Quantitative & Qualitative Analysis
In the physical sciences, data are typically measurable and quantifiable so that they can be meaningfully analyzed using inferential statistics. In the social and behavioral sciences, on the other hand, not all data of interest can be reduced to numbers in this way. Therefore, it is important that social and behavioral scientists collect both quantitative and qualitative data to better understand behavior. Qualitative research paradigms (i.e., field observation, survey research, and secondary analysis) give researchers a depth and breadth of the understanding of human behavior that cannot be otherwise gained. However, it is through quantitative research paradigms (i.e., experiments, simulation, field experiments) that hypotheses can be tested and meaningful predictions made of real-world behavior. Both qualitative and quantitative data and their concomitant research paradigms are important parts of social and behavioral science research, and are essential to understanding behavior and advancing these sciences.
When one thinks of research methodologies and the scientific method, variables, experimental design, and inferential statistical tools usually comes to mind. Although the experimental paradigm is assuredly part of the social and behavioral science researcher's toolkit, the questions investigated by social and behavioral scientists tend to be complex, and researchers frequently need to test models that contain numerous independent and dependent variables, as well as extraneous variables that may affect behavior but cannot be easily tested using inferential statistics. Further, research designs for use with human subjects are often limited by practical and ethical problems. For example, it may be logistically impossible to control all the variables that may affect the way that one is perceived in the workplace. Similarly, although it may be possible to manipulate the independent variable of whether one's spouse lives or dies in order to determine its effect on depression, it would certainly be both unethical and illegal to do so. In addition, constructs such as attitudes, opinions, and beliefs often do not translate well into quantifiable data on a scale with a real zero and equal intervals between points. Fortunately, social and behavioral scientists are not limited to the use of experimental paradigms in their quest to better understand and predict human behavior. There are a number of research paradigms available that can help researchers in their tasks. The choice of which tool to use depends on the goals of the research study and any practical considerations that may limit the degree to which the researcher can control the variables in the study.
The continuum of research paradigms can be broken into two general categories. Quantitative research comprises research studies in which observations are measured and expressed in numerical form, such as in physical dimensions or on rating scales. The results of quantitative research studies are typically analyzed through the use of inferential statistics. Quantitative research paradigms also offer the researcher varying amounts of control over the research situation. However, even when multivariate statistical tools are used for the data analysis, quantitative research paradigms are restricted in scope.
Qualitative research paradigms, on the other hand, enable the researcher to look deeper into the data but generally allow less control over the research situation. Qualitative research comprises studies in which observations are not or cannot be quantified, that is, expressed in numerical form.
Quantitative Research Paradigms
There are three primary paradigms for quantitative research: laboratory experiments, simulations, and field experiments. These research paradigms offer scientists various degrees of control over the research situation and the extent to which the situation realistically reflects the complexity of the real world. The results of these research paradigms typically allow researchers to apply deductive logic and reason from a general principle to predict behavior in specific instances. As is illustrated in Figure 1, the more the research situation reflects the complexity of the real world, the less control the researcher has over the situation. Laboratory experiments allow researchers the most control, over both the level of the independent variable that is experienced by the subjects and the various extraneous variables that can affect the outcome of the study. For example, if one wanted to determine how different personality types behave in specific situations, one could set up a simple laboratory experiment in which subjects with different personality types were exposed to situations that contained what the researcher believed to be the important characteristics of the situations included in the hypothesis. Subjects might be exposed to two confederates of the researcher, one who acted neutrally and another who acted aggressively. Responses of the subjects could then be measured against some objective criterion and the results statistically analyzed. This approach to data collection gives the researcher a great deal of control over the experimental situation (e.g., whether the subject is exposed to the neutral or aggressive confederate). However, interactions with an experimental confederate tend to be far removed from interactions with people during the course of an actual day in the real world. Because of this fact, the results of the controlled experiment would not necessarily be very generalizable.
If the researcher were willing to give up some degree of control over the experimental situation, he or she could design an experimental condition with more realism. For example, if the researcher is trying to predict how people with different personality types react to various customer attitudes in the workplace, a simulation could be designed. The subject could play the role of the sales clerk, and the experimental confederate could play the role of the customer. The confederates would be assigned to play the role either in a neutral manner or in aggressive manner. The researcher could then measure the reaction of the subjects to the confederates and statistically analyze the results. This simulation design is more complex than the laboratory design. In addition to interacting with the experimental confederates, the subject also needs to perform the tasks of the sales clerk. This brings many other variables into the situation that are not directly related to the research question, such as stress levels resulting from the sales tasks and previous experience as a sales clerk. As a result, the researcher has less control over the situation than he or she does in the laboratory. However, the simulation offers the researcher data from a more realistic scenario, meaning that the results will likely be more generalizable than the results of the laboratory experiment.
If the researcher were willing to give up even more control of the experimental situation, he or she could conduct a field experiment instead of a laboratory experiment or a simulation. In this paradigm, the confederates might interact with the subjects in a real-world sales setting. For example, the subjects could be actual sales clerks. During the course of their daily activities at work, they would interact with the experimental confederates, who would be posing as customers. This is an even more realistic situation than either the simulation or the laboratory experiment. However, such a field experiment also introduces numerous other variables that are not present in the other two scenarios. For example, the experimenter has no control over what other types of interactions the subject has during that day or how other extraneous variables might impact the subject's reaction.
A real-world situation tends to be very complex. The reaction of a subject to an experimental confederate may depend not only on the confederate's behavior but also on numerous other variables, such as how the subject is feeling that day, whether or not the sales counter is very busy, whether the subject's last interaction with a customer was positive or negative, and what types of interactions the subject has had with other store personnel or supervisors that day,...
(The entire section is 3634 words.)