Introduction
Correlation determines the relationship that exists between two variables through providing evidence. Regression, on the other hand, enables the researcher to understand that extent to which the change in the independent variable is caused by the alteration of the dependent variable. The scholarly article, comprised of the Lancet, the British Medical Journals, and the New England Journal of Medicine, which was published in 1997 is related to correlation and bivariate regression in the authors’ analysis of data and plots.
Common Errors
A lot of errors are present in the correlation analysis within the three medical journals. The authors did not show the clear-cut difference between correlation and regression and often failed to define the type of regression line drawn through an array. Besides, the inappropriate use of correlation coefficients was explicitly expressed in the article. The expected 95% confidence interval and a clear indication of the number of cases, when using and citing correlation coefficient, were not given in the article (Gardner, 1986). The authors also attached undue importance to significant outcomes in the correlation context while explanation and justification of the outliers in plots and computations were omitted. Furthermore, an appearance of heteroscedasticity in plots without compatible comments in the text was also a common error.
Delegate your assignment to our experts and they will do the rest.
Assessment
The authors used correlation or bivariate regression to determine the level of dependency between the variables that included risk factors, age, health status, and the income of each population. The use of correlation or bivariate regression was the appropriate choice because testing the existence and strength of the linear relationship between two variables is required in such analysis (Sweeney et al., 1998).
Conclusion
The authors displayed the data of the research findings and constructed tables that are comprehensible. The presented scatterplot provided the clear visual relationship between the variables, which made the judgment of the strength, linearity and the outliers easier.
References
Gardner, M. J., & Altman, D. G. (1986). Confidence intervals rather than p values: estimation rather than hypothesis testing. BMJ , pp. 292:746-50.
Sweeney, K. J., MacAuley, D., & Gray, D. P. (1998). Personal significance: the third dimension. Lancet , pp. 351 :134-6.
