# Regression: which parameters to use and how to plot the data

• I
• ducmod
In summary, Ducmod is learning about finance and regression, and wants to replicate the analysis in a file that he downloaded from a website. He does not understand which columns are used to perform a regression model. He has 3 independent variables (lagged DPS, annual EPS, and change in EPS) and would like to find predicted dividends.f

#### ducmod

Hello!
I am yet very weak in statistics, but I am learning some basic finance, and this requires to create regression.
Please, take a look at attached files - one excel that contains the results of regression and one screen shot of the window of StatPlus that I have to fill in. Before using my own data I would like to replicate the analysis presented in the file. I don't understand which columns are used to perform a regression model. There is part of data that computes variables for regression model, and it contains 4 columns. The goal is to find predicted dividends. So I assume that dividends per share is the dependent variable. But I still don't understand which columns I should use and how.
I will be very grateful for your help!
Thank you very much!
Edit: corrected the excel file, and added my results.

#### Attachments

• Screen Shot 2016-05-17 at 6.38.05 PM.png
34 KB · Views: 657
• regression_model_test.xlsx
53 KB · Views: 316
Last edited:
Hi Ducmod,
In your attached Excel sheet, the columns used for the regression are clearly marked. Columns G-J are titled "Data for Regression Model".

The model was described by:
##D_t = a + bD_{t-1}+cEPS_t+d(\Delta EPS)##
You are correct that the Dividend, ##D_t## is your dependent variable, and you have 3 independent variables for this model, ##D_{t-1}## is called the "lagged DPS", ##EPS_t## is the "annual EPS", and ##\Delta EPS## is the "change in EPS".

The coefficients for the model, a, b, c, and d, are listed in that order in your output table.

Hi Ducmod,
In your attached Excel sheet, the columns used for the regression are clearly marked. Columns G-J are titled "Data for Regression Model".

The model was described by:
##D_t = a + bD_{t-1}+cEPS_t+d(\Delta EPS)##
You are correct that the Dividend, ##D_t## is your dependent variable, and you have 3 independent variables for this model, ##D_{t-1}## is called the "lagged DPS", ##EPS_t## is the "annual EPS", and ##\Delta EPS## is the "change in EPS".

The coefficients for the model, a, b, c, and d, are listed in that order in your output table.