Two Discussion Board Post Due Sunday 27 November 1200 Eastern Standard Time

When designing research to answer a specific question or study a specific phenomenon, one must be sure that the research design is appropriate to comment on the issue being studied. To begin this research study, the researchers must first create a hypothesis, to determine what is being tested. In this situation, the hypothesis is that the percentage of defectives is dependent on temperature, humidity, and the level of artisan expertise. The testing of a hypothesis is extremely important in science, business, and even in the Bible. John instructs us: “beloved, do not believe every spirit, but test the spirits to see whether they are from God, for many false prophets have gone out into the world” (1 John: 4:1, English Standard Version). Similar to the way we are called to test the spirits of the world as Christians, these researchers should test all variables to determine what is affecting the quality of their glass shells. Therefore, a research study must be conduct that tests and comments on the percentage of defectives that each of these variables produces.

The goal of this research study is to offer a comparison of how many defectives are produced under each of the identified variables, there is no implementation of an outside variable and therefore the researcher is only looking at actual historical data with no influence from an experimenter. Since the data is only available after the fact and there is no opportunity to set research parameters before conducting the study to gather information, a quasi-experimental study design must be used. According to Cook and Campbell, “in quasi-experimental designs, outcomes may only be measured at the end of the study (rather than at the beginning and the end), there may not be a control group, or the patients (what is being studied) may be assigned to the control and treatment groups be methods other than randomization” (1979).

Considering the data that is available and the above definition of a quasi-experiment, some form of quasi-experiment must be used to research this data. The “intact equivalent design, in which the membership of the experimental and control groups in naturally assembled” is the most likely candidate (Cooper & Schindler, 2014, p. 209). Since there are two groups being studied, the functional and the defective glass shells being manufactured and these two groups are naturally separated and determined, this form of research makes the most sense to use.

In this experiment, the glass shells being manufactured will be separated into the two groups and then studied based on the three possible factors that could affect quality of the glass shells. It is noted that some experts believe that quality could be dependent upon production supervisor, however, since this data is not available for the entire time frame it should not be used. Cooper and Schindler states that instrumentation is a threat to internal validity that “results from changes between observations in either the measuring instrument or the observer” (2014, p. 203). Therefore, only having the production supervisor information for a portion of the days, this information should not be included in the same analysis as the other three factors. If the company were to really need to analyze this information, a separate study should be conducted in which the 242 days where supervisor was available were extracted from the data set and studied independently from the remainder of the period.

References

Cook, T. D., & Campbell, D. T. (1979). Quasi-experimental designs. Retrieved November 23, 2016, from http://ebp.uga.edu/courses/Chapter 7 - Observational and quasi-experimental/5 - Quasi-experimental studies.html

Cooper, D. R., & Schindler, P. S. (2014). Business research methods (12th ed.). New York, NY: McGraw-Hill. ISBN: 9780073521503.