The tutorial explains the essence of the standard deviation and standard error of the mean as well as which formula is best to be used for calculating standard deviation in Excel.
In descriptive statistics, the arithmetic mean (also called the average) and standard deviation and are two closely related concepts. But while the former is well understood by most, the latter is comprehended by few. The aim of this tutorial is shed some light on what the standard deviation actually is and how to calculate it in Excel.
The standard deviation is a measure that indicates how much the values of the set of data deviate (spread out) from the mean. To put it differently, the standard deviation shows whether your data is close to the mean or fluctuates a lot.
The purpose of the standard deviation is to help you understand if the mean really returns a "typical" data. The closer the standard deviation is to zero, the lower the data variability and the more reliable the mean is. The standard deviation equal to 0 indicates that every value in the dataset is exactly equal to the mean. The higher the standard deviation, the more variation there is in the data and the less accurate the mean is.
To get a better idea of how this works, please have a look at the following data:
For Biology, the standard deviation is 5 (rounded to an integer), which tells us that each student's score is no more than 5 points away from the mean. Is that good? Well, yes, it indicates that the Biology scores of the students are pretty consistent.
For Math, the standard deviation is 23. It shows that there is a huge dispersion (spread) in the scores, meaning that some students performed much better and/or some performed far worse than the average.
In practice, the standard deviation is often used by business analysists as a measure of investment risk - the higher the standard deviation, the higher the volatility of the returns.
In relation to standard deviation, you may often hear the terms "sample" and "population", which refer to the completeness of the data you are working with. The main difference is as follows:
Researchers and analysists operate on the standard deviation of a sample and population in different situations. For example, when summarizing the exam scores of a class of students, a teacher will use the population standard deviation. Statisticians calculating the national SAT average score would use a sample standard deviation because they are presented with the data from a sample only, not from the entire population.
The reason the nature of the data matters is because the population standard deviation and sample standard deviation are calculated with slightly different formulas:
Sample standard deviation |
Population standard deviation |
Where:
Having difficulties with understanding the formulas? Breaking them down into simple steps might help. But first, let us have some sample data to work on:
First, you find the mean of all values in the data set (x in the formulas above). When calculating by hand, you add up the numbers and then divide the sum by the count of those numbers, like this:
(1+2+4+5+6+8+9)/7=5
To find mean in Excel, use the AVERAGE function, e.g. =AVERAGE(A2:G2)
This is the part of the standard deviation formula that says: (x_{i} - x)^{2}
To visualize what's actually going on, please have a look at the following images.
In this example, the mean is 5, so we calculate the difference between each data point and 5.
Then, you square the differences, turning them all into positive numbers:
To say "sum things up" in mathematics, you use sigma Σ. So, what we do now is add up the squared differences to complete this part of the formula: Σ(x_{i} - x)^{2}
16 + 9 + 1 + 1 + 9 + 16 = 52
So far, the sample standard deviation and population standard deviation formulas have been identical. At this point, they are different.
For the sample standard deviation, you get the sample variance by dividing the total squared differences by the sample size minus 1:
52 / (7-1) = 8.67
For the population standard deviation, you find the mean of squared differences by dividing the total squared differences by their count:
52 / 7 = 7.43
Why this difference in the formulas? Because in the sample standard deviation formula, you need to correct the bias in the estimation of a sample mean instead of the true population mean. And you do this by using n - 1 instead of n, which is called Bessel's correction.
Finally, take the square root of the above numbers, and you will get your standard deviation (in the below equations, rounded to 2 decimal places):
Sample standard deviation | Population standard deviation |
√8.67 = 2.94 | √7.43 = 2.73 |
In Microsoft Excel, standard deviation is computed in the same way, but all of the above calculations are performed behind the scene. The key thing for you is to choose a proper standard deviation function, about which the following section will give you some clues.
Overall, there are six different functions to find standard deviation in Excel. Which one to use depends primarily on the nature of the data you are working with - whether it is the entire population or a sample.
To calculate standard deviation based on a sample, use one of the following formulas (all of them are based on the "n-1" method described above).
STDEV(number1,[number2],…)
is the oldest Excel function to estimates standard deviation based on a sample, and it is available in all versions of Excel 2003 to 2016.
Excel STDEV can accept up to 255 arguments that can be represented by numbers, arrays, named ranges or references to cells containing numbers.
