Correlation is referred to as the association between bivariate data, which is also described as data sets containing two observations. Scatterplots, also known as scatter graphs or correlational charts, provide excellent descriptive representations of the interrelationship between two quantitative variables. Each point in a scatter plot denotes a paired measurement of two variables for a particular subject, and every subject is represented by one point on the graph (Brase & Brase, 2015). Therefore, by noting the sequential position of the points throughout the chart, an individual can determine the direction and strength of the relationship. In regards to direction, when the plots produce a lower-left to upper right pattern, it can be concluded that there is a positive correlation between the two variables. Conversely, an upper-left to lower-right pattern suggests the existence of a negative correlation. Furthermore, when the plots lie on a straight line, a perfect correlation can be insinuated. Lastly, when the points are scattered randomly and do not show a linear trend, a zero correlation can be suggested.
When analyzing scatterplots, it is crucial not only to consider the direction of the relationship, which is negative, positive, or zero but also the magnitude of the correlation. The distance between individual points can represent this (Brase & Brase, 2015). The concept of drawing an imaginary oval is often used to help interpret the magnitude of the collinearity. A strong correlation between variables is represented by the presence of points that are close to one another and a small imaginary oval. On the other hand, a weak correlation is signified by large distances between points and a wide imaginary oval. In summation, scatter plots provide a good visual representation of the direction and magnitude of linear correlation between two quantitative random variables.
Reference
Brase, H., & Brase, P. (2015). Understanding basic statistics (7th ed.). Boston, MA: Cengage Learning.