Determine whether you are primarily collecting statistical data. This may involve tests of performance or recall, opinion surveys, physical measurements, or participants' responses to multiple-choice or short-answer questions. This is quantitative research, where the data are analyzed using statistical techniques.
Determine whether you are primarily collecting narrative or textual information. This may involve in-depth interviews with primarily open-ended questions, such as "Tell me about your family." This is qualitative research, where the data are analyzed using techniques of textual or thematic analysis.
Determine whether you are collecting both statistical and narrative/textual information. This is mixed-methods research, which combines aspects of both quantitative and qualitative research.
Determine whether you are administering an experimental treatment to one or more groups. An experimental treatment might be exposure to a special class or training technique, a medication, a nutritional supplement, radiation, or an experience, such as being in the audience of a play or a vignette, or being spun about in a centrifuge. If you are administering an experimental treatment, you are conducting experimental research.
Determine whether you have a quasi-experimental study. For example, you may be comparing groups that occur naturally, such as groups distinguished by sex or country of birth. Or, you may be comparing groups that differ by the choice of the participant, such as religion or occupation. You may be comparing groups that differ in other ways that you as the experimenter do not control, such as ethnicity. Studies like this are examples of quasi-experimental or correlational research.
Determine whether you are conducting psychometric research. This involves research primarily intended to develop an assessment tool. The tool may be meant to assess, for example, intelligence, mental illness or opinions about a specific topic.
Determine whether you are conducting descriptive research. This involves collecting behavioral observations or survey data, without dividing participants up into groups, without any experimental manipulation, and without the objective of developing a new psychometric instrument.
Note whether or not you assign participants to groups, where the groups differ in terms of the experimental treatment they receive, or whether they receive a placebo or sham treatment, conventional treatment, or even any treatment at all. An experiment with one or more groups who receive sham treatment, conventional treatment or no treatment is said to be a controlled study, because these are different types of control groups.
Note whether you assign participants to groups randomly. Experiments that use random assignment are stronger designs.
Label your experimental design accurately. If you lack either control groups or random assignment to groups, you have an experiment, but one with a weak design. It may be best to label such a design a preliminary study.