Eveland's Research Design Is Quasi-experimental The Sample Essay
- Length: 5 pages
- Subject: Business - Ethics
- Type: Essay
- Paper: #19849949
Excerpt from Essay :
Eveland's research design is quasi-experimental. The sample populations for the experiment are not randomly selected. There is structure to the experiment with more than one form of measurement during the research process. Quasi-experimental design includes multiple groups and multiple waves of measurement. While there were not a significant number of groups in the experiment, there was more than one. At more than one point in the experiment, measurements were taken.
According to researchers such as Trochim, there are a few different types of research design. A randomized experiment is just that -- it is a random assignment which is different from random selection. Random assignment is a way in which comparison in experiments may be better facilitated because it helps create similar groups within the experiment population. If there are any differences among the population it is because of chance and not because of some kind of manipulation or premeditation by the researchers. Random assignment is advantageous because with this type of research design, the population coagulates into groups or groups with similarities on its own. There is less direct influence of the division within the experiment or research population by the researchers. The quality of randomness allows for the data to guide itself or direct itself. Perhaps with this design researchers have a relatively greater since of detachment from the work and the experiment may operate more independently from researcher bias or manipulation.
Experimental designs are arguably the most frequently used type of research design. Trochim calls this kind of the design the gold standard by which other forms of design are compared. (2007). Internal validity is a critical factor in experimental design. This design is strongly reinforced and contingent upon internal validity. Internal validity is, too, strongly related to cause and effect, or as Trochim states, causal inferences. (2007).
Experimental design aims to create groups within each experiment that are similar or equivalent, whatever equivalency may mean within the context of a specific experiment. The aim is to create equivalent groups between the control group and the treatment group, whatever that may mean as defined by the parameters of the specific experiment. This is the type of design that many researchers use in aims to prove their hypotheses and theories. It is advantageous to use because it is well-known and used often. It may not reflect much creativity because of its frequent use and it may be the incorrect approach for every experiment. It is very challenging as far as the limitations of the researcher's control over groups, yet it may be useful in challenging researchers in their definitions and establishing internal validity. Perhaps it is a good method to test internal validity, but not the best option when the actual experiment is conducted. Within this design schema, there are several sub-categories of research design within experimental design including hybrid experimental design.
Quasi-experimental design, the method this paper argues Eveland uses, is again, roughly self explanatory. It is not quite fully experimental design, yet not quite fully random, but elements from both kinds of design are present. Hybrid forms of theory or design are often challenging to use and may not be as compelling, yet they offer a unique perspective and advantage in the research, data collection, and data analysis.
While this design may not be used as often as say, experimental design, it still is useful and yields quality results. It offers some of the better qualities of both of the aforementioned design schemas. It may be difficult for researchers to master elements of both designs, or it may assist researchers in their mastery of other research approaches because quasi-experimental design compels them to understand both perspectives. Eveland uses this design approach in his work. There is some balance between control and detachment in such a design approach.
Research projects move through stages just as people move through stages in life. Research stages and sequence follow roughly the same sequence; there is general consensus among those who study the stages of research. The sequence typically includes: planning, research set up, checking the research & design, the actual research & fieldwork, the data analysis, the interpretation or conclusion of the research, and the follow up after the research is complete. My position on this issue is that the sequence is logical and is followed with consistency with professional research.
Research questions are essentially connected to the form the research design takes. There are instances where the research question dictates the research design. There are other instances where the research design necessitates the type of question the research question will take. There are times when the research question implies or even restricts the type of research design the research will take. In some instances, the research design that a researcher selects can help him/her form a better or clearer research question. All of this is to say, that again, the research question and the design are critically linked. If these two fundamental elements of the research project are out of sync, then there is greater likelihood that the research project/study/experiment will fail or at the least not be internally valid or consistent.
External validity refers to the application of the research design, question, and/or results to other projects and populations outside of the original research project. External validity refers to the validity of inferences based on the researcher's experiments. External validity questions how well those inferences apply to other people or in general situations.
Eveland chose his participants based on their retirement status. He chose participants that had already retired and he chose participants that were of retirement age, or approaching retirement yet still working. He additionally conducted structured interviews at regular interviews, differentiating between those who relied more on in person informal meetings and those who relied more heavily upon technology such as networked computers for support. The selection of these participants allowed for him to narrow down his conclusions and actually make some viable generalizations. On the other hand, the mentality of retired persons and those of retirement age but are still working are quite disparate, which likely affected the results/data. I do agree that people who use information technology primarily to work are afforded greater flexibility and perhaps comfort when performing work tasks. This conclusion holds true outside of the study. I am uncertain as to whether the conclusion that those who work more with technology are linked moreso and rely more upon the group than those who did not.
The participants were randomly assigned to the controlled condition. In this case, I believe the internal validity is at least medium. There is a balance between what is random and what is controlled. This should provide at least adequate to substantial validity. Internal validity exists within an experiment if there is a clear and present causal relation between two variables. The experiment is internally valid with this kind of relation is demonstrated by and within the experiment. Once there is clear demonstration of internal validity, causal inferences may be made based on the relation between the variables. Causal relation occur when there is a linear relation between cause and effect; the cause and the effect exhibit covariation; and that this covariation cannot feasibly explained by any another reason besides the observed covariation.
If participants are randomly assigned, my position is that the when measuring the internal validity solely on that aspect, there is medium internal validity. Random assignment will statistically produce appropriate, intended, or hoped for outcome. Random assignment will statistically produce outcomes that are in opposition to the intended or hope for result. Essentially with random assignment, researchers win some and they lose some. Medium internal validity is a minimum. Internal validity can increase with random assignment if the research design and question complement each other such that random assignment plays in the favor of the research project or hypothesis, or if at least, random assignment is accounted for in the formation of the research question and research design.
There were not severe protection issues with this case. There may have been some mild inconvenience for those who are retired, as they might have other activities that they would rather spend their retirement doing. Behavioral research is governed by ethical principles that protect things such as the subject's self-respect, privacy, and life. Researchers in any field, including behavioral, face dilemmas with respect to the ethical treatment of their research subjects. Researchers often become quite engrossed and impassioned about their research, sometimes to the point where the subjects involved may be in danger or exploited.
The Federal Human Subjects Protection rules are in place to ensure that human subjects for research are protected from exploitation and unreasonable danger or harm in various forms. These rules have been in place in American society for not very long -- approximately three decades. It makes one wonder what researchers were able to get away with respect to their subjects before these rules were in place. It makes one wonder as well what kind of reception these rules…