Logical values and text representations of numbers supplied directly in the list of arguments are counted. In arrays and references, only numbers are counted; empty cells, logical values of TRUE and FALSE, text and error values are ignored.
STDEV.S(number1,[number2],…)
is an improved version of STDEV, introduced in Excel 2010.
Like STDEV, the STDEV.S function calculates the sample standard deviation of a set of values based on the classic sample standard deviation formula discussed in the previous section.
STDEVA(value1, [value2], …)
is another function to calculate standard deviation of a sample in Excel. It differs from the above two only in the way it handles logical and text values:
If you are dealing with the entire population, use one of the following function to do standard deviation in Excel. These functions are based on the "n" method.
STDEVP(number1,[number2],…)
is the old Excel function to find standard deviation of a population.
In the new versions of Excel 2010, 2013 and 2016, it is replaced with the improved STDEV.P function, but is still kept for backward compatibility.
STDEV.P(number1,[number2],…)
is the modern version of the STDEVP function that provides an improved accuracy. It is available in Excel 2010 and later versions.
Like their sample standard deviation counterparts, within arrays or reference arguments, the STDEVP and STDEV.P functions count only numbers. In the list of arguments, they also count logical values and text representations of numbers.
STDEVPA(value1, [value2], …)
calculates standard deviation of a population, including text and logical values. With regard to non-numeric values, STDEVPA works exactly like the STDEVA function does.
A variety of standard deviation functions in Excel can definitely cause a mess, especially to unexperienced users. To choose the correct standard deviation formula for a particular task, just answer the following 3 questions:
To calculate standard deviation based on a numeric sample, use the STDEV.S function in Excel 2010 and later; STDEV in Excel 2007 and earlier.
To find standard deviation of a population, use the STDEV.P function in Excel 2010 and later; STDEVP in Excel 2007 and earlier.
If you want logical or text values to be included in the calculation, use either STDEVA (sample standard deviation) or STDEVPA (population standard deviation). While I can't think of any scenario in which either function can be useful on its own, they may come in handy in bigger formulas, where one or more arguments are returned by other functions as logical values or text representations of numbers.
To help you decide which of the Excel standard deviation functions is best suited for your needs, please review the following table that summarizes the information you've already learned.
STDEV | STDEV.S | STDEVP | STDEV.P | STDEVA | STDEVPA | |
Excel version | 2003 - 2016 | 2010 - 2016 | 2003 - 2016 | 2010 - 2016 | 2003 - 2016 | 2003 - 2016 |
Sample | ✓ | ✓ | ✓ | |||
Population | ✓ | ✓ | ✓ | |||
Logical values in arrays or references | Ignored | Counted (TRUE=1, FALSE=0) |
||||
Text in arrays or references | Ignored | Counted as zero | ||||
Logical values and "text-numbers" in the list of arguments | Counted (TRUE=1, FALSE=0) |
|||||
Empty cells | Ignored |
Once you have chosen the function that corresponds to your data type, there should be no difficulties in writing the formula - the syntax is so plain and transparent that it leaves no room for errors :) The following examples demonstrate a couple of Excel standard deviation formulas in action.
Depending on the nature of your data, use one of the following formulas in Excel 2010, Excel 2013, and Excel 2016:
=STDEV.P(B2:B50)
=STDEV.S(B2:B10)
As you can see in the screenshot below, the formulas return slightly different numbers (the smaller a sample, the bigger a difference):
In Excel 2007 and lower, you'd use STDEVP and STDEV functions instead:
=STDEVP(B2:B50)
=STDEV(B2:B10)
When discussing different functions to calculate standard deviation in Excel, we sometimes mentioned "text representations of numbers" and you might be curious to know what that actually means.
In this context, "text representations of numbers" are simply numbers formatted as text. How can such numbers appear in your worksheets? Most often, they are exported from external sources. Or, returned by so-called Text functions that are designed to manipulate text strings, e.g. TEXT, MID, RIGHT, LEFT, etc. Some of those functions can work with numbers too, but their output is always text, even if it looks much like a number.
To better illustrate the point, please consider the following example. Supposing you have a column of product codes like "Jeans-105" where the digits after a hyphen denote the quantity. Your goal is to extract the quantity of each item, and then find the standard deviation of the extracted numbers.
