The planning and conducting of research that will aid in the advancement of any behavioral or social science requires the choice of a research paradigm that appropriately balances the researcher's ability to control the research setting while maintaining an adequate and accurate representation of the complexity of the real world situation. When working with human subjects, the researcher is further required to take into account various ethical considerations to ensure that no physical or psychological harm occurs to the subjects as well as that the data are as valid as possible. For this purpose, professional codes of ethics have been developed to guide researchers in the ethical conduct of experiments and studies. In particular, researchers need to guarantee as much as possible the confidentiality of subjects' private information, acquire informed consent from research participants, and take all precautions possible during research planning, implementation, and dissemination to adequately and accurately present their research findings.
The conduct of the scientific research, by which sociology and other behavioral and social sciences advances, can be both challenging and rewarding. The antecedents of human behavior that we see all around us are complicated and interrelated, requiring the creative application of the scientific method in order to gather data to better understand and predict human behavior. There are a number of research tools available to social and behavioral scientists which supports this task. As shown in Figure 1, these research paradigms offer scientists various degrees of control over the research situation and the degree to which the research situation realistically reflects the complexity of the real world.
Laboratory experiments allow researchers the most control not only over the level of the independent variable that is experienced by the subjects, but also over the various extraneous variables that can erroneously affect the outcome of the study. For example, if one wanted to determine the relationship between how an individual dresses at work and how that person is treated by others, one might set up a simple laboratory experiment in which pictures of individuals in various types of attire (e.g., dark business suit, business casual, casual) are presented to subjects who are then asked to rate the professionalism of the person in the picture. This approach to collecting data to test the research hypothesis gives the researcher a great deal of control over the experimental situation (e.g., how the people in the picture are dressed). However, rating the "professionalism" of people portrayed in photographs is far removed from real world settings, so the results would not be widely applicable. If the researcher is willing to give up some control over the experimental variables, s/he would be able to design an experimental condition with more realism.
A simulation could be set up in which subjects interact with experimental confederates who dress in various types of attire as specified by the experimenter. This research paradigm still offers the researcher a great deal of control over the experimental situation (e.g., she can specify exactly how the confederates will dress), but its increased realism concomitantly gives the researcher less control (e.g., extraneous variables such as the way the confederates talk, their attitude, and other variables not related to the research hypothesis can affect the response of the subjects). Further, although a simulation is more realistic than a laboratory experiment, it still only remotely emulates the real world situation.
Giving up a little more control in favor of a higher degree of realism, the researcher could conduct a field experiment in which confederates interacted with the subjects in a real-world business setting. However, this situation would allow for the greater possibility of the influence of extraneous variables than the more controlled simulation and laboratory experimental paradigms. In some respects, this can be both an advantage and disadvantage. Although one's attire in the workplace has been shown to affect the way that one is treated, the way one is treated also depends on many other variables as well (e.g., behavior, grooming, attitudes of the other person, competence). The complexity of these variables can be better seen in field settings than in more controlled paradigms.
Real world situations tend to be very complex, however, particularly when one is trying to determine what variables affect human behavior. In many cases, it would be virtually impossible for a researcher to sufficiently articulate all the real world variables that influence behavior in a way that would allow a hypothesis to be empirically tested using inferential statistics. Statistical tools are available for modeling real-world behavior, but these typically require the collection of vast amounts of data from real-world observation. For such tasks or for the purposes of collecting individual observations for the application of inductive reasoning, more realistic research paradigms are needed. For example, although a researcher might be able to use a more controlled research paradigm to collect data on various levels of the dependent variable (e.g., business attire, business casual, and casual dress), in truth there are virtually infinite combinations of the ways that people can dress at work. Is a dark suit more impressive than light suit? If so, does the suit need to be black, or would dark gray or navy blue be just as impressive? Does the suit need to be plain or do pinstripes add to the professional aura? The list of permutations on just this one level of attire is seemingly endless. Similarly, how does one best define the way that a person is "treated at work"? Once again, in the real world there are seemingly endless ways in which this can be defined ranging from the politeness or friendliness with which they are treated by peers, supervisors, and customers to the hard data of number and frequency of promotions, amount and frequency of raises and bonuses, scores on performance appraisals, just to name a few measures.
Field studies are examinations of how people behave in the real world. For example, a researcher might either directly or unobtrusively observe how various people are treated in the workplace, recording his/her observations both on the treatment received as well as the person's attire. Another approach would be to employ the paradigm of survey research. Subjects could be interviewed by a member of the research team or asked to fill out a questionnaire regarding the way that they typically dress at work and how they perceive the treatment they receive from peers, supervisors, and customers. This could be combined with other information such as the amount and frequency of raises, bonuses, or promotions.
Theoretically, survey research allows researchers to gather the most information about the situation under investigation. However, although a very thorough interview or survey instrument can be written that would hypothetically gather all the data needed for the researcher to make decisions about the antecedents of treatment in the workplace, such instruments are often more lengthy than the potential research subject's attention span. Further, as opposed to the other research techniques, surveys and interviews are not based on observation. Therefore, there is no way to know whether or not the information being gathered from the subject is true. As a result, information gathered from research paradigms more to the right of the continuum in Figure 1 tend to be difficult to empirically test and determine the validity of the underlying hypothesis. So, the behavioral researcher is left with a dilemma.
The Middle Road
At first glance, it might seem that the best research paradigms lie somewhere in the middle of the continuum shown in Figure 1. Paradigms in the middle of the continuum still allow the researcher a good deal of control over the experimental situation while allowing for more widely applicable results due to the increased realism of the research situation. This tempting rule of thumb, however, is weakened by the fact that the ultimate goal of most researchers is to do research that adequately and accurately reflects the real world situation (i.e., is highly realistic) so that it can be extrapolated and used in predictions. It also allows a great deal of control over the variables so that the results can be statistically analyzed and the probabilities of the...
(The entire section is 3806 words.)