The scatter plot gives you the opportunity to investigate the connection between two variables.
After choosing this option you are first asked to select a horizontal and a vertical variable. Then a screen with the scatter plot, and several options, opens.
In tab pages you can choose between four regression models: linear, quadratic, exponential and power. Each page will display the formula of the model and the corresponding statistics.
On the right-han side of the screen there are several buttons for more options.
Variables (horizontal and vertical)
With these buttons you can select other variables.
Divide
With the Divide option you can divide the data into subgroups. For example, you can divide a file of students into a group of males and a group of females.
Then you can choose between having one scatter plot, with different colors for the subgroups; or several scatter plots, one for each subgroup.
All model information is now presented by subgroup
Title for scatter plot
With this option you can create a title above the graph. This is usually a good thing to do.
Tests (only for undivided data)
There are two tests in this program.
The first is a test on the regression coefficients. If the errors in the measurements are independent and normally distributed with equal variances, then the estimates of these parameters have t-distribution.
The degrees of freedom depend on the situation. In the linear and the exponential model this is the number of pairs minus 2. In the quadratic model this is the number of pairs minus 3. The most important hypothesis to investigate is whether a parameter can be zero.
The second test is to see whether the whole model is a good model. The null hypothesis is that all coefficients are zero. The alternative hypothesis is that at least one of the coefficients is not zero.
Here you will find the estimates and the accompanying ANOVA TABLE.
Regression curve
The regression curve is plotted
Bands
The confidence curve drawn in green shows the confidence interval for the mean/expected value of “Y”for each value of “X”
The confidence curve drawn in red show the confidence interval for the predicted value of “Y”for each value of “X”.
The level of the confidence interval can be entered and the default value is 0.95.
Example for the case of regressing weight in relation to height.
The green confidence interval is an estimate of the average weight of people of a given height. The red confidence interval is an estimate of the possible weight of one person, chosen at random, of that height. Our estimate of the weight of one person of a particular height is going to be more uncertain than our estimate of the average weight of people of that height.
Scatter plot or Residual plot
Either the scatter plot or the residual plot is shown’
Residual
Residual is the difference between the observation and the fit.
Buttons
With the print button you obtain a print option and options to copy the graph to the clipboard, send to Word, or save as file.
Offers help on the item currently being processed.
You leave the active window and return to the previous window.
The graph is immediately pasted into Word.