Field Data Collection
Data for behavioral research can be gathered in a number of ways. Although in experimental paradigms researchers have a great deal of control, the results of such studies often have limited generalizability and cannot account for the great complexity of variables experienced in the real world. On the other hand, field data collection techniques such as field observation, field research, and unobtrusive measures offer the researcher little or no control but are rich sources of information about the way people actually act in the real world. The data gathered using these methods enable researchers to apply inductive reasoning to real world data so that the variables contributing to behavior can be better understood and testable hypotheses can be formulated. Field data collection tools are an important part of the behavioral scientist's toolbox and can add to our understanding of human behavior in significant and important ways.
Keywords Data; Ethics; Field Observation; Hypothesis; Inductive Reasoning; Inferential Statistics; Operational Definition; Scientific Method; Subject; Survey; Survey Research; Unobtrusive Research; Variable
Field Data Collection
Research in the behavioral and social sciences can be both fascinating and challenging. Not only does social science research require the application of the scientific method, it also requires the use of a great deal of creativity in order to obtain data concerning intangible constructs, phenomena that change when they are directly observed, or complex situations affected by numerous extraneous variables not directly related to the hypothesis being tested. The complexity of social science research was vividly and hilariously illustrated by a recently published cartoon. Captioned simply, "If Einstein had been a social scientist…," the cartoon showed the back of a wild-haired individual scribbling furiously on a blackboard. Rather than the expected E=MC2 equation, however, the concept of "relativity" was expressed in terms of the square root of parents multiplied by sibling rivalry, the square of in-laws divided by the first marriage, and numerous other nonsensical terms. The resultant equation filled the blackboard while the researcher continued to articulate variables and relationships that needed to be considered.
In the behavioral and social sciences, there often seems to be a plethora of variables that need to be taken into consideration in the quest to understand and predict behavior. As a result, the findings of research studies frequently bring up more questions than they answer. For example, a researcher might want to determine the relationship between the time it takes to readjust after the death of a spouse and the time that the couple had been married. This is a simple enough relationship at first glance, at least until one takes into account other considerations: How dependent had the spouses been on each other? Does the surviving spouse have a strong support network of family and friends that can help in the readjustment period? Does the surviving spouse have a strong religious faith? Did the couple have children who also survived? Was this a first marriage? If not, what caused the end of the first marriage? The list of possible variables other than length of the marriage that might also have an effect on the outcome of the relationship is seemingly endless. Even if a researcher could articulate all the major factors to be considered and design a research paradigm that could be analyzed using inferential statistics, he or she would still face an ethical problem in collecting the research. It would simply be unethical to randomly assign married people to the various experimental conditions and then manipulate whether or not their spouses lived.
Because of the complexity of social science issues and the ethical considerations in the treatment of experimental subjects, social and behavioral scientists are often required to be very creative in the operational definition of their variables and concomitant data collection methods. One way to collect data is through survey research, in which data about the opinions, attitudes, or reactions of the members of a sample are gathered using a survey instrument that is administered in a paper-and-pencil (or electronic) form or by an interviewer. Surveys have the advantage of being able to collect information on non-tangible constructs (e.g., feelings, attitudes, and opinions of subjects) that are difficult to collect directly. Further, surveys can be relatively inexpensive to administer, as they require no manipulation of variables and have relatively low costs associated with data collection. However, even well-written surveys cannot provide the researcher with the answers to all questions of interest. What one says and what one does are often two different things. Research subjects may lie on a survey instrument in order to look good to the researcher (or even to themselves) or give responses that are not well thought out because they are not motivated to participate honestly in the survey. On the other hand, subjects' actions in response to the manipulation of an independent variable tend to reflect their real reaction. However, in addition to situations where it is unethical to manipulate variables to measure subjects' reactions, there are also instances where the mere fact that a subject knows that a researcher is watching may change his or her behavior.
Further, survey instruments do not yield the same kind of neat interval or ratio data that are gathered in most experimental designs. This makes the data problematic to analyze. Many commonly used inferential statistical tools assume that the data being analyzed have been randomly selected from a population that has a normal distribution and require data that are interval or ratio in nature. This means that not only do the rank orders of the data have meaning (e.g., a value of 6 is greater than a value of 5) but so do the intervals between the values. In the physical sciences, such assumptions are typically easy to meet: it is clear that the difference between 1 gram of a chemical compound and 2 grams of a chemical compound is the same as the difference between 100 grams of the compound and 101 grams of the compound. Physical measurements have meaning because the weight scale has a true zero (i.e., we know what it means to have 0 grams of the compound) and the intervals between values are equal. However, it may not be quite as clear that the difference between 0 and 1 on a 100 point attitude scale is the same as the difference between 50 and 51 or between 98 and 99. These are value judgments, and the scale may not have a true zero. For example, the scale may go from 1 to 100 and not include a 0. Similarly, even if the scale does start at 0, it may be difficult to define what this value means. It can be difficult to articulate how a score of 0 on this scale differs significantly from a score of 1 or even what a score of zero means (e.g., does a score of zero mean that the person has no opinion?). In addition, one cannot tell from the scores why the subject assigned the values, a question often of interest to social scientists. Even if the various points on the scale were well-defined, different people may give vastly different responses to indicate the same attitude. Ratings are also subjective, and although numerical values may be assigned to them, they do not necessarily meet the requirement of parametric statistics that the data can be at the interval or ratio level.
Field Research and Observations
One way to collect better data in some situations is through the observation of subjects in a real world setting. This can include field research, the collection of unobtrusive measures, and field observations. Like experiments and surveys, these methods also have advantages and disadvantages. However, they provide the researcher with additional tools for data collection that can aid in the quest to understand and predict behavior. Like all research tools, these methods should be selected only after careful consideration of what data are needed, what the practical and ethical limitations are in collecting the data, and what the statistical limitations are for analyzing the data. Although field research tools offer the researcher less control than laboratory experiments and simulations, they have the advantage of allowing subjects to be observed in a natural setting where the intrusion of the experimenter is unlikely to be noticed.
The Uses and Drawback of Field Research
Field observation and research typically allow the researcher no control over the experimental situation. Therefore, this approach to data collection is often considered inferior to other methods. However, it must be remembered that it is through the application of inductive reasoning to individual observations in the real world that testable hypotheses are generated. This is not only one of the first steps in the scientific method, it is also an essential step without which more controlled data collection could not be conducted. Further, because of the complexity of human behavior in real world situations, it is often beneficial to observe people in field settings in order to better understand the interaction of variables causing their behavior. On the one hand, field observation and research frequently do not yield high quality, quantitative data that can...
(The entire section is 4124 words.)