MathMathter
New member
- Joined
- Nov 23, 2013
- Messages
- 21
Is there a definitive guide for statistics variable names and symbols? I have a calculus background, and I'm learning statistics. I'm struck by both the inconsistency and unintuitiveness in the naming conventions for variables and the inconsistency in the characters used to represent the variables. It seems like this branch of math could use a better language to describe it.
For example, I have two books that I'm learning from: Probability and Statistics for Engineering and the Sciences by Jay L. Devore (4th Edition) and The Complete Idiots Guide to Statistics by Robert A Donnelly Jr. (2nd Edition). Taking the example of the sample proportion, it can be written as a lower-case 'p' (when upper-case 'P' is used for the population proportion) or as either p sub s or p hat (when lower-case 'p' is used for the population proportion). Then, to specify the standard error of the proportion, you use lower-case sigma with a subscript of [whatever you use for sample proportion...unless it's p sub s (then, you just use p? Why? That's what the Donnelly book does and what I've seen on-line)]. And then the first table in this stattrek tutorial (http://stattrek.com/estimation/standard-error.aspx?tutorial=stat) indicates that it's not correct to refer to sigma sub p as the "standard error" of p use "SE" instead. Also, how does "standard error of the proportion" become a short-hand way of saying "standard deviation of the sample proportions" anyway? What sense does it make to take out "sample" and change "deviation" to "error". I understand that languages sometimes develop in arbitrary/haphazard ways, but somehow it's disappointing to see this in math. The language of calculus seems much more pure and the formulas more intuitive...not liking statistics so far.
For example, I have two books that I'm learning from: Probability and Statistics for Engineering and the Sciences by Jay L. Devore (4th Edition) and The Complete Idiots Guide to Statistics by Robert A Donnelly Jr. (2nd Edition). Taking the example of the sample proportion, it can be written as a lower-case 'p' (when upper-case 'P' is used for the population proportion) or as either p sub s or p hat (when lower-case 'p' is used for the population proportion). Then, to specify the standard error of the proportion, you use lower-case sigma with a subscript of [whatever you use for sample proportion...unless it's p sub s (then, you just use p? Why? That's what the Donnelly book does and what I've seen on-line)]. And then the first table in this stattrek tutorial (http://stattrek.com/estimation/standard-error.aspx?tutorial=stat) indicates that it's not correct to refer to sigma sub p as the "standard error" of p use "SE" instead. Also, how does "standard error of the proportion" become a short-hand way of saying "standard deviation of the sample proportions" anyway? What sense does it make to take out "sample" and change "deviation" to "error". I understand that languages sometimes develop in arbitrary/haphazard ways, but somehow it's disappointing to see this in math. The language of calculus seems much more pure and the formulas more intuitive...not liking statistics so far.
