Chapter 11 - Inferences about Means of Two Independent Samples Flashcards
Dependent variable
The measured outcome of interest.
Independent variable
Also called treatment variable. The variable whose value defines group membership.
True experiment
A study were the experimenter assigns subjects to groups at random. The experimenter has total control over the treatment variable and randomly assigns a level of that variable to each subject.
What does the randomness of true experiments guarantee?
It guarantees that treatment variable will be assigned to each subject entirely independent of any naturally occurring characteristics of that subject, which is why the variable is called the “independent” variable in a true experiment.
What other studies have the same general form of general experiments but the experimenter does not have control over the treatment variable?
Quasi-experiments, in situ studies, and passive observational studies.
What is a status or classifying variable?
In quasi experiments, in situ studies and passive observational studies the experimenter does not have control over the treatment (independent) variable. E.g. the assignment to groups is not random and controlled by the experimenter - gender may be a naturally occurring characteristic of the subject - which is therefore called status or classifying variable (not independent variable).
Is the statistical or computational analyses of true experiments and other studies the same?
Yes.
What is robust?
Insensitive to violations of the assumptions.
What do we need to look at in order to reject the null hypothesis for a two sample test?
The smaller the standard deviations of the two groups, the more readily we will reject Ho. The amount of overlap, and therefore the decision as to whether to reject Ho, depends not only on the distance between the means but also on the width of the distributions.
