An experimental treatment is applied to a group of people, and then a measure of interest is taken. The experimenter examines the measure to see if it is in the hoped-for range.

A group is measured on a variable of interest. The group receives the experimental treatment. Then the measure is again taken. The experimenter then compares the pretest with the posttest to see if the hoped-for change has occurred.

Two naturally-occurring groups are selected. One Group receives the experimental treatment, and the other does not. Both groups are then measured. The experimenter then compares the two sets of measurements to see if there is a difference.

About all can be said for these designs is that sometimes they are the best you can do.

Science is about comparison and contrast. Without comparing what happens to people with a particular treatment versus what happens to similar people without the treatment, with care taken to insure that conditions are equivalent in both groups minus the treatment, you really cannot be sure that your treatment actually produced any results observed.