Chapter 7 Flashcards
(38 cards)
Randomized Experiments
The allocation of sampling units to groups or conditions done through random assignment. (Random Assignment) also known as randomization.
Sampling Units
A general way of referring to the participants, subjects, groups, or objects being studied (the units sampled from the population).
Propensity Matching
A statistically correct matching procedure often used in observational studies with very large samples.
Between Subjects-Design or Nested Design
When participants are exposed to one condition each.
Method of Agreement
If X, then Y;
X = PResumed Cause Y= Presumed Effect
Sufficient Condition
Adequate to bring the effect.
Method of Difference
If not X, then not Y; if the presumed effect Y does not occur, the presumed cause X is absent.
Necessary Condition
Implies that the cause is indispensable (You need the cause for the Effect)
Within-Subjects Design
Same subjects trying out different conditions. Also known as repeated measures design or cross-design.
Factorial Design
consisting of two levels of a variable. IV; Gender > Male;Female
Mixed Factorial Design
A design that consists of both between and within-subject factors.
Counterbalancing
To deal with the problem of systematic differences between successive treatments or measurements, you would use counterbalancing.
Latin Square Design
A square array of letters and numbers which represent treatment conditions. Each letter appears once and only once in each row and in each column.
Material Causality
Substance or substances which are thought to be necessary for the movement of something or for the coming into being of an event.
Formal Causality
Plan or development that gives meaning to the event.
Final Causality or Teleological Causality
the action is “goal directed” Refers to the objective or purpose of the event.
Efficient Causality
Refers to the activating or energizing force responsible for the event.
Covariation
Conjunction between cause and effect is not necessarily constant but is likely or probable.
Temporal precedence
“Priority” the assumption that cause X does in infact precede the Cause Y
.. some relationships can become a vicious circle of bidirectional cause and effect.
Internal Validity
Scientists attempt on logical and empirical rounds to rule out plausible rival explanations between presumed cause and presumed outcome
Quasi-Experimental
Resembling - Resembles a randomized design but it is not random.
Biased Selection
How sampling units assigned to different conditions were selected for those conditions.
Bias implies that the selection procedure resulted in groups that were systematically dissimilar in characteristics even before the experimental intervention.
Pre-Experimental Design
Designs viewed as too primitive and are seen as especially vulnerable to bias selection and causal misinterpretations.
One Shot Case Study
X - O
X= Exposure to the intervention (Experimental Treatment) O= Observation or measurement.