done to establish a cause and- effect relationship. It means controlling for all other variables and ensuring internal validity (e.g. experiments). • Causal studies face the challenge of establishing attribution: To what extent can we attribute the variable y to the variable x? • Impact studies are one form of causal studies.
- 12. Research designs Causal vs. Correlational studies: • A correlational study is done to define the important variables associated with a problem. Examples • Does smoking cause cancer? • (A causal study question) • Are smoking and cancer related? • (A correlational study question)
- 13. Research designs Causal vs. Correlational studies: Examples • Fears of future terrorist attacks at the Coast led to an unprecedented number of European tourists cancelling their holiday visits. • Increases in interest rates and property taxes, the global recession and recent terrorist attacks have considerably slowed down direct foreign investment in Vietnam.
- 14. Research designs Unit of analysis:
- 15. Quantitative research • Research focus: Explain & predict. • Research approach: Deductive (using data to test theory). • Characteristics: – Examines relationships between variables – Numerical (precise) measurement. – Data analysis using statistical techniques. – Probability sampling techniques (Large sample size).
- 16. Quantitative research • Characteristics: – Incorporates controls to ensure validity of data (experimental design). – Structured & standardized data collection instruments. – Researcher is independent from those being researched (to minimize bias). – Subjects of the study are called respondents.
- 17. Quantitative research Research design types: – Experiment: true or quasi-experiment – Survey research – Archival research (makes use of both recent and historical records and documents) + Use of secondary data to build time series models of economic data e.g. + Time series analysis of the effects of electric power network expansion on economic growth in Kenya (Gewa, 2013).
- 18. Quantitative research • Research focus: Describe and interpret. Research that aims to tell the researcher how (process) and why (meaning) things happen as they do. The purpose is to gain in-depth understanding. • Research approach: Inductive (using data to build theory) – In practice, much qualitative research uses an inductive approach, where inductive inferences are developed and deductive ones tested iteratively throughout the research e.g. in grounded theory (Saunders et al., 2012).
- 19. Qualitative research Characteristics: • Researcher aims to gain an in-depth understanding. • Non-standardised data collection (emergent and flexible) that uses multiple methods. • Unstructured or semi-structured data collection instruments. • Non-probability sampling techniques (Small sample size).
- 20. Qualitative research Characteristics: •Researcher is dependent on the participants for cognitive access to their data. The researcher must therefore build rapport and demonstrate sensitivity to gain the trust and confidence of the participants. •Research takes place in the natural setting of the participants. •Data analysis through thematic content analysis.
- 21. Qualitative research Research design: – Action research e.g. Employee relations in a large telecommunications company undergoing a merger/acquisition. – Case study research e.g. Tourism development in Isiolo county. – Narrative Inquiry e.g. ‘Lest we forget’: Stories of the 2008 Postelection violence survivors. – Grounded theory e.g. Awareness of dying by Glaser & Strauss (1967). – Archival research e.g. “The Making of a Nation” by Hilary Ng’weno (2008).
- 22. Mixed methods research • Research focus: To enhance research quality. • Research approach: May use either a deductive or inductive approach, or combine both. – A theoretical perspective may be used to provide some direction for the research and limit its scope. – Quant/Qual. data can be used sequentially to develop a richer theoretical perspective.
- 23. Mixed methods research • Characteristics: – Uses multiple data collection techniques and analytical procedures to answer the research questions (Quant & Qual. data). – Quant/Qual. research is partially integrated within a single phase, usually sequentially (simple design). – Fully integrated Quant/Qual. research with multiple phases running concurrently (complex design).
- 24. Experiment • Often regarded as the gold standard against which the methodological rigour of other research strategies is assessed. • Commonly used in psychology and medical research (randomized control trials). • Its purpose is to examine possible cause-and-effect relationships among variables.
- 25. Experiment To establish that a change in the independent variable (IV) causes a change in the dependent variable (DV), all four of these conditions should be met: 1. The IV and the DV should co-vary. 2. The IV (presumed causal factor) should precede the DV. 3. No other factor should be a possible cause of the change in the DV. 4. A theory is needed to explain why the IV affects the DV.
- 26. Experiment • This is an experiment done in the natural environment in which life goes on as usual, but treatments are given to one or more groups (circumcision in HIV/AIDS prevention). • The experimental and control groups in the field experiment may be composed of households within a certain radius, or from different households in the same location.
- 27. Experiment • Internal validity – the degree of confidence that variable X causes variable Y. • External validity – the extent of generalizability to other settings, people or events. • Field experiments have more external validity but less internal validity compared to lab experiments. (The reverse is true for lab experiments.)
- 28. Experiment Factors affecting the validity of experiments:
- 29. Survey • This is a popular business/management research strategy. • It allows the collection of quantitative data that can then be analysed using descriptive and inferential statistics. • It is critical to ensure that your sample is representative if you intend to extrapolate findings to the whole population.
- 30. Survey • It is also important to ensure that you get a good response rate (30% minimum). • The questionnaire needs to be well designed to ensure internal validity as well as pre-tested to improve reliability.
- 31. Case study • This strategy explores a research topic within its context. • The acknowledgement of contextual variables makes it an antithesis of the experimental strategy where these variables are highly controlled. • It is a relevant strategy if the researcher seeks a rich understanding of the context of the research.
- 32. Case study • The case study is often used for exploratory and explanatory research. But it is also highly descriptive. • One strength of the case study is its ability to accommodate both quantitative and qualitative methods (document review, interviews, observation and questionnaires). • Data triangulation is therefore possible.
- 33. Case study • Yin (2009) distinguishes four case study strategies: – Single case vs. Multiple cases. – Holistic case vs. Embedded case. • A single case is usually chosen because it is an extreme or unique case (outlier) or because it is typical enough to develop theoretical propositions.
- 34. Case study • Multiple cases are used when the focus is replication of findings across cases. • Literal replication occurs when similar results are predicted to be produced for each case. • Theoretical replication occurs when a predicted variation is realized due to a difference in a contextual factor.
- 35. Case study • Yin (2009) advocates for a deductive approach to case studies where theoretical propositions are proposed and tested through an iterative process of examining rival propositions. • It is also possible to use an inductive or abductive approach where data is first collected and propositions generated, following which the propositions are tested.
- 36. Case study Research design types:
- 37. Case study Research design types:
- 38. Case study Research design types:
- 39. Case study Research design types:
- 40. Case study • A major pitfall regarding the embedded case design is when the researcher focuses only on the sub-unit level of analysis and fails to return to the larger unit of analysis. – E.g. A study of organizational effectiveness may involve a survey of individual employees as a subunit of study. The researcher needs to collect data from other sources within the organization otherwise the study will become an employee rather than an organizational study (Yin, 2009).
- 41. Case study Yin (2009) offers the following advice in selecting case study designs: A. Single or Multiple Case Design? Multiple case designs are preferred over singlecase designs. At least two cases are adequate. B. Closed or Flexible Designs? Flexibility, because a case study’s design can be modified by new information or discovery during data collection.
- 42. Case study C. Mixed methods designs – mixing case studies with other strategies? – Embedded case designs may rely on quantitative techniques to collect data about the sub-unit. – The case study may be a component of a larger mixed methods study e.g. a survey of schools involves a case study of one of more schools. • It is possible to mix case studies with other strategies but more time-consuming than single methods.