November 18, 2012
I will compare and contrast quasi-experimental research and true experiments by addressing their weaknesses and strengths. Throughout my project I will give a detailed description of my experimental method used, as well as a thorough justification of why I selected this method as well as my sampling plan. I will also identify the target population, any ethical issues and my expected results. I will discuss the internal validity, external validity, limitations to my conclusion as well as my recommendations for future research. True Experiments
True experiments are considered the most accurate form of experimental research and are used to prove or disprove a hypothesis, or theory. True experiments are excellent for showing a cause-and-effect relationship. There is a random assignment of subjects or groups to treatments in true experiments with only one variable manipulated and tested. Random assignment controls for extraneous variables. The difference between participants or groups is based purely on chance. The strength of true experiments is causal control and strong internal validity. They are high on internal validity and variable being measured is clear. Randomization is the number one contributor to making an experiment a true experiment. Long term and reasonable sized sample groups; groups usually of more than 5-10 participants, are better for randomization. The strength of randomization is that it creates two or more groups that are approximately equivalent at the start of the research. As a weakness of true experiments, bias may be in the overall research design. The ethical considerations in true experiments are also a weakness. True experiments are seldom a good source of descriptive data about meaningful populations in experiments. The factors jeopardizing internal validity and changing a true experiment to a quasi-experiment are; history, maturation, testing, measuring of instruments, statistical regression, interactions with selection, subject mortality and selection. The nature of control groups can influence internal validity in this experimental design as well. Quasi-Experiments
Quasi-experiments, natural occurrences or in-situ experiments, are not randomly assigned and the subjects are beyond the control of the experimenter. A quasi-experimental design lacks one a control and or control group to be considered a true experiment. Studies with different levels of treatments, and people or groups assigned to treatments without random assignment are considered to be a quasi-experiment. Quasi experiments are not typically performed in the laboratory and study just groups of people, but can be used to study plots of land in an agricultural or environmental study as well. The benefits of quasi-experiments are practicalities and expense. It is easier and cheaper to set up a quasi-experimental design then it is to set up a controlled experiment. Variables are not always able to be manipulated so there are limitations to correlation research. In quasi-experiments there is more generalization of results to different people or other situations, which is external validity, a strength for quasi-experiments. Without generalization the results are not valuable to science. Pre-selection and randomization of groups are not one of quasi-experimental research strengths. Without proper randomization, statistical tests can be meaningless. Some treatment groups are initially formed on the basis of performance; high, medium, low and cannot be experimentally induced. The fact that quasi-experiments use intact groups and can introduce systematic errors, they are not always reliable. Individuals can enter different treatment levels through self-selection or because a researcher has "paired" individuals that they believe are somehow similar....