Methodology, Research Design & Concepts in inferential Statistics Flashcards
Hypothesis Construction & Testing
Define methodology
A system of explicit rules and procedures upon which research is based and against which claims for knowledge are evaluated (Frankfort-Nachmias and Nachmias 2005, 13).
What are the features of quantitative methodology?
- Measurement of Variables = Numerical Data
- Objectivity
- Universal truth or laws (of human behaviour)
- Statistical testing
- Testing theory and Hypotheses (Deductive reasoning)
- Generalizability (related to large samples)
What are the 3 research design options?
- Classical Experimental Design
- Quasi-Experimental Design
- Survey Design
Discuss the classical experimental design
Features of the classic experimental design:
1. Two groups
i. Experimental group
ii. Control group
The groups are equivalent (have the same characteristics) except that the experimental group is exposed to the independent variable and the control group is not.
Cases are randomly assigned to each group
Researchers measure the dependent variable in both groups:
* Before the experimental group receives the independent variable – pre-test
* After the experimental group receives the independent variable – post-test
The difference in pre-test and post-test results are compared.
What are the benefits of the classical experimental design?
- Helps us understand the logic of all research designs.
- Able to measure the effect of an independent variable and infer causation.
What is the quasi-experimental research design?
Similar to the experimental design with one main difference – the researcher does not randomly assign persons to groups. They usually involve the study of more than one group of persons, often over an extended period of time (Frankfort-Nachmias and Nachmias 2005, 131).
Elements of classical experimental
Classical Experiment
- Emphasis on internal validity
- Assess cause and effect in a relatively artificial environment
- Participant are randomly assigned to experimental and control groups
- Control is maintained throughout study
Elements of quasi-experimental design
Quasi-experiment
* Emphasis on external validity
* Describe ‘real’ events in a naturally occurring environment
* Groups are naturally occurring/existing; no random assignment
* Full control not possible
What are the elements of descriptive statistics?
Graphical:
* Frequency tables
* Charts
Numerical:
* Central Tendency
* Variability
* Position (Z-scores & percentiles)
What are the elements of inferential statistics
Estimation:
* Point estimate
* Interval Estimate (confidence interval)
Hypothesis Testing:
1. Concepts
* Significance level (alpha)
* Probability value (p-value)
* Error (Type 1 & 2)
2.Statistical Tests
* Pearson Correlation
* Linear Regression
* T-tests
* ANOVA
* Chi-square
Why is reporting research findings necessary?
Reporting research findings is important for dissemination and for synthesis and evidence-based management (EBM). Primarily, the importance lies in dissemination across conferences, journals, books, and increasingly digital media.
What are some factors to consider when designing a study?
hypothesis, research objectives, literature review, theory, research ethics,
resources (financial, physical, human), time, research skills, validity, reliability,
components of research design, selection of research design, data collection
method, sampling
What are the data collection methods?
- Questionnaire
- Observation
- Focus Group
- Interview
Point vs Interval Estimate (confidence interval)
A point estimate is a single number. Whereas, a confidence interval, naturally, is an interval. The point estimate is located exactly in the middle of the confidence interval. However, confidence intervals provide much more information and are preferred when making inferences.
A confidence interval will provide valid result most of the time. Whereas, a point estimate will almost always be off the mark but is simpler to understand and present.
The confidence level refers to the percentage of instances that a similar study will capture the true mean (accuracy) of the population being tested.
if we want to be 95% confident that the parameter is inside the interval, alpha is 5%.
Type 1 vs Type 2 errors
Type I error occurs if the null hypothesis is rejected when it is true (a false positive)
Type II error occurs if the null hypothesis is not rejected when it is false (false negative)
