Basic research investigates the truth of existing theory or attempts to discover new knowledge that will further the development of theory. Applied research is designed around a specific problem, and the research questions, when answered, will help solve one or more aspects of the problem.

An example of basic research would be a study investigating the therapeutic value of the three critical attitudes and values postulated by Person-Centered Therapy. Another might look at whether any of the value changes postulated by Beatrice Wright are, in fact, associated with adjustment. Still another might study individuals with disabilities to see which of the stage models for adjustment seems to be the most relevant.

An example of descriptive research would be a study designed to see if individuals who are members of minority groups receive the same extent of rehabilitation services as individuals who are members of the majority group. The independent variable here would be group membership, and the dependent variable might be the total dollar expenditure on consumer rehabilitation programs.

Descriptive research is correlational in nature. It usually investigates whether one or more groups differ from one another on one or more dependent variables. In the simplest design, there would be only two groups and one dependent variable. Stated another way, there would be one independent variable with two levels and one dependent variable. An example would be investigating whether the IQs of men and women were significantly different or not. The independent variable here would be gender (with two levels or groups, men and women), and the dependent variable IQ scores for all the studied men and women.

Experimental research is different from descriptive research in that there is a control group. Rather than being correlational, it seeks to establish cause and effect. In the simplest design there would be one independent variable with two levels and one dependent variable with subjects randomly assigned to the two levels of the independent variable. An example would be a test of the effectiveness of a new AIDS medication where individuals with AIDS were randomly assigned to two levels of an independent variable that might be called "Treatment." One group would receive a placebo and the other the medication being studied. The dependent variable would be some measure of health or improvement. Such a study should be "double blind" to eliminate possible contamination by the experimenter or treatment staff expectations. If the group receiving the medication scored significantly higher on the dependent measure, a cause and effect relationship would be a reasonable conclusion.

In group based research, subjects are sorted into groups they naturally fall into (as when, for example, demographic factors are being studied), or randomly into experimental groups defined by the researcher(s). There are one or more independent variables and one or more dependent variables, and in the simplest design there is one independent variable with two levels and one dependent variable. Group based statistical tests are then employed to look for possible differences.

In single subject research only one person is studied, although there maybe cross validations of the procedures with other single subjects. A baseline on the behavior(s) of interest to the researcher(s) is established, and then an intervention is made that may have some potential to affect the behavior(s) under study. A post measurement of the behavior(s) is then made to see if there is any difference from the pre intervention baseline.

Dependent Variable: The factor that is being measured and free to vary within the different levels of the independent variable being studied. It is usually scores obtained by subjects on some test or instrument.

Independent Variable: This is the variable that may or may not have some relationship to values taken by the dependent variable. This is what the study is trying to determine.

Null Hypothesis: The null hypothesis is simply the statement that there will be no significant difference in the values taken by the dependent variable for the different levels of the independent variable being studied.

Type I Error: An old term that means the null hypothesis was mistakenly taken to be false when it was in fact true (i.e., the researchers concluded there were significant differences when there were not).

Type II Error: An old term that means the null hypothesis was accepted as true when it was in fact false (i.e., the researchers concluded there were no significant effects when there actually were).

Significance Level: A probability value set for concluding there are in fact differences in the studied data not due to chance. A significance level of p < .05 means the data differences would occur by chance, given the sample size, only 1 time in 20. A significance level of p < .01 means that only 1 time in 100, given the sample size, would the identified differences result simply from chance alone.

Sample: The group of subjects being studied. The sample should be representative of some larger population (that it was drawn from) so that results obtained for the sample may be generalized to that population. To the extent that the sample differs, from the larger population it claims to represent, the generalizability of findings are reduced. This is a common limitation of research studies, and in some cases very problematic.

Population: The larger group the study is hoping to learn more about. Because populations are usually too large to study completely (i.e. measure every member), a sample is taken and the results for the sample are generalized to it.

Generalizability: The extent to which results found for the sample can be extrapolated and applied to the population it claims to represent. A sample should be well representative of a population if results are to be considered applicable to that population.