Sampling

Answer

Deciding on Appropriate Research Design

Upon formulating a hypothesis, a researcher’s next significant task is selecting the most appropriate study design for the research.  A researcher should always have it at the back of his/her mind that research design is far more important compared to analysis of research findings.  Once a wrong or inappropriate design is used, the validity of the research gets compromised.  This essay identifies and appropriate study design for my research hypothesis and addresses several issues pertinent to the research design.

Research Question

Students express varying views about different courses they undertake at the university. Biostatistics causes anxiety among many students for various reasons.  Therefore, one may want to determine what the views of senior students are towards the course a given university and whether these views are dependent on gender. With this in mind, my research will be answering the following research question: 

“Do perceptions about difficulty in biostatistics course vary by gender?” 

Hypotheses

Gender does not predict perceptions about difficulty of biostatistics course.

Cross sectional design is appropriate for this research question.

Components of the Design

As stated above, a cross sectional design would be appropriate for this research question. A cross sectional study is an observational study in which all the variables to be studied are measured at the same point in time, in a defined population (Pandis, 2014; Gordis, 2013). The researcher merely observes and measures variable of interest in the study subject without taking part in changing the variables (Metse et al., 2016; Pandis, 2014).  Cross sectional studies may be descriptive or analytic.  For this research, the analytic type is applicable because association between perceptions about difficulty in biostatistics and gender is examined.  Both variables are measured at the same point in time.

Would this design be considered experimental or observational? Why?

This design would be considered an observational study because there is no randomization of subjects or intervention and no researcher manipulation.  Data is obtained from subjects within their usual setup (Stineman et al., 2016). Observational studies resemble experiments because they concern treatments or interventions of exposure to certain variables and the effect these cause (Pandis, 2014).  The difference with experimental studies is that in experimental studies, the researcher plays a role in manipulation the independent variables of interest and ensuring that groups in the study are comparable through random assignment (Gordis, 2013).  In this research there will be no randomization of subjects or manipulation of the variables under study.  The variables will be measured from the students under their usual setting.

Why did you choose this design over others? Be sure to note all advantages and disadvantages of employing this design to address your particular research question.

The choice of a given research design is dependent on the research question, concerns regarding validity, ethical and practical considerations (Spencer et al., 2014).  A cross sectional design is chosen over an experimental design for many reasons.  Relative ease and quick in implementation, data on all variables are collected at the same time, the treatment variable may be harmful and it would be unethical to give it to human subjects (Stineman et al., 2016).  In some cases, the treatment or the independent variable under study may be beyond reach of experimental manipulation by the researcher (Straub et al., 2016). In this research, the biostatistics course is undertaken by all the senior students and the researcher cannot manipulate the views of the students towards the course.  Additionally it would be practically easy to measure the variable under study since data can be collected at one point in time.  While it is a relatively simple and inexpensive to implement research design, it also comes with a number of disadvantages.  The researcher’s inability to have control over factors other than the independent variable of interest, make this research design unable to predict causality (Stineman et al., 2016). It is not possible to determine that perceptions about difficulty of the biostatistics course are only influenced by gender which leads to difficulty in interpretation (Spencer et al., 2014).  It is also not easy to determine temporal relationships between the variables, because they are all measured at the same time (Gordis, 2013).  This design is also susceptible to bias as explain below.

To what types of bias is this design most vulnerable? Please explain.

Bias is a fault committed by the researcher in the conduct of a study, which leads to a false association between the independent variable and the dependent variable (Gordis, 2013). The important ones for cross sectional studies include selection bias and self-selection bias (Spencer, et al., 2014).  The process of sampling students for this study may not be random and representative of the entire student population, because the sample may conveniently include only students currently undertaking the biostatistics course.  Additionally, if the response rate is not adequate, the non-respondents may be the students who fail to attend classes when the course is taught, leading to self-selection bias (Gordis, 2013; Pandis, 2014).

Cross sectional studies are also prone to information bias due to self-reporting where respondents tend to provide what they feel to be socially acceptable answers rather than the truth (Gordis, 2013; Spencer et al., 2014).  In this research, depending on several factors, students may think it is favorable to be seen to have positive perceptions about the course.  These scenarios usually result in bias of the measures, in this case, the perceptions about difficulty in biostatistics, and gender, if one type of gender systematically does not respond to the questionnaire.(2017) study employed, there are very little room for generalizing the study’s findings.