Introduction
The absence of an independent variable in research should be viewed as a cause of concern. The point is, any quasi-experimental study has to involve the manipulation of one or more independent variables that already exist (such as race, gender, age). These variables may be either categorical or continuous. Even though research lacks randomization and control, it is still aimed at identifying a causal relationship between an explanatory variable and a response variable. The fact that the given scientific work does not have an independent variable call into question not only the design but also the feasibility of the study.
Main text
Spearman’s correlation is a commonly used method for the determination of the strength and direction of the association between variables. This approach assumes that variables are ordinal, and the values of variables are monotonically related to each other. In general, Spearman’s correlation is a good tool for analyzing whether the correlation between the data exists. However, several factors need to be taken into account when reflecting on whether this method is appropriate for particular research. Firstly, the variables must be ordinal (categorical ones which, though, have numerical values and may thus be ordered). This very assumption leads to two important concerns that make the whole study rather doubtful. One issue is related to the usage of solely ordinal variables in the quasi-experimental research study that needs to rely on the raw data. Another issue considers the applicability of the findings of this study.
Secondly, it is worth mentioning that the conduction of the correlational analysis may not allow for answering the research question. After Spearman’s rank correlation coefficient has been determined, one needs to measure the p-value to accept or reject the null hypothesis (Moore, Notz, & Fligner, 2015). This procedure will help identify whether Spearman’s rank correlation coefficient is statistically significant. Thirdly, Spearman’s correlation is a non-parametric test, which means that it was designed for samples that have non-normal distribution. However, given that the sample of the discussed research study has a normal distribution, it appears that the use of Spearman’s correlation is not justified. The normal distribution is a statistically desirable one, and a more accurate Pearson correlation coefficient needs to be used to measure the linear relationship between two variables. In the given case, this is a more appropriate type of correlational analysis for the determination of a statistically significant relationship.
Nevertheless, it is crucial to take into account some limitations of the correlational analysis. The correlation does not imply causation, which means that changes in one variable may not necessarily cause changes in another variable. Even if there is a strong relationship between two variables, this does not mean that one of them causes another. Given that research is aimed at identifying the association between two dependent variables, it is unclear if the results obtained could apply to healthcare practice. Another drawback of the correlational analysis is the impossibility of going beyond the data that is given.
It needs to be stated that in statistics, association and correlation are used interchangeably, though the former is a more generalized term. Association analysis helps uncover any kind of relationship between two or more variables or bivariate data. A collection of association rules is applied to express any data relationships (Johnson & Bhattacharyya, 2019). One could note that this method is often utilized to discover any interesting patterns in large data sets. However, it does not consider that some of the connections may be spurious since they can happen simply by chance. Even though association analysis has several application domains, such as market research and bioinformatics, correlational analysis is the main tool in scientific research.
The correlational analysis measures a specific form of association, in particular, a linear one. Results obtained from this method (such as correlation coefficients) can be used to explain whether there is a strong or weak and negative or positive association between variables. The value of the correlation coefficient describes how changes in one variable affect the other variable within the whole data set. It is possible to say that in contrast to association analysis, the correlational analysis makes statistical inferences about the connections between variables. In other words, this type of analysis may be used to make predictions about the data behavior. Therefore, one could state that association analysis and correlational analysis are not equivalent in determining relationships between variables.
Conclusion
All the findings regarding the methodology of the described research study discussed above impact one’s decision about whether to use the obtained evidence to inform practice change. This is because they undermine the usability and reliability of the given quasi-experimental study and the results it yields. Based on the scenario provided, there are several important reasons why this research should not be used for the further implementation of a change. Firstly, the study does not have an independent variable, so it is unclear how the quasi-experiment was conducted. Secondly, an inappropriate level of correlational analysis was chosen to identify the association
between variables. Thirdly, no statistical inference regarding the causal relationship can be made solely on the basis of the correlation coefficient, so the study’s conclusion cannot be utilized to support a practice change.
References
- Johnson, R. A., & Bhattacharyya, G. K. (2019). Statistics: Principles and methods (8th ed.). Hoboken, NJ: John Wiley & Sons.
- Moore, D. S., Notz, W. I., & Fligner, M. A. (2015). The basic practice of statistics (7th ed.). New York, NY: W. H. Freeman & Company.