Nursing: Evidence-Based Practice

Description:

Reflect on the research articles you have reviewed to date that relate to your clinical question. Please refer to all the unit resources (e.g., readings, recordings, announcements, etc.) to facilitate this process. With these thoughts in mind, post your thoughts on the following:

In the articles you have reviewed, what statistical tests are being used for data analysis?
Are the statistical tests being used appropriately for the type of data being produced and the question being asked?
Do the authors appropriately indicate implications of the data analysis, or do they overstretch?
In the articles that have I reviewed listed below, I need to know what statistical tests are being used for data analysis?

· A key study for my PICOT project is titled “How Accurate are Self-Reports? An Analysis of Self-Reported Healthcare Utilization and Absence When Compared to Administrative Data” (Koerner & Zhang, 2017). The authors evaluated the data using Pearson’s correlation and multivariate logistic regression models.

· Correlation is a statistical method used to assess a possible linear association between two continuous variables, while Pearson’s correlation coefficient is a test statistics measuring the statistical relationship, or association, between two continuous variables (Mahan, Clark, Anderson, Koller, & Gates, 2017).. Pearson’s correlation coefficient (r) is a measure of the strength of the association between the two variables; it is the best method of measuring the association between variables of interest because it is based on the method of covariance, and covariance is a measure of the joint variability of two random variables (Koerner et al., 2017). If the data is normally distributed, in other words, bell-shaped, and symmetrical about the mean, tests such as the histogram, and normality tests may be used to evaluate if the distribution is normal (McCrum-Gardner, 2008).

· Are the statistical tests being used appropriately for the type of data being produced and the question being asked?

· The majority of studies may be evaluated with the proper evaluation of over 100 statistical tests to choose from. Tests are chosen depending on the type of research questions being asked, the type of data being analyzed, and the number of data sets or groups being evaluated.

· The first question to be asked is what scale of measurement the data will be, nominal, ordinal, or interval. The study in question involves interval data. The second question is the reason for the analysis; as an example, paired groups or independent groups?

· Yes, the statistical tests are appropriate for the type of data being produced, and for the type of questions being asked.

· Do the authors appropriately indicate implications of the data analysis, or do they overstretch?

· Based on our readings and videos, I saw no evidence of data-stretching. For example, data was collected over two calendar years, a lengthy, non-selective inclusion period. A potential confounding factor was that all participants had to have been enrolled in the company health insurance program a minimum of 320 days as a condition of inclusion in the study. Another potential confounding factor was that only hourly employees were included in the two-year study. In addition, pregnant employees were excluded from the study, as these employees are atypical of the general employee population in that their healthcare utilization would be temporarily altered. These factors may have influenced the results, by excluding participants with different reporting habits. After recognizing these potential limitations, their findings mirrored multiple studies previously conducted.

· The authors also noted potential improvements for future studies, including inclusion of a more diverse employee population, and examining why there exists a difference in self-reporting based on participant sex.