Studies design (Statistics)
Published in
1 min readJan 17, 2019
Types of studies:
Sample study
: To estimate ONE certain parameter of a population.Observational studies
: To compare TWO parameters of a population WITHOUT affecting them.Experiments
: To compare TWO parameters of a population and DOES affecting one and does not affecting the other one.
Samples
or surveys
- Problem 1: Qualitative and quantitative data
- Problem 2: Representative samples
- Problem 3: Biased wording in survey questions
- Problem 4: Sampling methods
- Systematic sampling: 100% members of ALL groups chosen.
- Stratified sampling: Some members from ALL groups chosen.
- Random sampling: An adequate number of members chosen, each an equal chance of being in the sample.
- Cluster sampling: 100% members from SOME groups are chosen.
Observational studies
& experiments
Observational study
: Measure or survey members of a sample WITHOUTtrying to affect them.Controlled experiment
: Apply some treatment to one of the groups, while the other group does not receive the treatment.
Notes
- Randomized experiments are designed to
suggest causation
Correlation
is WEAKER thancausation
- To answer a question about a causal relationship, we need to perform an experiment with a
treatment group
and acontrol group
. - While
sample study
need a part of relative members,Observational study
need ALL members.