Research Design

What is the order of a research design?

1. Variables and unit

2. Hypothesis

3. Design

4. Cases/sample

5. Measurement and data

6. Method

7. Threats and ethics

What is involved in variables and unit?

IV, DV, Unit, Controls

What is involved in hypothesis?

H1 and H0

What is involved in design?

Quant, Qual or mixed, longitudinal or cross-section, what’s compared, explain design, justify design

What is involved in cases/sample?

Population, scope, selection method, selection bias check

What are the selection methods?

Random, all, purposive, quota

What is involved in measurement and methods?

What is measured and where it comes from for each variable, validity, limitation

What is involved in method?

Describe, Simple relationship, model, interpret, outliers

What is involved in threat and ethics?

2-3 threats and defence, internal validity, external validity, reliability, reproducibility, ethics

Simple relationships

Categorical X and Y - cross-tabulation - % of each X in each Y
Categorical X Continuous Y - compare group means
Continuous X and Y - Pearson’s correlation - shows strength of linear pattern

Models to use

Continuous DV - OLS (Linear) Regression
Binary DV - Logistic Regression

Experimentation Case

What it is: Run an experiment and test results

Use when: there’s something to show or tell people

Strength: Internal validity

Weakness: Artificial, weak external validity

Quasi-Experimentation Case

What it is: Real life assigns treatment; compare treated vs untreated before/after.

Use when: Something is introduced that changes things

Strength: External Validity, Time Order

Weakness: Selection Bias, no control

Cross-Sectional Case

What it is: Looking at a cross section of population at one time

Use when: Many units, X can’t be assigned

Strength: Generalisable, Cheap

Weakness: No time order

Longitudinal

What it is: Trend looks at changes over time, panel is sample of individuals over time

Use when: Change over time matters

Strength: Fixed time order,

Weakness: Cost, Confounders

Intro Strengths

Situates RQ in literature

Identifies a clear gap

Clear research question

Road map of the paper

Intro weaknesses and remedies

Claims with no citation - Cite peer-reviewed, official sources

No roadmap - add explanation of sections

Answer built into RQ - Reframe neutrally

Doesn’t synthesise studies - group by theme and have them interact

Conclusion Strengths

Links findings back to literature

Restates RQ and summarises findings

Acknowledges limitations

Suggests future research

Conclusion weaknesses + remedies

Introduces new evidence - move it to the analysis section

Overclaims - Be more honest about what findings mean

No limitations or future research - include it

Doesn’t link back to intro - show what findings mean