Glossary
Blocking
The process of grouping similar experimental units together based on a variable known or suspected to affect the response, before randomly assigning treatments within those groups.
Example:
When testing different exercise routines, blocking participants by their initial fitness level helps ensure that groups are comparable before the routines are assigned.
Comparison
A fundamental principle of experimental design that involves using at least two treatment groups, often including a control group, to provide a baseline for evaluating treatment effects.
Example:
To see if a new teaching method is effective, you need to use comparison by having one group learn with the new method and another with the traditional method.
Completely Randomized Design
An experimental design where experimental units are assigned to treatment groups entirely by chance, without any prior grouping.
Example:
Assigning 50 students to either a 'traditional lecture' group or an 'online module' group purely by drawing names from a hat is a completely randomized design.
Confounding Variable
A variable that is related to both the explanatory and response variables, making it difficult to determine if the explanatory variable is truly causing the observed effect.
Example:
If a study finds that people who drink more coffee also live longer, but doesn't account for exercise habits, 'exercise habits' could be a confounding variable.
Control (Principle)
The principle of keeping all other variables constant besides the explanatory variable to minimize their influence on the response variable and reduce confounding.
Example:
In an experiment on plant growth, ensuring all plants receive the same amount of sunlight and water is an example of control.
Control Group
A group in an experiment that does not receive the treatment or receives a placebo, serving as a baseline for comparison.
Example:
In a drug trial, the control group might receive a sugar pill instead of the actual medication to see if the drug has a real effect beyond expectation.
Double-Blind
An experimental design where neither the subjects nor the researchers interacting with them know who is receiving which treatment, minimizing bias from expectations.
Example:
In a double-blind study for a new antidepressant, neither the patients nor the doctors administering the pills know if they are giving or receiving the actual drug or a placebo.
Experiment
A research method where a treatment is intentionally imposed on experimental units to observe its effect on a response variable, aiming to establish cause-and-effect relationships.
Example:
To determine if a new fertilizer increases plant growth, a scientist applies the fertilizer to one group of plants and observes their growth, conducting an experiment.
Experimental Units
The individuals or objects to which treatments are applied in an experiment.
Example:
In a study testing a new dog food, the individual dogs participating are the experimental units.
Explanatory Variables (Factors)
Variables that are manipulated by the researcher to see their effect on the response variable.
Example:
In an experiment comparing different types of exercise for weight loss, 'type of exercise' (e.g., running, swimming) is the explanatory variable.
Matched Pairs Design
A special type of randomized block design where subjects are paired based on a similar characteristic, or each subject receives both treatments in a random order.
Example:
To compare two different types of running shoes, a matched pairs design could involve having each runner test both shoes, one on each foot, or having identical twins each wear a different shoe.
Placebo
An inactive treatment that looks identical to the real treatment, used to account for the psychological effects of receiving any treatment.
Example:
When testing a new pain reliever, participants in the control group might receive a placebo pill that contains no active ingredients.
Placebo Effect
A phenomenon where subjects respond favorably to any treatment, even an inactive one, due to their belief that they are receiving a beneficial treatment.
Example:
A patient feeling less pain after taking a sugar pill, simply because they believe it's a powerful painkiller, is an example of the placebo effect.
Random Assignment
The process of assigning experimental units to treatment groups by chance, which helps to balance out lurking variables and reduce bias.
Example:
When testing two different apps for learning a language, using random assignment ensures that students are not sorted into groups based on their prior language ability.
Randomized Block Design
An experimental design where experimental units are first grouped into 'blocks' based on a known variable that might affect the response, and then treatments are randomly assigned within each block.
Example:
To test a new fertilizer, you might use a randomized block design by grouping plots of land by soil type (e.g., sandy, clay) and then randomly assigning fertilizer within each soil type block.
Replication
The principle of using multiple experimental units in each treatment group to ensure that results are not due to chance and to reduce variability.
Example:
To be confident that a new plant food works, you wouldn't just test it on one plant; you'd use replication by applying it to many plants in the treatment group.
Response Variables
The outcomes measured after applying treatments in an experiment; these are what researchers are trying to see change.
Example:
If you're testing a new study technique, the response variable might be the score on the next exam.
Single-Blind
An experimental design where either the subjects or the researchers (but not both) do not know which treatment is being administered.
Example:
If students know they are getting a new teaching method but the teacher doesn't know which students are in which group, it's a single-blind experiment.
Treatments
The specific conditions or levels of the explanatory variable applied to the experimental units.
Example:
If an experiment tests different dosages of a new medication, '10mg', '20mg', and '30mg' would be the treatments.