Common Sources of Error in the Research Design
It is frequently said that your research is no better than your research design. As we demonstrated, it is not difficult to specify the basic steps in preparing a list of research procedures that explain the basic research design. However, it is more difficult to recognize errors in the basic research design, especially for students who have little experience in empirical research. One of the best ways to become familiar with problems in a research design is to study other investigators’ research by reading their results in articles.
A number of common errors have occurred in the past and these errors are listed below with an explanation of the error. By becoming aware of these shortcomings it will be possible for you to upgrade your methodology before starting your data collection. Some of these are not easily avoidable, but they can be minimized by carefully considering possible common research errors.
The specific errors, which occur due to poor research design and its poor implementation, are as follows (Isaac, 1978):
The Halo Effect
This is the tendency for an inapplicable feature of a unit of study to influence the relevant feature during a favorable direction. Generally, a powerful initial positive or negative impression of a person, group, or event tends to influence the ratings on all resulting observations. Impressions formed early during a series of observations usually have an effect on later observations: or impressions supported high or low-status attributes of the unit of study have an effect on observations on unrelated attributes. A lot of vague and impressionistic the variable to be rated, the additional power is that the effect; the additional specific and clearly defined the variable, the less evident is that the effect.
Rating Errors
In addition to the halo effect, three tendencies plague the validity of ratings:
· The over-rater error - rating subjects in general on the side of leniency or favorableness.
· The under-rater error - rating subjects in general on the side of severity or un-favorableness.
· The central tendency error - rating subjects toward the middle of the scale. This often occurs when the observer is unusual with or unsure about what is being rated.
The Hawthorne Effect
In an industrial efficiency study performed at the writer Plant of Western electrical in Chicago during the 1920s, it had been observed that to single out a group of employees for special research makes them feel and act differently compared to regular employees. The effect of this was to bring about a consistent increase in productivity in spite of changes in the working conditions intended to both increase and decrease efficiency.
Explanations for this effect point to the factors of (a) novelty; (b) awareness that one is a participant in an experiment; (c) a modified environment involving observers, special procedures, and new patterns of social interaction; and (d) knowledge of results in the form of daily productivity figures and other feedback, ordinarily not systematically available.
The Experimenter Bias Effect or “Self-Fulfilling Prophecy” in Research
This is the bias you, as a researcher, are inclined to project into your methodology and treatment that shapes the data in the direction of your foregone conclusions. What you waiting for to see”, where you direct your interest, what you disregard or forget, what you keep in mind of record, and even the way you interrelate with subjects to alter your own hope and motivational states, all can influence the results to fit your preconception.
The Placebo Effect
Placebo is inert or neutral information given to subjects as if it were the active treatment variable itself. In medical research, the placebo is commonly an injection of sterile water or with chemicals inert pill administered so the topic cannot distinguish between neutral and active ingredients. This allows the observer to separate treatment effects from psychological reactions included by the treatment situation involving, in particular, the patient’s expectations and suggestibility. For this reason, background fundamentals in an educational research setting often turn out to be the agents of transform in addition to the active treatment variable.
For example, in a research study to examine the effects of videotape playback on increasing participation in class discussion using treatment and control groups, it might turn out that the importance in contrast to the presence of videotape recording equipment and any operator or observer was of equal or greater. Unless the same equipment and personnel are present in at least some of the control classes as a placebo condition the com effectiveness of the playback may be misconstrued.
Post Hoc Errors
After this, so as to this.” several actions follow each other without having a cause-and-effect relationship. The explanations for this vary all the way from easy coincidence to advanced relationships with different factors. For instance, an increase in the ocean’s temp is directly related to the quantity of drowning on the California coast. We don’t conclude that the warming surf, per se, is a lot dangerous however it attracts a larger number of swimmers, exposing a lot of individuals to the probability of drowning.
The error of Misplaced precision
To collect information with great care and precision however inside the framework of a faulty design invalidates the findings and therefore the careful work that created them. when properly evaluated, this can be additionally demoralizing to the employees related to the project.
Beware of “Typical” Case Studies
Studies based on “typical” cases are sometimes biased and not representative. Such cases too typically are a lot of ideal than typical, or higher match the reporter’s biases than others, which could be cited. A defendable difference would be to select and review many cases every which way, a procedure seldom followed.
The Law of the Instrument
This refers to the human inclination to become attached to a particular instrument and apply it as an across-the-board resolution to each problem. As Resta and Baker (1967) place it, “If you provide a tiny boy a hammer, he can realize that everything he encounters needs pounding” Most of the favored devices, innovations, and panaceas in management are vulnerable to this response. In such instances, the focus is on the instrument or procedure and not the matter, putting the cart before the horse. Examining the matter initially then seeking appropriate instruments and procedures that best fit the properties of the matter avoids this error.