Experimental Design¶
Introduction to Experiments¶
Experiments¶
Components:
- Experimental unit
- Treatment
Variables:
- Explanatory variable
- Responsive variable
Observational Study¶
Do not influence the individual
Types:
- Retrospective
- Data is collected then analyzed
- Prospective
- A sample is observed for a period of time into the future
Well-Designed Experiments¶
Confounding Variables¶
graph LR;
A[Explanatory] --> C[Responsive]
B[Confounding] --> C
A --> B
- In an experiment, confounding values need to be controlled.
- In an observational study, confounding values should not and cannot be controlled
Well-Designed Experiments¶
- \(\geq 2\) treatment groups (a control group counts)
- Randomization: Treatment groups are formed by randomly assigning treatments
- Replication: \(\geq 1\) experimental unit per treatment group
- Confounding variables are controlled
Control Groups, Placebos & Blind Experiments¶
Control Groups¶
A treatment group in an experiment that are purposefully given either
- an inactive form of the treatment
- or the pre-existing treatment
Placebo¶
A dummy treatment that the experimental units believe is an active treatment
- Control groups can use placebos as inactive forms of treatment
The placebo effect is the response made by experimental units to a placebo
Blind Experiments¶
A single-blind experiment is the one in which the participants do not know which treatment they are receiving
- Helps reduce the placebo effect
A double-blind experiment is the one in which participates and researchers do not know which groups are assigned which treatments
- Helps reduce the placebo effect
- Reduce any bias
Completely Randomized Design¶
All experiment units are randomly assigned
- Creates treatment groups that are as similar as possible
- Helps balance out any effects of confounding variables across the groups
Randomized Block & Matched Pairs Design¶
Block¶
A group of experimental units who have something in common that may affect how they respond to a treatment
- It should only be done if the researcher believes the blocking variable could affect the results
- Blocking allows natural variations in response to treatments to be distinguished from those variations that were due to the blocking variable
Randomized Block Design¶
- Experimental units are separated out into blocks
- Experimental units are randomly assigned the different treatments within each block
- Each block needs to be randomly assigned all of the treatments
- A randomized block design is generally better than a completely randomized design
- If blocking variables are unknown or the sample size is very large, completely randomized designs should be used
- Larger samples tend to introduce more blocking variables
- So each block has smaller sample size
Matched Pairs Design¶
- Each block has only two experiment units which are matched either naturally or manually based on some common factor
- Two treatments
- A matched pairs design is better than completely random design
- Easier to distinguish the effectiveness of the treatment