About

My first visions of the website started several years ago. That was before I retired. At the time I was working for the US Federal Government on surveys and censuses. I felt that dealing with criticisms of statistical issues was at least for me something of a conflict of interest in that I did not feel free to comment on statistics released by the federal government or on uses of those statistics. I acquired the domain name badstats.org and managed to put all of three posts.  So now I am retired and not longer have such constraints. I also did not like the purely negative connotations of the domain name so when I retired I decided on the new name goodstatsbadstats.

The impetus behind this site is my longstanding frustration with the way statistics are used. It has been said “There are lies, damn lies, and statistics.” I would say that statistics themselves is not the problem. Rather the problems are in how they are constructed and how they are interpreted. With that in mind the readers can expect me to got beyond the statistics themselves to raise questions about how they were derived and also to delve into the logic that is used in the analysis of the data. Sometimes statistics are misused. But at other times it is an error in presentation, analysis, logic or understanding. Then there also is the continuing issues of lack of data. I am a firm believer in data driven decisions. Thus I expect those who are making “statistical claims” to provide in a readily available forum the data need to evaluate the accuracy of their claim and the quality of the data being used to make those claims – be those media reports, press releases, and professional papers.

So I will deal with many topics in this blog – not just the statistics themselves. The comments that I make are best viewed as making use of an opportunity to provide a “teaching moment” and not just as a negative criticism of the data presentation. I will also be adding in some data analysis and commentary of my own making.

By way of background my name is Larry Cahoon. I have a PhD in Statistics and have spent over thirty-five years working on government surveys and censuses. I can be reached at LarryC@goodstatsbadstats.com.

And of course whatever I have to say is just my opinion………….

3 Comments

  1. Andrea says:

    Hi,Larry,

    Glad to find your blog here.

    First of all let me introduce myself in case of any misunderstanding… I’m Andrea, 19-year-old college girl from China. I was struggling with the problems in Data Analysis course on Coursera when i saw one of your comments in the forum. So, sorry to say but yes, i’m here for help… Hope you’re not bothered by my visiting.

    I started to get confused as the course moved on into the second week. Honestly speaking I was not that enthusiastic into Statistics, signed up out of mere coincidence, but I was currently in a time looking for what I really love in life and i do want to finish this course with a satisfactory performance, be it a try or anything. In the first week, I had already spent a lot of effort in knowing the Coursera platform, installing R, dealing with all sorts of problems like time zone, studying materials, languages and so on. Though it was exciting, I was also really tired..and don’t know if I could make it to the end.

    Is the workload and the amount of lecture in Data Analysis course alike with that in a good US university? I’ve always longed to be in one of them, though it’s a bit hard to realize.

    One more question, is it common that I am confronted with so much trouble with data analysis now, or am i just being less dedicated or less intelligent?

    Thank you very much!
    Andrea
    2013.2.3. 23:45 UTC+8:00

    • Larry says:

      Andrea,

      Glad you found the blog. As for the data analysis course it is difficult to make a direct comparison to the usual fare for college course here in the States. The general rule of thumb is that you should expect 3-4 hours of work on your own for each hour of lectures. But this course is so different and how much time that is needed will depend very much on your background. If you don’t know much, or have done very little programming learning R is going to take a good bit of time. If you done some programming the R is easier to pick up. I have found that it has a steep learning curve and the official documentation that is available is not very clear to someone who is not already familiar with R.

      I still don’t know how much the course is going to get into statistics. My take at this time is that his goal is more to discuss how to do an analysis and how to present an analysis than to turn it into a statistics course. If someone wants a statistics course I would not be recommending this one. I looked at the example exercise he posted yesterday for the first analysis assignment and found it to be a very simplistic problem. I’m not sure I like that, but I think it is consistent with his goals. What I do like about the course, and I think this is good for anyone doing any kind of science is that he seems to be trying to make people aware of the pitfalls of any type of analysis. I hope he continues along those lines.

Leave a Reply

Your email address will not be published. Required fields are marked *

  • Subscribe to Blog via Email

    Enter your email address to subscribe to this blog and receive notifications of new posts by email.

  • May 2017
    S M T W T F S
    « Oct    
     123456
    78910111213
    14151617181920
    21222324252627
    28293031  
  • Recent Posts