Research Methods Flashcards
Types of research methods
Define Quantitive vs Qualitative research
- Quantitative:
- Measures “how much”
- Traditional methods all quantifiable: controlled experiments or surveys - Qualitative:
- Measures understanding “why, what, when, and how”
- Methods: interviews, focus groups, narratives, case studies etc
Types of research
Define cross-sectional vs longditudinal study methods
- Cross-sectional:
- Measures taken at one time point
- Issues with causality unless you conduct an experiment - Longitudinal:
- Measures taken at two or more time points
- Better case for cause and effect, limitations if not experimental
- Confounding or “third” variables
Types of research
Define experimental va quasi-experimental designs
- Experimental: Researchers randomly assign participants to one condition, treatment, or group.
- Quasi-experimental: Group of interest cannot be manipulated, occurs naturally (pre-existing) or unethical to manipulate. For example;
- Gender, sex, race, ethnicity, socioeconomic status
- Tobacco, alcohol, or drug use
Issues with social experiments
Define ecological validity in social experiments and why this may pose an issue.
- Modelling social phenomena in a controlled lab environment
- Not always accurate representation of normal contexts
Issues with social experiments
Define the use of confederates in social experiements and why this may pose an issue.
- Actors who play the role of someone in the study.
- Not always convincing… use of online, virtual confederates seems more reliable than using in-person
confederates
Types if research methods
Define Meta-Analysis
Effect size?
Use of statistics to combine results of several individual studies addressing similar questions into a single pooled measure of an effect size (strength of association between variable).
Identifying Variables
Define: Construct
Attempt to capture a pattern of behaviour that is difficult to capture directly due to vagueness
- e.g. depression, racial prejudice, self-esteem
Identifying variables
Define: Operational definition
Attempt to identify way to capture a construct
- Can be quantitative or qualitative
- Vary in reliability, but there are usually endless options
Explicit vs implicit measures
Explain what is meat by explicit operationalizations
- Conscious and deliberate
- Self-reported by participant (e.g., survey)
Explicit vs implicit measures
Explain what is meant by implicit operationalizations
- Unconscious and automatic
- Gut reactions or impulsive decisions made by participant
- Behavioral tasks, natural observation, facial expressions, etc.
Explicit vs implicit measures
Implicit Association Test (IAT)
- Measures strength of automatic associations between concepts
- Flower vs. Insect (Attitude Object)
- Pleasant vs. Unpleasant (Evaluation) - Interpreted as implicit preferences or attitudes
- When two concepts are associated it is easy to respond quickly and correctly when categorizing
- Flower-Pleasant
- Insect-Unpleasant
- This leads to quicker/faster responses to these pairings. - When two concepts are not associated it is difficult to respond quickly and correctly when categorizing
- Flower-Unpleasant
- Insect-Pleasant
- This leads to slower responses to these pairings
Explicit vs implicit measures
Explain The D-Score
A Measure of bias from IAT data
- Positive values = congruent bias e.g. White-Good, Black-Bad
- Negative values = incongruent bias, e.g. Black-Good, White-Bad
Estimated by subtracting reaction times (RTs) between congruent and incongruent pairings, divided by SD of RTs
Explicit vs implicit measures
What are some issues with IAT?
- Makes assumptions about groups and forces categorization
- May prime participants with stereotypes, rather than assess them
- Some targets may not have appropriate comparisons
- Alcohol vs water? Alcohol vs soda?
- Gender equality initiatives?
Open Science
Exploratory analysis
- Examining data for patterns, relationships, or trends without a specific hypothesis in mind
- often used to generate hypotheses or insights that can be further investigated in confirmatory (hypothesis-driven) research
- Since exploratory analyses involve multiple comparisons and data exploration, there is a higher risk of chance findings
Open Science
Brian Nosek
- Co-founded OSF in 2012
