Bancova News and Announcements

How Much SAS do you need to be a professional SAS programmer?

Picture of John Zhang
How Much SAS do you need to be a professional SAS programmer?
by John Zhang - Monday, 9 November 2015, 8:01 PM

I raised the question because I see many students are working in the wrong directions.

Two wrong directions should be avoided:

1) certifications-oriented.

SAS certification is only useful if you take the opportunity as enhancing your knowledge and skills. If your goal is just for passing the exams, it's useless. People who do not know SAS at all can get a certificate. Professionals know this. Therefore, SAS certificates might be the least respected certificates among all computer languages.

2) knowledge coverage-oriented. Some students think there more they know about SAS, the more likely they can get a job. Wrong. To know something on the surface is easy. To make the knowledge as part of the knowledge and skills set is difficult. To master a language is like to drive a car: the more you use it on the daily basis, the more likely it become your own.

The right directions:

1) focus on projects. No matter how small the projects are, you should pick one. I use SAS to process my credit cards account records and bank statements. I also use SAS to find houses in the and the build models for the prices.

2) focus on the basic skills, not every thing. What are the basic skills then? Well, SAS base is the most basic one. In fact, you can do almost all data processing and important analyses by just using SAS base. You do not have to use proc sql to process data; you do not have to use ods to output SAS results.

As a stat major, the key to a successful SAS programmer is the programming "spirit" we need to develop, and that spirit is built by using a certain skills sets again, again, and again on a certain projects.

Picture of John Zhang
Bad Example: sliding on the surface
by John Zhang - Monday, 9 November 2015, 8:02 PM

Most SAS teaching at many universities are of the style "sliding on the surface".
It has the following features:
1) it teaches you how to find the right answers to your statistical problems using a SAS procedure or statement.
2) it uses SAS default outputs to present the results.
3) it assumes the data are ready for analysis without asking the basic questions about how the data come along and how as statisticians we should manipulate and process the data before we can use them.