Why do we use a square and square root method in the standard deviation formula?

Click For Summary

Discussion Overview

The discussion centers on the theoretical underpinnings of the standard deviation formula, specifically the rationale for using squaring and square rooting in its calculation. Participants explore the implications of these mathematical operations in relation to measuring distances in data sets and their connection to the normal distribution.

Discussion Character

  • Exploratory
  • Technical explanation
  • Conceptual clarification
  • Debate/contested

Main Points Raised

  • One participant questions the necessity of squaring values in the standard deviation formula, suggesting that using absolute values could avoid issues with negative values and might be more accurate.
  • Another participant argues that squaring values is akin to measuring distance, drawing a parallel to the distance formula in geometry, and notes that absolute value is not differentiable at zero, which complicates its use.
  • Some participants assert that the root mean square definition is significant because it relates to the normal distribution, although the exact nature of this relationship is questioned.
  • One participant states that the normal distribution is defined by two parameters: the mean and the standard deviation, and mentions the empirical rule regarding the distribution of data within one standard deviation of the mean.

Areas of Agreement / Disagreement

Participants express differing views on the appropriateness of squaring values versus using absolute values in the standard deviation formula. While some agree on the connection to the normal distribution, the discussion remains unresolved regarding the necessity and implications of the squaring process.

Contextual Notes

Participants have not fully explored the implications of using absolute values versus squaring, nor have they resolved the mathematical nuances related to differentiability and the properties of the normal distribution.

gsingh2011
Messages
115
Reaction score
1
I would like to understand the theory behind the standard deviation formula. The way it was explained to me, you have to subtract each value from the mean and square it to avoid the negatives canceling out the positives. After multiplying and dividing by the correct frequencies, we have to square root our sum to correct the effects of squaring it. This explanation doesn't make sense to me because we could have just used absolute values to avoid the positive/negative problem, and it seems more accurate. So why do we use this square and then square root method?
 
Physics news on Phys.org
It is, basically, measuring the distance from the data to the norm, just like [itex]\sqrt{(x- a)^2+ (y- b)^2+ (z- c)^2}[/itex] for distance in the plane.

Of course, you could measure "distance" by |x- a|+ |y- b|+ |z- c| and you could calculate such a number for a set of data but "absolute value" is not a very nice function- it is not differentiable at 0.

The most important reason for using "root mean square" definition is that it is the one that shows up in the normal distribution.
 
HallsofIvy said:
It is, basically, measuring the distance from the data to the norm, just like [itex]\sqrt{(x- a)^2+ (y- b)^2+ (z- c)^2}[/itex] for distance in the plane.

Of course, you could measure "distance" by |x- a|+ |y- b|+ |z- c| and you could calculate such a number for a set of data but "absolute value" is not a very nice function- it is not differentiable at 0.

The most important reason for using "root mean square" definition is that it is the one that shows up in the normal distribution.

How does it "show up" in the normal distribution. I know there is a relation, like one standard deviation on each side of the mean is 68%, but why does there have to be a root mean square in order to have this correlation?
 
The normal distribution is actually a family of distributions with two parameters. These are the mean and the standard deviation (or the variance).

The 68% ~ one standard deviation is simply a fact of the shape of the normal distribution.
 

Similar threads

  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 4 ·
Replies
4
Views
2K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 15 ·
Replies
15
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 7 ·
Replies
7
Views
4K
  • · Replies 21 ·
Replies
21
Views
4K
  • · Replies 4 ·
Replies
4
Views
2K