Research is a systematic searching for information—a process of inquiry. What distinguishes scientific research from other research are the same characteristics that distinguish science in general from other ways of knowing: Scientific research is a systematic process of inquiry that is based on combined empirical-rational principles. That combination makes heavy demands on the adequacy of the information it uses.
A scientific assumption is a statement or an idea that we tentatively accept as being true, without examining it further. Scientists share several basic assumptions about nature and the role of science in understanding nature:
Empirical observations constitute the facts of research. Facts are those events that can be directly, empirically observed. In psychology, observed facts include the physiological structures of participants, the physical conditions around them, and the behavior of the participants and the other organisms around them, including the researcher. The major category of facts studied in psychology is the behavior of organisms.
In addition to studying behavioral facts, psychologists also study nonbehavioral events such as emotion, attitudes, intelligence, thinking, and so on. These events are not classified as facts because they cannot be directly observed; instead of being observed by the researcher, they are inferred by the researcher. Drawing an inference is an intellectual process in which we arrive at ideas based on observations and/or other ideas. The inference is a creation of the researcher.
Nonobservable inferred events—such as gravity, electricity, memory, or the ego—are rational ideas that have been created by the researcher. They are not facts. The ideas constructed in this way by the researcher are called constructs. Once constructed they are used by the researcher analogically, i.e., as if they were actually facts.
Constructs help to explain relationships among facts. In science we can develop constructs based on facts. Further, we can then use the constructs as bases for predicting new facts. By systematically putting together facts and constructs we create models, which are representations of parts of the universe. The real universe and any phenomenon in it can be represented by models. These models help us to organize knowledge about the reality that the models represent. Conceptual models in science are important in helping us to organize facts and constructs. All models share the following characteristics:
When we use facts as a basis for developing constructs, our thinking moves from a specific event (an observation) to a more general idea (a construct). This rational step from specific to general is called induction or inductive thinking. Moving rationally from a general construct to a more specific fact or idea is called deduction or deductive thinking. Both are rational thought processes that together form a cornerstone of science. {02:01}
A theory is a formalized set of concepts that organizes observations and inferences and predicts and explains phenomena. Theories are carefully constructed from empirical observations and constructs using both inductive and deductive logic. We make the distinction between several types of theories. Inductive theories begin with a solid empirical base and gradually build to more abstract levels. Deductive theories represent the more traditional theories in which the abstract concepts are central and deductions from the theories are the focus of research. Functional theories involve an equal emphasis on both induction and deduction. A fourth type of theory is the model, which is an analogical representation of reality. Models are usually more limited in their scope and less developed than other theories. {02:02}
Models can be developed to represent any events, including the events that occur within scientific research. The model of research presented here—like all scientific models—is an incomplete and tentative representation of reality. This model is labeled Research: A Process of Inquiry. It includes two major dimensions: the phases of research, and the levels of constraint.
Psychological research usually proceeds in an orderly manner, from initial ideas to problem definitions, to procedures design, to empirical observations, to data processing, interpretation, and communication. These are the phases of research. Each has its own characteristics, and different work is accomplished in each. (02:03}
Idea-generating phase. All research begins with an idea, sometimes quite vague. The researcher's interest in the idea is critical.
Problem-definition phase. Vague ideas are not sufficient. In good scientific tradition, we must put more demands on these ideas to clarify and refine them. In the problem-definition phase, the researcher systematically develops and refines initial ideas. This usually involves a thorough search and understanding of the relevant literature.
Procedures-design phase. In the procedures-design phase, the researcher systematically determines what observations are to be made and exactly how they are to be made. Here ethical issues must be considered (e.g., are participants placed at risk, and if so, how are they to be protected?).
Observation phase. In this phase, the researcher carries out the procedures that were determined in the previous phase. The observation phase is central in all science. Empirical observations constitute the facts of research. When the researcher records the observed facts, that record constitutes the research data. In the remainder of the research process, it is these data that are analyzed, interpreted, and communicated.
Data-analysis phase. In most psychological research the data will be in the form of a numerical record representing the observations, and those numerical data must be put into some order and further processed. Statistical procedures are used to describe and evaluate numerical data. Determining which statistical procedures to use is not difficult; the choice is largely determined by the nature of the question and the observational procedures.
Interpretation phase. Having statistically analyzed the data, we continue to make sense out of them by interpreting the statistical results in terms of: (1) how they help to answer the research question, and (2) how this answer contributes to current knowledge in the field. We take the answers found in the research and see how well they fit into our theoretical predictions and constructs.
Communication phase. Science is a public enterprise, and one of its most basic components is communication among scientists. This communication occurs through both oral presentations at scientific meetings and written accounts in journals and books. The APA has published guidelines for the preparation of scientific reports. A great deal of communication among scientists occurs informally. {02:04}
Each research project proceeds through phases, as discussed above. Research is also carried out at various levels of constraint, i.e., levels of demands made on the adequacy of the information and on the processing of that information. Thus, within science, some methods of research are more demanding than others. This concept—levels of constraint—provides the second dimension of our model of research. It is a dimension of precision, structure, and control. By levels of constraint we mean the degree to which the researcher imposes limits or controls on any part of the research process.
Naturalistic observation. The naturalistic level of constraint involves the observation of the participants' natural flow of behavior within their natural environments. No constraints are placed on the participants.
Case-study method of observation. Case study is a somewhat higher level of constraint, and the researcher does intervene with the participant's functioning to some degree. Case-study research is not limited to research on psychopathology or psychotherapy; it is a set of methods that can be applied to many issues.
Correlational research. Correlational research requires greater constraint on the procedures used to measure behavior. The main focus is to quantify relationships between two or more variables. Therefore, we must use precise (constrained) procedures for measuring each variable.
Differential research. Differential research involves an explicit comparison between two or more groups of participants defined by some preexisting variable. In order to make the comparison meaningful, all groups must be treated in the same way. That means that the settings and observational procedures must be constrained across groups. If done properly, the only thing that is not the same across groups is the level of the variable that defines the groups.
Experimental research. Like differential research, experimental research involves an explicit comparison between two or more groups. A major difference between the differential and experimental research is the way participants are assigned to the groups. In experimental research, participants are assigned to groups in an unbiased manner, usually randomly.
All of the levels of constraint are properly scientific when used appropriately, and we must let the nature of the question determine the level of constraint used in answering the question. Problems arise when researchers inappropriately mix levels of constraint. In doing research we should refine questions so they can be answered at the highest constraint level possible given current knowledge in the field and ethical and practical constraints on the researcher. Finally, it is important to note that all research involves the study of relationships among events, and the types of relationships vary from one constraint level to another.