If the students were randomly assigned, then a pretest would not be necessary because the differences between the students will be eliminated. The differences are eliminated because they are randomly assigned, which will leave the differences to chance. On the other hand, if the students were not randomly assigned, then it would be necessary to administer a pretest. This is because there could be some students who are more fluent in French than their peers. The pretest allows the instructor to measure how much the students improved over the semester. An example is if a student scored 35% on the pretest and then a 90% on the final, they improved greatly. But if a student scored 85% on the pretest and then a 90% on the final, they displayed minimal improvement. Due to the fact that this experiment is quasi-experimental (Trochim, W.M. 2006) in this question, the instructor should not administer a pretest in this case only if they plan to randomly assign students to the four classes. This is a quasi-experiment (Trochim, W.M. 2006) because the students were not randomly assigned to one of their four French classes, but when they are randomly assigned, the experiment will resemble a true experimental …show more content…
In this experiment, the experimenter should attempt to gain control over as many aspects as possible. The first aspect of control is that the population should be randomly assigned. This will ensure that all possible confounding variables are eliminated. These confounding variables include differences in the level of vocabulary that the students possess, etc. Another aspect of control is that there is only one teacher. Because there is only one teacher, confounding variables such a different styles and personalities will be eliminated because all the students will have the same teacher. Another aspect that should be controlled in this experiment is the language and the level of the language being taught. By teaching the same level of language in four classes, the differences between the classes will not be because some students are more experienced or were exposed more than others. This will increase internal validity. In addition, the way these courses are taught should be standardized so that all students are taught the same information in the same way, and are given the same tests. The only thing that should vary is the independent variable. By doing so, the experiment’s internal validity will increase because any variation between the control and experimental groups can be attributed the independent variable, which is the use of programmed learning material. An