Quasi-Experimental Research Design

This research study employs quasi-experimental research design; this is based on the objectives of the research study which I will say is justified because the researchers intended to utilize time series in measuring the causal effect resulting from marriage initiatives across the 50 States that the study is focused on. Thus, because the sample cases for this research study cannot be randomized and involves time series, then quasi-experimental is the most appropriate research design to use.

However, looking at the study design there are two issues to critique; one is the inability of the study to control against all confounding factors. In this case for instance the researchers uses race and economic factors to control against the variable of poverty, but we know there are other factors that can confound poverty such as education level, discrimination and other factors which are not controlled in this study. In the same way the study has also not completely controlled against the prevalence rate for divorce.

Regarding the suitability and accuracy of the sample size used in this study, I would say this does not apply as the research utilizes the whole study population which means that sampling in this case is not necessary. This inability to randomly select study cases is in fact one of the characteristics in quasi-experimental research designs as we have seen.

Even on the 11 States that were picked from the list of the total 50 states which the authors describe as “high-activity states” no randomization strategy was used but rather the States with high rates of marriage initiatives were the ones used (Kickham and Ford, 2009). All the same, because this research study is designed to utilize the whole population of the study then the implication is that the research findings are more accurate since no generalization is necessary.

My Research Plan

My research topic is “A research on the impact that video games playing by young students (aged 7-15 years) has on their academic performance and level of interaction with their parents”. The research objectives for this study are two;

  1. To determine the impact that video game playing has on the academic performance of young students aged 7 to 15 years.
  2. To determine the impact that video game playing by young students has on their level of parental interactions and bonding.

Hypothesis: The amount of time spent by young students (aged 7-15 years) is directly correlated with their academic performance in school and level of parental bonding at home.

To research this relationship I propose to use a case-control study design in which the findings of the study cases are controlled by the data obtained from a similar group that does not have the characteristic of interest which in this case is the activity of video game playing or for that case amount of time spent on playing videos. This is because the case-control study will be able to identify if there is a causal-effect relationship between the two groups of young students that are being compared based on the variables identified which are academic performance and the parental level of interaction.

The sample size used for this research study will be based on the total population of students that the study intends to generalize its findings, this figure can be obtained from the “t test for Two Independent Samples” once the effect size is estimated if we were to assume that alpha is 0.05 and the statistical power is 0.80 (Kovacs, 1985). In determining what cases are selected for each group I suggest simple random sampling be used.

Study population

Since this research study seeks to determine the impact that video game playing has on the academic performance and their level of parental interactions of young students aged 7 to 15 years, then the study population for this research study will be school going children within a certain locality (town y) that falls within this age bracket. Thus, the inclusion criteria for this research study will be two factors; young persons aged 7-15 years, who are enrolled in school and who reside within the specified locality, town y.

It is from this population that the sample of study, comprising of the two groups (case and controls) will be pooled from, which will make up the actual sample size that would be studied during the research study. Because the research study also intends to investigate the level of parental interaction between the subjects of study from the two groups, then their parents will also be included in the research study. This is necessary so as to measure the level of interaction that exist between the selected cases and their parents; to control for single parent families the study will require the assessment questionnaire to be filed by just one parent who the student feels is most close to them. The assumption that we make for this study is that all students have at least one parent living with them.

Sampling

A sample refers to a subset of a particular population size; it is the proportion of population that is usually studied by a researcher since it is impossible to study the whole population due to time and costs resource limitations (Kovacs, 1985). Since we have already determined our population of interest to be young students enrolled in school aged 7-15 years residing within town y, all that we now need to do is come up with a method that objectively selects required cases from this population.

The type of sampling method that is used must be based on a scientific theory backed with a rationale of choosing that particular method which should be relevant to the type of research study (Adams and Schvaneveldt, 1985). Therefore for this research study I propose simple random sampling because it is a technique that ensures there is no bias in selecting cases from the population of study (Leedy, 1993). At the end of this process I intend to have finished selecting the students that I will base this research on, who will be paired with one of their parents.

Simple random sampling is a sampling technique that relies on a simple probability method of choosing desired sample size by random selection (Mugo, 2010). Ideally, simple random sampling involves assigning all cases of the population with numbers; once this is done the required study cases are then determined. The study population will be the sample size which is randomly picked by selecting cases from a list of numbers that were assigned to the population and then matching these selected numbers to specific cases that would make up the sample population (Newman, 1994).

In this case the process of selecting samples will be done twice; first sample for the case group who in this case comprise students who play video games and another category of students, control group whom in this case are students with exactly the same attributes but who do not play video games will be selected. The population from which these two groups of students are derived from will thus be differentiated based on this particular variable.

For instance, in order to select 200 cases from a population of let’s say 2000 young learners enrolled in school, in town y aged 7-15 years who plays video games, will require assigning numerical numbers to all the cases as the first step. The second step is randomly selecting 200 numbers from the total list of 2000 numbers which represent the whole population, probably by blindly picking tags with numbers from a basket.

The last step is matching the chosen 200 numbers with the population cases which will now be our sample size. This procedure will be repeated all over again when selecting sample cases of the control group who in this case are students who do not play video games; this is how randomization is achieved in a research study since the sample size becomes a true representation of the larger population of the study and consequently the research findings of the study can be generalized for the whole study population.

The eligibility criteria used for selecting sample size is the same with that required for the whole population, this means that as long as a person is included in the study population one has equal chances of being chosen as a sample case. The major advantage of simple random sampling is that selected cases reflect the actual characteristics of the target population since the selection of sample is done randomly (Babbie, 1992).

Another advantage of choosing a sample using simple random method is that all cases have equal chances of being chosen as samples (Babbie, 1992). Finally, it is a relatively simple and straight forward method of randomly selecting samples quickly from a given population (Babbie, 1992).

References

Adams, G. & Schvaneveldt, K. (1985). Understanding Research Methods. New York; Longman Inc.

Babbie, E. (1992). The Practice of Social Research. California; Wadswork Publishing Company.

Kovacs, A. (1985). The Research Process: Essential of Skill Development. Philadelphia. Davis Company.

Kickham, K. & Ford, D. (2009). Are State Marriage Initiatives Having an Effect? An Initial Exploration of the Impact on Divorce and Childhood Poverty Rates. Public Administration Review, 1: 846-856.

Leedy, P. (1993). Practical Research: Planning and Design. New York; Macmillan Publishing Company.

Mugo, F. (2010). Sampling in Research. Web.

Newman, L. (1994). Social Research Methods. Boston; Allys & Bacons.