Research methods and study design Flashcards
Experimental design
- Select population (only apply to selected population with SAME characteristics.
- Independent/ Dependent variable
- Controls/Experimental groups
- Random participants
- Randomly assign individuals to group
- Measure results
- Test hypothesis
Independent variable
Manipulated by researchers
Dependent variable
Measured, quant! Statistics! Reproducibility.
Operational definition
Specifications, precision
Control group
Critical; needs to be homogeneous, not same treatment
Placebo effect
Unrelated variables can make an effect. Double-blind to counter it.
Attrition
Participants dropping out of study
Randomized block techinque
Randomly assign individuals to group based on variables of interest to make groups similar.
Psychometrics
Measure psychological variables.
a. Between-subjects
b. Within-subjects
If both are used together is called “Mixed method research”
Between-subjects
Comparison both groups
Within-subjects
Same group, at different time periods
Type 1 error
False positives
Type 2 error
False negatives
Significant difference
Measured difference large enough that is not due to chance
p value
from 0 to 1. Probability difference observed in experiment by chance. Lower value, stronger relationship.
Sample size
Minimum 30 participants. Larger sample sizes are preffered.
External Validity
flaw make it hard to apply to others
Internal validity
Inherent flaw in design. When not taking considerations some variables
Predictive validity
Psychometrics
Demand characteristics
Participants behave the way they are expected to. Affect internal validity.
Ethical considerations in experimental design
- Disclosure given to participants
- Debriefing
Debriefing
Participants are told AFTER experiment what was done and purpose. If psychological vulnerability happened, treatment and counseling are given at the same time with debriefing.
Correlational studies
Pearson correlation.
Positive- both increase or decrease
Negative: opposite effects.
No correlation: 0
Etnographic studies
D[eth of analysis without interference. Studying culture, way of life.
Twin studies
Nature vs nurture. Heritability
Longitudinal studies
Interval measurements dependent variables over long time frames. Expensive, time/resources extensive, high attrition rate
Cross-sectional studies
Data collection or survey of population or sample at specific time.
Case studies
In depth exploration in an individual
Confounding variables
Extraneous variables not accounted for in the study; lack of useful controls
Lack of control group
Affect internal validity (design)
Phenomenological studies
Researchers studying themselves. Attempt to understand own perception. Introspective details. Difficult to generalize data. Small samples sizes, could be more than 1.
Survey
No accurate responses. Must be checked for consistency and accuracy,
Archival studies
Already collected data
Biographical studies
Observational studies
Individuals observed and outcomes measured with no attempt to control outcome
Extraversion
how energizing individuals find social interactions with others.