Common Mistakes in Sampling

Index: Common Mistakes in Sampling

How to Investigate Persons

A common mistake in educational research is to investigate persons from the appropriate population simply because they are available. For example, a researcher might select all subjects from one school because she happens to know the school principal and is sure that the principal will grant permission to do the study in that school. The problem with this strategy is that the research results cannot be applied with much confidence to other objects who are members of the same population. Suppose that the researcher has selected all subjects from one school, and subsequently finds that subjects exposed to teaching method A learn significantly better than subjects exposed to teaching method B. A principal in another school car, legitimately raise the question: How do I know that teaching method A will be superior in my school. Generalization from one school to another or from one sample of students to another is risky, unless the researcher has selected subjects by means of appropriate sampling techniques.

Common Mistakes in Sampling

How to select subjects

An even worse error is to select subjects who are not even in an appropriate population for the contemplated research, merely because they are easily available. For example, many studies on the effects of different types of psychological counseling or therapy are conducted using normal college students who have no need for counseling but who are drafted as subjects merely because they are students in a class in psychology. It is doubtful whether the responses of such students to counseling bear any relationship to responses of persons in the real target population, that is, persons who seek counseling because of serious emotional problems.

College sophomores, who are available but often inappropriate, have been the subjects for so much research in education and psychology that the use of sophomores in research projects finds its way into many of the jokes about research workers in these fields. Some studies suffer relatively little from using available subjects. For example, exploratory studies on the effects of drugs upon behavior are not seriously weakened by the use of available subjects. Whenever the research worker washes to generalize the results to specifically defined populations, however, the use of subjects merely because they are available is inappropriate.

Common Mistakes in Sampling

Bias

Many educational field studies are biased because the research worker chooses experimental and control groups from different populations. For example, some early studies on the effectiveness of TV instruction used high school students receiving conventional instruction as a control group, but used adults who wanted to complete their high school education in home study as an experimental or TV-instruction group. The age, interests, motivation, and dropout rates for the two groups are very different, thus making the results of such studies meaningless.

Occasionally identification of a suitable sample is sufficiently rime-consuming and expensive to warrant the use of shortcuts. Nevertheless, the possible effects of shortcuts should be carefully studied before they are used. Terman's famous study of "gifted children" provides an example of a sampling shortcut that seriously affected the research results. In this study Terman wished to locate 1000 children with IQs over 140 on the Stanford-Binet Intelligence Scale. The Stanford-Binet test is expensive and time-consuming to administer. Therefore, rather than test many pupils who had little chance of obtaining a score of 140, Terman decided to ask teachers to suggest the names of students whom they considered superior. Only those students nominated by the teacher were tested. The difficulty with this procedure is that teachers tend to underestimate the intelligence of pupils who create disturbances and are not cooperative in the classroom. Thus Terman's sample does not include this type of individual, and as a result of this shortcut in selecting the original sample, all the findings of his important study must be qualified. Terman's results, instead of being applicable to gifted children in general, refer primarily to a particular type of gifted child.

The method of selecting a sample population is critical to the whole research project. If research findings are not generalizable to some degree beyond the sample used in the study, then the research cannot provide us with new knowledge, cannot advance education as a sdence, and is largely a waste of time.

Common Mistakes in Sampling

How to Select a Sample

3The sample should not only be selected so as to be representative of the population from which it is drawn, but it should also be large enough so that the investigator can be reasonably sure that if he had drawn a different sample using the same procedures he would have obtained similar results in his research. If an adequate number of subjects is not included in the sample, then one's confidence in the research findings will be shaky.

There is a difference between the characteristics of a sample and the characteristics of the population from which the sample was drawn. This difference, which can be estimated for random samples, is called sampling error. Sampling error is a function of the size of the sample, with the error being largest when the sample is small. Research findings based on a sample of two or three subjects are apt to be highly unreliable.2 If we studied another sample of this size, it is quite likely that different findings would be obtained. It is important to select a sample of adequate size in order to produce research data that reliably approximates the data that would be obtained if the entire population were studied.

This chapter discusses sampling techniques that enable the researcher to select a sample that is representative of a larger population. We also discuss procedures that the researcher can use to determine the sample size needed for a given study. Although our discussion is concerned primarily with the selection of subjects for a research project, the student should note that the sampling techniques discussed here also pertain to the selection of events or objects for research. For example, sampling techniques would be used if the researcher wished to select a sample of class periods for systematic observation, or a sample of textbooks in order to do a content analysis.

Another Research Tips

Labels:

Infoskripsi Bandulan, Malang and work as an Administrator at Infoskripsi Corp.
0
Newer Post
Home
Older Post
The item being reviewed 4 5 24