Thanksgiving turkey, Blue Jays playoff baseball, and SAS ANALYTICS all in one day!
Thanksgiving is a wonderful holiday. The air is crisp, fall colours are at their peak, family and friends abound, and a veritable feast is at hand. This year was especially good because Thanksgiving Day included three of my favourite things: roast turkey, Blue Jays playoff baseball and coding in SAS/STAT.
First the baseball. If you are a SAS User residing in the U.S. then you are thinking that baseball is over by Thanksgiving, right? That’s true, unless you live in Canada. The Toronto Blue Jays defeated the Texas Rangers on October 9 (Thanksgiving in Canada) to win the American League Division Series. Let’s not discuss the AL Championship Series.
The roast turkey, as a Thanksgiving tradition, needs no explanation. The real story here is how SAS/STAT found its way on to my Thanksgiving Day menu.
It is natural to enjoy perks from family members’ occupations: retail employee pricing, airline benefits, medical advice, etc. Are there any perks from having a statistician in the family? Yes indeed. And one who knows SAS programming, even better.
My niece, Meaghan, sent me a note before the Canadian Thanksgiving weekend to let me know that she was bringing home research data from the University of Ottawa, where she is a Masters candidate in the Department of Biology. We were gathering at her parents’ home in Annan Ontario, overlooking the Georgian Bay of Lake Huron (exquisite view!). She had some questions regarding the statistical analyses. When we sat down to review the study and the data, I noticed that she was using two software packages: MS Excel (okay, I guess), and another package that will remain nameless. I joked that I could not help her until the offending file was deleted from my view. Meaghan explained that she would like to use SAS, but that it is too expensive. Ah ha! “Not so, mon ami.”
It did not take long to explain that SAS University Edition is free for use to students and faculty and suggested she download it. Within a few hours, Meaghan had SAS installed on her MacBook (!). In no time she was able to execute the SAS program that I helped her code to import her data from a CSV file, transform it for a mixed model analysis, and run several candidate models using the GLIMMIX Procedure. Meaghan’s study is a classic split-split-plot design where the plots, in this case, are rabbits. Seven replicates were measured in the smallest unit, and different rabbits were “measured” at 3 times: 4, 8, and 12 weeks. But it is better if I turn this blog over to Meaghan to explain the study and how she is learning SAS.
A new SAS user is born - Meaghan’s story
In my study, forty-five rabbits were randomly allocated into two different experimental groups: experimental and control. The thirty rabbits in the experimental group were subjected to the experimental intervention approved and overviewed by the Ethics Committee at my Institution. The rabbits were sacrificed in groups of ten, 4, 8, and 12 weeks post-intervention. Both shoulders were harvested in the experimental group, creating experimental and contralateral groups. For the control group, both shoulders were harvested to make the control group. The experiment gets more complicated and for each specimen, data collection was in three distinct locations of the same tissue and for each location seven distinct areas were characterized by microscopy.
As you can imagine, a graduate student with basic statistical training is going to have trouble determining what approach is the most appropriate to compare all data and draw statistically supported conclusions. When I began my statistical analysis I was attempting to dig deep and bring back the knowledge I had learned in the courses I had taken during my undergraduate studies. After two months of struggling with various software packages, I was desperately looking for some help to get me out of the statistical abyss. Lucky for me, Canadian Thanksgiving was quickly approaching and Thanksgiving means family. My family is composed of quite a few wonderful individuals, but not all of them will make the journey a day early after hearing my statistical analysis was in dire need of assistance. Uncle John to the rescue! Following a quick walk to stretch his legs after the three-hour drive, we got down to business. Knowing Uncle John is a dedicated SAS user, I was very hesitant to show him my previous attempts using a software package that shall remain nameless. When Uncle John asked why I didn’t use SAS, I explained that people in my laboratory had used the aforementioned software for years and were not likely to change their ways anytime soon. My dream statistical software package would include four main things: a refresher in statistical analysis, software that is user friendly, something I could use remotely from my laptop and one that had a good troubleshooting guide. This is when Uncle John said that one magical word that graduate students around the world all appreciate, ”free”. SAS is free to students? This is statistical GOLD.
Once he’d convinced me that the software was, in fact free(I really had a hard time accepting this fact, if you can’t tell by the number of times I’ve repeated myself), I began downloading it. SAS University Edition requires very little set-up. You simply download two files from a very clearly laid out website and install them onto your computer. From there, the download guide walks you through how SAS University Edition works. You use the downloaded packages to remotely access the server through your browser of choice. To be honest, I don’t completely understand how this part works but I do know it works well, is very user friendly and gets me FREE access to SAS University Edition.
SAS also offers two courses to help you get started, and they’re free too! The first course is SAS Programming 1: Essentials, and let me tell you, it is definitely essential. This course focuses on getting you familiar with the basics of SAS programming. Things start simple, with importing data sets and accessing that data. As you progress through the eleven lessons of this course you begin to get more comfortable with the SAS University Edition interface and generating reports that are relevant and accurate for your needs. Right now, my breaks from lab work let me go through these interactive training videos at my own pace and as many times as I want. For someone who doesn’t have the time to complete a course without interruption, this module has been created in a very user friendly way. They are very interactive, saving exhausted university students from flipping through the pages at 2am without actually digesting the information, they allow you to go through the information as many times as necessary, and you can access the course whenever you want. The second available course is Statistics 1. Although I haven’t started this course yet, I have a feeling it is going to be the refresher course I was looking for to bring back all that information that I need to be able to make informed decisions about where to take my statistical analysis. So far, SAS University Edition has provided me with all the tools I need to be successful in my statistical analysis and research. I strongly encourage all students to give this software package a try because once you start using it, you will not look back!