Advantages and disadvantages of quasi-experiment
The quasi-experimental design has several advantages and disadvantages, which are worth noting. According to Jackson (2012), one of the main advantages of the quasi-experimental design is that it is logically easy to manage. A researcher will find it easy to conduct a quasi-experiment because of this fact. Another advantage of this design is that generalization is always possible when using a control group. It means that a researcher can easily use a small control group to generalize the entire population. Despite the above advantages, quasi-experiment has several disadvantages. One of the main disadvantages of this experimental design is that it has a limited ability to compare studies because of the possible variations that may exist. Another major disadvantage of using this method is that it has questionable validity. As Pelham and Blanton (2012) note, many researchers do not trust the validity presented by this design because of the small number of participants that are always used, among other reasons.
According to McBurney and White (2010), the fundamental weakness of the quasi-experimental design is that it lacks random assignment that is common in actual experiments. Lack of random assignments is considered a weakness because it reduces the internal validity of an experiment. This weakness matters because with the internal validity affected, making a causal claim becomes difficult to even after the completion of the research. It is for this reason that some researchers avoid it when conducting research that requires a high degree of both internal and external validity.
Assigning participants into groups randomly
If a researcher assigns the participants into various groups, it is always assumed that at the beginning of the study, the groups have the same capacity. However, this assumption is wrong because the groups will have different individuals with different capacities. It is also a fact that once the study comes to an end, the groups will have several differences. The capacity of a group will depend on the capacity of the individual members of the group and how well they work as a team to achieve a common goal. In such cases where it is not possible to assume equivalence at the beginning or end of a study for the groups involved, a researcher may exercise tight control of the experiment in all the teams. In this case, the researcher will ensure that the behavioral patterns of the groups in terms of discussions, or any other relevant issue, are universal.
Presence of a threat in a given study
It is always important to try and avoid the major weaknesses or threats in a given study. However, there are cases where a researcher fails to identify these threats to a time when the study is completed. The outcome of a study may suggest that a given threat might have been experienced. For example, a major deviation of the outcomes from that of the existing literature may suggest a given problem in the study. As Jackson (2012) notes, a deviation of the outcomes from that of the existing literature may not necessarily mean that the research is invalid. However, if the variation is significant, and there is no proper justification proving why the previous studies were wrong, then this may be the strongest indication of validity threat in the study.
Types of designs
Non-equivalent control group, pretest only design is one of the most commonly used research designs in social sciences (McBurney & White, 2010). This research design involves conducting a study on two groups or individuals who are believed to share specific characteristics. It is important to note that one outstanding feature of this study is that it is not randomized as other studies. In this design, the researcher will choose two individuals or groups that he feels have similar characteristics. It is important to note that this design does not effectively address the threat to validity, as discussed by Trochim and Donnelly (2008) in chapter seven of their book.
Non-equivalent control group pretest-posttest involves using two groups or individuals where one receives treatment, and the other (control group) does not. At the end of the study, the two groups are tested to determine if there is a significant difference between them. It is important to note that the design may be affected by several threats to its validity.
McBurney and White (2010) define a cross-sectional study as a research method that “Utilizes different groups of people who differ in the variable of interest but share other characteristics” (p. 64). The groups may share several demographical characteristics, such as their social status, levels of education, among other factors. These researchers note that cross-sectional researches are always observational. They take the form of descriptive research. One of the biggest threats that this design face is that sometimes the participants may change their normal behavior during the study, which would result in the collection of wrong data.
According to McBurney and White (2010), Regression-discontinuity study design is an approach of study that is used to determine how effective a given intervention program is when addressing a given problem. In this research design, the participants are selected based on a predetermined cut-off score. For instance, a researcher may want to determine the effectiveness of a program in elevating poverty among the people living in absolute poverty. In that case, the researcher will have a given level of income upon which one will be disqualified from being part of the study. The validity of this method is based on the ability of the researcher to identify the participants who rightfully belong to a given desired group.
Reasons why quasi-experimental designs are used more often
According to Pelham and Blanton (2012), in many studies, quasi-experimental designs are often preferred for experimental designs. This may be attributed to several reasons. One of the main reasons why this design is more common is that it is easy to manage. Researchers find it easy to conduct a quasi-experiment, especially when they have limited time to present their study. Another reason that makes this design very popular among social scientists is that generalization is always possible when one is using a control group. Based on the results obtained from the control group, a researcher may conclude a given group.
