/*Please comment if you find any mistakes*/
A treatment is what an experimenter would like to do among the subjects.
(“In experiments, a treatment is something that researchers administer to experimantal units”)
The quantitative aspect of a treatment is controlled and described by ‘level’.
(“Treatments are administered to experimental units by ‘level’, where level implies amount or magnitude.”)
The variable of an experiment set by the experimenter.
(“A factor of an experiment is a controlled independent variable; a variable whose levels are set by the experimenter.”)
- Independent variable
The variable that is set by the experimenter.
- Dependent variable
The variable that is observed or measured.
A measure is the process that leads to a single unit of observation.
(“It is the same thing as a dependent variable”)
- Level of measurement
The types of data we can obtain from an experiment (nominal, ordinal, ratio, and interval).
A trial is an instance of the treatment that produces a tuple of values of the variables (independent and dependent).
The degree to which variables change together.
- Within-subjects factor
The factors designed for and are consistent across individual subject groups.
A factor all of whose levels are experienced by each subject;
(“factors associated with measurements made on an individual subject”)
Example: when measuring/evaluating a technique, changing conditions like task, usage condition, etc.
- Between-subjects factor
The factors designed for comparing between different subject groups.
A factor each level of which is only experienced by one subject.
Example: comparing various techniques by asking groups of people to use each of them.
- Factorial design
Given multiple factors (each with various levels), a factorial design exhausts all possible combination of these factors in the experiment.
- Main effect
The effect of a given factor over the dependent variable. (“It is the effect of the factor alone averaged across the levels of other factors”)
The treatment where the presence of multiple factors result in a unique effect – beyond the sum of their individual effects.
- Within-subjects design
All factors are within-subjects factors.
Designing the treatment for a subject group to see how the within-subjects factors affect the dependent variable(s).
- Between-subjects design
All factors are between-subjects factors.
Designing the treatment for across subject groups to see how the between-subjects factors affect the dependent variable(s).
- Mixed factorial design
There are within-subjects and between-subjects factors.
A variable to correlates to both independent and dependent variables.
A control group/condition is a baseline that receives neutral or no treatment, and is used to be compared with other groups/conditions that yield results and observations.
- Carryover effect
When multiple factors are applied, the effect of one factor affects the forthcoming ones.
Example: the fatigue of using the first technique ‘carries over’ to the second technique.
To eliminate carryover effect, the possible orderings of applying multiple factors are distributed across the subject groups.
Example: Latin Squares.
- Balanced design
In a balanced design each factor runs the same number of treatment for its levels
(“In the Design of Experiments a Balanced Design (Balanced Experiment) is a factorial design in which each factor is run the same number of times at the high and low levels.”)
- Nominal variable
Variables that represent identities, e.g., different techniques, different medicine.
(“A set of data is said to be nominal if the values / observations belonging to it can be assigned a code in the form of a number where the numbers are simply labels.”)
- Categorical variable
Variables that can be sorted to a definite set of categories.
(“A set of data is said to be categorical if the values or observations belonging to it can be sorted according to category. Each value is chosen from a set of non-overlapping categories. “)
- Ordinal variable
Variables that can be ordered but the difference between a given pair depends on their definitions, as opposed to mere calculation.
- Continuous variable
Variables whose domain is a range of (infinite number of) values, as opposed to a finite set of values.
(“A set of data is said to be continuous if the values / observations belonging to it may take on any value within a finite or infinite interval”)
- Scalar variable
- Fixed effect
A fixed effect is a factor whose levels are deliberately chosen and thus of interest to the experiment, e.g., the different techniques and postures used in studying a new interaction technique.
- Random effect
A random effect is a factor whose levels are randomly chosen and thus usually not studied, e.g., the participants recruited for studying a new interaction technique.
- Mixed-effects model
Some factors are fixed. Some are random.
- Long format
Each row only contains one trial
- Short format
Each row contains all the trials for one subject