Scientific knowledge is one way among many ways of knowing (epistemology). The scientific methods rests upon empirical observation and reason. Scientific evidence requires the testing of statements that potentially can be shown to be false (disconfirmed).
In order to determine a cause-and-effect relationship, 3 conditions must be met:
Behavioral science researchers generally are interested in exploring concepts or constructs such as personality, aggression, maternal behavior, perception, learning, etc. Concepts are abstract generalizations. Theories are constructs (concepts) that attempt to explain natural phenomena. Because they are abstract, they are not directly observable. Theories are tested by generating hypotheses, which are then confirmed or disconfirmed. Even without a theory, research questions are framed as hypotheses - testable statements or propositions that can be supported or rejected by the evidence.
For clarification, the concepts covered in a theory are operationally-defined, that is, they are defined by the techniques used to measure them. These definitions become the variables used in research. Variables have values or levels. These are the dimensions by which they vary. The dimension can be continuous (varies along a continuum), or categorical (the variation is discrete -- one level or another).
Variables play three roles in research:
Reliability refers to constancy or consistency in measurement. Reliable procedures are repeatable and replicable. Validity is the credibility or genuineness of results. Internal validity means that the instrument or procedure measures what it is supposed to measure. External validity means that the results are generalizable beyond the immediate situation. Research methods vary with regard to their internal and external validity.
Terms to know (define each before clicking to see the definition in a pop-up message)
categorical variable constant continuous variable data dependent variable disconfirmation divination empirical empiricism epistemology external validity extraneous variable factor falsifiability hypothesis independent variable internal validity levels (of a variable) phenomenon proposition reliability third variable problem theory validity values (of a variable) variable