Session 2: Resaerch desing Flashcards
Resaerch designs
- cross-sectional
- longitudinal (chronological sequence, lagged design, cross-laged desing)
- experimental
Cross- sectional design
- independent and dependent variable are measures at the same point in time
- are there any differences between specific groups?
Limitations of the cross-sectional design
- limited regarding the possibility to examine causality (and the direction of effects)
- unable to identify changes over time
Longitudinal design
the longitudinal design involves at least two (three) measurement waves
3 options for longitudinal design
- chronological sequence
- lagges desing
- cross- laged desing
Chronological sequence
- measure the hypothesized flow of effects at different points in time
Lagged desing
modelling change (controlling for the DV at previous measurement wave)
Cross-lagged design
- a study of the relationship between two or more variables across time in which one variable measured at an an earlier point in time is examined with regard to a second variable measured at a later point in time, and vice versa
- extending the lagged desing
What to keep in mind when designning a longitudinal study
- time of measurements
- attention bias
- participants burden
Timing of measurements
- What is the existence interval of a construct?
- How long does an effect take to unfold?
- How long does an effect hold?
Attention bias
- Loss of survey respondents over time
- Problem: generalizability of results may be in question
- Testing attrition bias: t-test between responders and non-responders based on T1
Participants burden
the more frequent the measurements, the shorter the questinaires should be
Experimental design
a desing that assigns units to controlled conditions based on some chance process such as the toss of a coin
Four essential features of ecperimental desing
- manipulation of the IV (create different levels)
- control of extraneous variable
- randomization
- control group
Manipulation of the IV…
- how many variables do you manipulate?
- With how many levels?
- high fidelity manipulation, manipulation check
Control for extraneous variables
Extraneous variable: anything that varies in the context of a study
- make it difficult to detect the effect of the IV and the Dv
- we have to hold them constant (control for them)
Random assignment
usning a random process to decide which participants are tetsed in which conditon
- primary way to control for extraneous variables across conditions
Randomization: two criteria
- each particpant has an equal chnace of being assigned to each condition
- each participant is assigned to a condition indecently of other participants
Why is it important to have a control group?
to avoid biases due to maturation, history, and testing
Experimental design: Counterfactual
- if i change the cause, what is the effect
- attempt to eliminate confounds and alternative explanations
Repeated measures expeirment
participants are surveyed several times during th experiment
(before - baseline vs. after manipulation)
High internal validity
study is conducted in a way that supports the conclusion that the IV caused any observed differences in the DV
What to keep in mind when designing an experimental study?
- single-factor design vs. mult-factor design
- high fidelity manipulation
- manipulation check
- confoudnig variables
- random assignment
- repeated measures experiment
- experiment have a high internal, but rather low external validity
