This topic has the following purposes:
Introduces estimation procedures in i-Regand with OLS as an example.
Carry out diagnostic tests
Revise and re-estimate the model to obtain the best model
Send and share regression output
By following each step in this topic, you are expected to be familiar with menus in i-Regand. You will find it very easy and convenient to run any regression methods in this apps. But first, we will introduce you to navigation buttons.
There are only seven navigation buttons that you have to be familiar with. These buttons will enable you to move from one screen to another. Just try one by one and follow the instructions, you will know what they are. It will not harm your smartphone. The buttons are:
Add a new estimation to the current project sheet
Create a new project sheet
Auxiliary menu (right hand side), carry out post estimation tests or seek help.
Send and share results via email, WhatsApp, Shareit, or other apps.
Save output to a new spool or append to an existing spool
You can find these navigation buttons
Add an estimation equation
A project sheet consists of variable list, estimation list, and output spools. Variable list is where you save any variable needed in the current or future estimation. Estimation list is a record of estimation procedure you have done and needed in the future. Output spools accumulate whatever estimation you need to save. The spools may be shared or sent via email, WhatsApp, Line, Bluetooth, etc.
It is always advisable to save your work on a project sheet. You will find it very convenient especially when you need to repeat or modify your previous projects. Just recall a project sheet and run the regression.
You can save or delete variables to or from variable list. Use the auxiliary menu right to the variable, if you want to delete it.
The i-Regand also allows you to save and add estimation list. Just tap on add icon, select the estimation procedure (OLS etc), specify the name, and run. You can recall back the estimation later on for modifications or repeating the estimation.
Output spools facilitate you to accumulate various estimation results in one or few out files. Every time you do the estimation, you can add the results into a spool without discarding the old ones. Once the results are displayed, touch the save icon, specify the name of destination spool, and save. To display what is the content of a spool, just double tap on a spool name, and the output spool will be displayed. To share and send a spool, tap on share icon, select the medium (email, Bluetooth, etc), and send.
We will analyze the determinant of commuting time of Bill Griffiths, a professor of econometrics at Melbourne University. The data, commute.csv, contains 231 observations with four variables TRIPTIME, DEPART, REDS, and TRAINS. Here is the story.
Every morning between 6.30 and 8.00 AM, Bill goes from his house at Carnegie to his office at Melbourne University. TRIPTIME measures how many minutes he spends driving his car each trip. He postulates that the amount of time depends on the time he departs from his house (DEPART), the red light he encounters each time he commutes (REDS) and the number of trains that he has to wait for at the Murrumbeena level crossing (TRAINS). He himself recorded these data for 231 working days in 2006.
Our task is to run and compare two following regression models:
We have to decide which model is better and statistically satisfactory.
Creating a project sheet and retrieving the data
The file, commute.csv, is located somewhere in your smartphone. The apps will automatically identify all CSV (or Excel and text) files, no matter where you place it. If your intended file does not show up, just press the refresh button.
To utilize the data, first you need to create a new project sheet. Here are the steps for creating a project sheet and retrieving the data:
Step 1. Tap on the create project file button (bottom right side).
Step 2. The screen will ask you to fefine the data type: time series, cross-section, or panel. Press time series. You will see different response from the apps when you press other data types. Just try it.