Analyse data
The applet provides a number of graphs and regression procedures. There are also a few procedures for transforming data. The data window contains a spreadsheet and a menubar from which you select the appropriate statistical analysis. For instance if a logistic regression is selected from the menu, a dialog box is shown. Using the dialog box select a dichotomous response variable and one or more explanatory variables. Interactions are added to the model by selecting two or more variables in the left panel, and clicking the button add interaction". Model terms are removed from the analysis by clicking the remove button. Click "done" to start the analysis.
File menu
Data files
Opens a text file with example data, the contents of this file are displayed in the datasheet.
Clear
Clears the contents of the spreadsheet
Quit
Closes this program
Edit menu
Copy
Copies a value from the spreadsheet.
Cut
Removes a value from the spreadsheet
Paste
Pastes a value in the cell in the spreadsheet
Data menu
Calculate
Creates new variables by calculating sums, products,
quotients, and differences of two variables.
Transform
Calculates a transformation, such as the logarithm, or square root of the variables.
Invalid values are replaced by missing values. The following Mathematical functions are available.
- ln. Returns the natural logarithm of a nonnegative number.
- exp. Returns the value of e raised to a power. For example, exp(1) returns 2.71828.
- sin. Returns the sine of an angle.
- square root. Returns the square root of a nonnegative number.
- abs. Returns the absolute value of a number.
- cos. Returns the cosine of an angle.
- tan. Returns the tangent of an angle.
- log10. Returns the logarithm to the base 10.
Datatype
Changes the datatype of a variable. The default datatype
for a variable is numeric. Use the menu to change the
datatype to factor. Parameters for factor variables are calculated using the cornerpoint parameterisation. This means that for a factor with n levels the first level is set to zero. Using this parameterisation, the value of the first level is included with the constant term. The estimates for the levels 2,¼,n describe the deviation from the first level. In words, the first level has the role of a baseline measurement.
Delete
Deletes a variable or observation. This operation cannot be undone.