The purpose of correlation analysis is to determine whether there is a connection between the items under study. On the other hand, regression analysis basically investigates the impact of the independent variable X on the dependent variable Y. It’s important to note that the presence of correlation does not imply the existence of a regression relationship.
There are two ways to complete a regression analysis on the DiVoMiner platform:
- Under [Statistical Analysis] – [Basic Statistics], you can run a one-way linear regression as follows:
Drag and drop a categorical variable into the [Dimension] column, drag and drop two numerical variables into the [Numerical value] column, use “Summation” as the statistical method, select Scatter chart on the right side, check the box of [Unary regression analysis], select the regression analysis method, then the regression analysis of the variables can be calculated.
- The second way is to produce a multiple regression analysis in [Statistical Analysis] – [Advanced Statistics] by:
Click [Create a Calculation Task], choose the statistical methods ([Linear regression] or [Linary logarithmic regression]) base on your need. Fill in the “Task Name”, select the database, X-axis and Y-axis variables, you can specify a specific range of data in [Filter condition], confirm the number of data, click [OK] to execute the algorithm.
完成后，可点击模型右侧【查看数据】，查看数据结果。
