EXAMPLES - TWO BETWEEN GROUP FACTORS

EXAMPLE 1

You are interested in whether psychologists, psychiatrists and psychiatric nurses have different views on the degree of biological causation of depression, but you also wonder if gender of clinician may predict their views. You cannot randomly assign people to profession or sex, so you have to use NATURALISTIC GROUPS. Your target sample is 20 male psychiatrists, 20 female psychiatrists, 20 male clinical psychologists, 20 female clinical psychologists, 20 male psychiatric nurses and 20 female psychiatric nurses.

The first between group factor is profession (psychiatrist, psychologist, nurse).

The second between group factor is gender of clinician (mail versus female).

These factors may also be called the INDEPENDENT VARIABLES or PREDICTOR VARIABLES.

Your measure of views about causation is the DEPENDENT VARIABLE.

Each of the independent variables may be associated with a MAIN EFFECT - that is to say either profession or gender of participant may be associated with significantly different views on causation of schizophrenia.

Additionally or alternatively, there may be an INTERACTION EFFECT between profession and gender of participant. This means there is an association with views on causation that is not just the sum of any association found for profession and any found for gender of participant.

EXAMPLE 2

You do an experimental study investigating the differential effects of therapy type and gender of therapist.  There are three therapy types, and each participant may have either a male or female therapist. Participants are randomly assigned to therapy type and either a male or female therapist.

The first between group factor is therapy type (of which there are three).

The second between group factor is gender of therapist (mail versus female).

These factors may also be called the INDEPENDENT VARIABLES or PREDICTOR VARIABLES.

Your measure of outcome is the DEPENDENT VARIABLE.

Each of the independent variables may be associated with a MAIN EFFECT - that is to say either therapy type or gender of therapist may be associated with a significantly different outcome.

Additionally or alternatively, there may be an INTERACTION EFFECT between therapy type and gender of therapist. This means there is an association with outcome that is not just the sum of any association found for therapy type and any found for sex of therapist.

Back to Which Test Home Page