A scatter chart is a graphical representation of data in the form of points which represent 2 variables. It is common for scatter charts to also be referred to as scatter graphs or scatter plots. Scatter charts distribute points along the X axis and Y axis in accordance with their values and their nature, i.e., dependent or independent.
Scatter charts are employed when the user would like to analyze numerical data to determine if there is a relationship between pairs. If the points fall on a curve or line, a relationship is revealed. More specific cases of usage are listed below for review:
- Analysis of paired numerical data
- Analysis of dependent variables that appear to have multiple values for each independent variable
- Analysis of cause and effect
- Analysis of effects that appear to have a cause in common
The general process for creating a scatter chart is simple. First, gather the data pairs that are suspected to have a relationship. Create a graph with the dependent variable on the y axis, and the independent variable on the x axis. Identify pairs by marking the plots which intersect, and also place plots side by side in an easily identifiable way. If there is a noticeable pattern, meaning a curve or line, you can stop and then perform correlation analysis; if not, you would then divide the graph into 4 quadrants. The procedure for dividing the graph is presented below for review:
- Draw a horizontal line after counting x/2 points from the top of the graph to its bottom.
- Draw a vertical line after counting x/2 points from the left side of the graph to its right
If the quantity of points is odd, find the middle point, and draw a line through it. Omit the points found on a line, and count the total points within each quadrant. Add the totals of the quadrants which are diagonally opposite. Find the sum of all points, and find the smaller sum of opposing quadrants. Finally, utilize a trend test table to find the limit for N.