Workshop on MegaStat, R and R Commander for Master of Finance Students (MFin 604)

September 2, 2016 (9:00 - 12:45 and 1:30 - 5:15)

Both sessions will cover the same material

Room: DSB-505

 

Mahmut Parlar [Home page]

 

(DeGroote School of Business, McMaster University, Canada)

 

Cellphone use policy

 

We will start the workshop by first introducing MegaStat (an Excel add-in). We will then justify the use of R and Rcmdr in statistical analysis. These will be followed by covering a few examples of the most important topics. These introductory comments will be helpful when we start MFin 604.

(a) MegaStat (Installation and some examples)

 

In the first ten chapters of MFin 604 we will use the Excel add-in MegaStat 10.2 (available for Windows and Mac) that can be downloaded at the following link:

Publisher's web site

Installation is quite easy and it involves two steps. Here is the documentation (MegaStat_Installation.pdf) which is also available at the publisher's web site link.

 

Two other useful files from the publisher's site: (1) MegaStat_FAQ.pdf, (2) MegaStat_Users_Guide.pdf.

 

Here are a few examples of using MegaStat: (We will look at these problems in more detail later in class.)

 

a.1 Graphics

 

Actual GMAT  scores of entering DeGroote MBA classes (2003 to 2005). Let's look at the 2003 scores.

IQ Scores  of 500 individuals.

a.2 Hypothesis testing

DeGroote GMAT scores (2005) : In 2005, the mean GMAT scores of our entering students was approximately 630 with a standard deviation (s.d.) of approximately 45. Before the admissions are finalized, we have data for 25 newly admitted students with a sample average score of 625. Do we have a good reason to believe that the true mean (when admissions close) will be at least as much as in 2005? Shall we reject the null hypothesis H0: μ ≥ 630? We can solve this problem assuming that the s.d. for will still be 45. To do this, you will need to enter the following four items in the first column: Parameters; 625; 45; 25. Check MegaStat's Help file on Hypothesis Testing for further details.

 

(b) R, and R Commander (Installation and some examples)

 

Why R and R Commander?

 

Basics. Installation instructions, introduction to R and R Commander (Rcmdr) and R functions for basic statistical models

 

Regression. R functions for multiple regression and its ramifications