Conclusion
The analysis above demonstrates that when conducting a study, a researcher can take different approaches. It is clear that when conducting a study, a researcher will always need to choose the best design based on the nature of the study. Different research designs will be appropriate for different designs. The most important thing that a researcher should do is to select a design that meets the expectations of the study. As Pelham and Blanton (2012) argue, the validity and reliability of research do not only depend on the participants chosen, but also the design was chosen for the study. The following research question can be used with a single-group post-test only design:
What is the impact of small class sizes on the quality of learning in our local schools?
This is a typical example of a question that can be used when handling a single-group post-test design.
Quasi-Experimental Designs Part II
Research questions in the study
The study by Goldberg (1990) seeks to address the following research question:
What is the effectiveness of television advertisements which are targeted to children?
Goldberg’s rationale for the study
In this study, Goldberg (1990) wanted to determine if television commercials have any impact on children. This was after a Quebec law eliminated all the adverts targeted to children on Quebec television stations. This law meant that English-speaking children could get toy advertisements on American television commercials, which were predominantly English. On the other hand, French-speaking children living in Quebec were locked out of television commercials. The study was focused on determining how well English speaking children knew about the toys as compared to the French-speaking children. Based on the approach taken in the study, it is apparent that the study was designed to contribute to theory. It would be right to say that the results of the study contribute to the theory. The researcher not only confirmed that children exposed to advertisements had better knowledge of the toys, but also explained the relevance of exposure to cognitive learning and development.
Construct addressed in the study
In this study, the underlying theme is that children learn a lot from what they are exposed to during their development. The study proves that media has a major role in defining the kind of knowledge that children acquire when they develop. To put the constructs into operation, the researcher took advantage of the existing law that prohibited Quebec television stations from advertising to children. The law did not affect the English speaking American television stations. It meant that English speaking children in Quebec could still have access to the advertisements. This was not the case for French-speaking children. This way, it was easy to determine the relevance of media on children’s knowledge development.
Independent and dependent variables in the study
As Trochim and Donnelly (2008) note, in every study, it is important to define the independent and dependent variables clearly. In this research, the independent variable was having access to television commercials on toys. The dependent variable was the knowledge that children have about toys. This knowledge included the types of toys, how to use them, where to get them, their prices, and the value they offer to their users. All these dependent variables largely depended on the independent variable of having access to television commercials.
Type of research design used in the study
The researcher used a non-equivalent control group pretest-posttest. This is so because the researcher used two groups that were relatively at the same level of knowledge before the experiment. Still, after one was exposed to a given environment, a difference emerged.
Internal and external threats to validity
It is clear from this study that the researcher was keen to eliminate threats to external and internal validity to enhance its reliability. The internal validity of the study was enhanced by using the same person to collect data from all the participants. The interviewer used questionnaires to collect data from the children with the help of their former teacher, who they all knew. The teacher could speak both English and French. The research also used the same questions for the two sets of students but in different languages spoken by the participants (French and English). External validity was addressed in the selection of the participants. One threat that was not addressed in this study was the fact that some French-speaking children could access information about toys from other sources, making them as knowledgeable about the toys as the English-speaking children. This would be a threat to the interpretations because it would be easy to conclude that the advertisements do not affect while, in essence, they have major effects on children’s knowledge.
The researcher believes that the conclusions made by Goldberg (1990) are convincing. The sources used by the scholar to conclude are reputable. The research design used also is also appropriate for the research. This makes it easy to trust the findings and conclusions made in this study.
References
Goldberg, M. E. (1990). A quasi-experiment assessing the effectiveness of TV advertising directed to children. Journal of Marketing Research, 27(4), 445-453.
Jackson, S. L. (2012). Research methods and statistics: A critical thinking approach. Belmont: Wadsworth Cengage Learning.
McBurney, D., & White, T. L. (2010). Research methods. Belmont, CA: Wadsworth Cengage Learning.
Pelham, B. W., & Blanton, H. (2012). Conducting research in psychology: Measuring the weight of smoke. New York: Wadsworth.
Trochim, W. & Donnelly, J. P. (2008). The research methods knowledge base. Mason: Thomson Custom.