Experiments Flashcards
What is a natural experiment?
An experiment where one has control of background variables, but uses no manipulation or intervention
What is a observational study?
An experiment where one uses no manipulation or intervention and has no control over background variables.
What is repeatability?
there is enough information available about an experiment so that a competent person other than the original experimenter can repeat the experimental procedures
What is reproducibility?
If a competent repetition of the same experiment yields the same result
What is replicability?
If a competent, independent expert in the spirit of the original experiment, with independent data, analyse method, laboratories, instruments etc. yields the same result
What is an experimental artefact?
Experimental artefacts are experimental errors. It is something that was unintentionally created in the experiment. An example is the Hawthorne effect: the workers were more productive as a consequence of them being observed.
Why does experimental artefacts occur?
They come about through:
1) Misunderstanding the nature of the experimental manipulation
2) Wrongly describing the background conditions of the experimental setting
3) Falsely reading the experimental observations
Name three defining features of an experiment.
1: Control of background variables through manipulation.
2: Intervention on target variable through manipulation.
3: Observation of difference produced by the intervention.
What is a field experiment?
In a field experiment, one maintains the found conditions in both treatment and control group, varying only the intervention on the feature of interest.
What is a laboratory experiment?
In a laboratory experiment, one constructs the background variables in both the control group and treatment group and then perform interventions on target variables. This type of experiment has a high level of control.
What is a simulation experiment?
Here one constructs a representation of the real-world object/scenario in a computer and then intervenes on target variables to then observe the difference.
State 4 necessary steps in a measurement procedure.
1: Definition of the concepts to be measured.
2: Specification of the measurement operation.
3: Specification of a measure and definition of a unit.
4: Representation by numbers/a measurement scale.
State five criteria for what makes a good operationalization
1: The concept to be observed is defined.
2: The property of interest is connected to an observable property.
3: Connection between the property of interest and observable property is valid to the best of our knowledge.
4: The connection between the property of interest is sufficiently stable for practical purposes.
5: The observable property is observable with sufficient precision in practice.
What is an operationalization?
An operationalization links a feature of interest which is not directly observable or measurable to an observable property.
What is construct validity (when speaking about operationalization)?
It is a measure on to what degree an operationalization successfully measures what it is intended to measure.
