Statistics Websites (still under construction)
The Web is popular, in all senses of the word. That many people like it, is not surprising. What does warrant notice, though, is the sense of "popular" where anybody who wants to can put up a web site about anything they want to. Caveau Emptor is the warning that should be the most often heard on the Internet. Not only buyers of tangible products, but especially "buyers""of information (like statistics) that are being presented on the Web. I saw one site where "correlation" was defined as a positive number between 0 and 1... Whether or not you are familiar with the concept of correlation, suffice it to say that correlation is a number between -1 and 1 that....(the rest is unimportant here). My point is that no one censures such sites, so the person surfing the statistics sites has to look at everything with a critical eye. Back in the day when university professors were the major dispensors of knowledge, this filtering was one of their major functions. Nowadays, they are chasing grants and making money for their institutions, leaving us up to the filtering job ourselves.
Nevertheless, the top sites I have found are mostly university web sites and are excellent, so thank goodness some of those professors are out there taking care of us:
Statistics Web sites fall into the following categories:
- Book publishers trying to sell a book - There's nothing wrong with this. The author is most likely a university professor, researcher, or other who has been carefully screened by the publisher. We must remember that their purpose is to convince you to buy the book, so the presentation will be selected from the book material and will not be so exclusively pedagogical.
- Software companies trying to sell their software - Again, the purpose is to sell you their software, so you may not have the r code or the SAS code, unless it is SAS doing the website. Note that, in general, the SAS documentation is good.
- Professional consultants trying to sell their services (including commercial schools) - The purpose is to entice you to enlist the consulting services, so the objective is to show you what they can do - not necessarily to show you how to do it yourself.
- University consulting groups advertising their consulting (free for the U, but they hope for paying external customers) - Here we are entering into the more altruistic world of statistics.
- Statistics Community: Professional Statistics Organisations (ASA, etc.) or University professors of Statistics taking care of their students - Of course, there are good teachers and bad, but in general they are making their best effort to explain things to you. Often, they give you their code (and often the data). Methods will be oriented toward the content of the course you have found, but, dependent on your google skills, you are most likely in the right place.
- Independent statisticians (regardless of their job) trying to "do it right" - I suppose I am the guilty party here. There are many better sites than mine, but I hope to have something to contribute, and I also hope that this review is also of value. I do this out of passion, not any financial remuneration. I am counting the minutes until retirement. But let's be honest, if you want to hire me, my contact information is below.
- Talented amateurs with experience and knowledge, but not necessarily the "official jargon"
- Untalented amateurs trying to "show what it really is, not like those blue sky theoreticians tell you"
Some good sites for 2020:
- UCLA Institute for Digital Research and Education (type 4 from above) - note the resources tab.
- R Bloggers
- The Analysis Factor
- Univ of Bristol Centre for Multilevel Modelling
- Princeton, sort of
- Amstat (type 5 above)
- SAS (type 2 above)
- Datacamp (type 3 above)
- Quick-R by DataCamp
Please come back to this site soon. And thank you for your comments and suggestions:
Contact me at: email@example.com