Pulling the quantity to another column is not a problem:
=RIGHT(A2,LEN(A2)-SEARCH("-",A2,1))
The problem is that using an Excel standard deviation formula on the extracted numbers returns either #DIV/0! or 0 like shown in the screenshot below:
Why such weird results? As mentioned above, the output of the RIGHT function is always a text string. But neither STDEV.S nor STDEVA can handle numbers formatted as text in references (the former simply ignores them while the latter counts as zeros). To get the standard deviation of such "text-numbers", you need to supply them directly to the list of arguments, which can be done by embedding all RIGHT functions into your STDEV.S or STDEVA formula:
=STDEV.S(RIGHT(A2,LEN(A2)-SEARCH("-",A2,1)), RIGHT(A3,LEN(A3)-SEARCH("-",A3,1)), RIGHT(A4,LEN(A4)-SEARCH("-",A4,1)), RIGHT(A5,LEN(A5)-SEARCH("-",A5,1)))
=STDEVA(RIGHT(A2,LEN(A2)-SEARCH("-",A2,1)), RIGHT(A3,LEN(A3)-SEARCH("-",A3,1)), RIGHT(A4,LEN(A4)-SEARCH("-",A4,1)), RIGHT(A5,LEN(A5)-SEARCH("-",A5,1)))
The formulas are a bit cumbersome, but that might be a working solution for a small sample. For a bigger one, not to mention the entire population, it is definitely not an option. In this case, a more elegant solution would be having the VALUE function convert "text-numbers" to numbers that any standard deviation formula can understand (please notice the right-aligned numbers in the screenshot below as opposed to the left-aligned text strings on the screenshot above):
In statistics, there is one more measure for estimating the variability in data - standard error of mean, which is sometimes shortened (though, incorrectly) to just "standard error". The standard deviation and standard error of the mean are two closely related concepts, but not the same.
While the standard deviation measures the variability of a data set from the mean, the standard error of the mean (SEM) estimates how far the sample mean is likely to be from the true population mean. Said another way - if you took multiple samples from the same population, the standard error of the mean would show the dispersion between those sample means. Because usually we calculate just one mean for a set of data, not multiple means, the standard error of the mean is estimated rather than measured.
In mathematics, the standard error of mean is calculated with this formula:
Where SD is the standard deviation, and n is the sample size (the number of values in the sample).
In your Excel worksheets, you can use the COUNT function to get the number of values in a sample, SQRT to take a square root of that number, and STDEV.S to calculate standard deviation of a sample.
Putting all this together, you get the standard error of the mean formula in Excel:
Assuming the sample data are in B2:B10, our SEM formula would go as follows:
=STDEV.S(B2:B10)/SQRT(COUNT(B2:B10))
And the result might be similar to this:
To visually display a margin of the standard deviation, you can add standard deviation bars to your Excel chart. Here's how:
This will insert the same standard deviation bars for all data points.
This is how to do standard deviation on Excel. I hope you will find this information helpful. Anyway, I thank you for reading and hope to see you on our blog next week.
12 Responses to "How to calculate standard deviation in Excel"
How to graph the normal distribution curve in excel 2013
Hi Benard,
You can find the detailed instructions here:
http://www.tushar-mehta.com/excel/charts/normal_distribution/
New topics, new learn. Thanks Svetlana..
Hi, I'm trying to come up with a formula that copies contents of a cell only if the value of that cell being copied is -1 or less, then converts copied negative number to a positive value. A Here's a formula I tried =if(Q2>=0,"",Q2)*-1
(P2 is copying data from Q2) but instead if leaving cell blank it displays a 0. How do you think I should change the formula to have a cell stay blank instead of populating it with 0?
If anyone is also wondering it was as simple as moving *-1 inside parentheses
=if(Q2>=0,"",Q2*-1)
Hello
I have to calculate the standard deviation of the centre of pressure displacement which expresses the variability of the centre of pressure at specific points in time. I will do the calculations in excel, however, I would like to know is the standard deviation formula the same for any variable for which SD is used?
Thanks this has been really helpful. Can you tell me does excel calculate Stdev.s or Stdev.p when you add “Standard Deviation” from “error bars” in “Add chart element” ?
Hi Libby,
For error bars, Excel actually calculates standard deviation as a single value form the given Y-values, assuming that the Y-values are individual values from one sample. As the result, the default error bars drawn by Excel are the same for all data points in a series. The formula that Excel uses for calculating standard deviation error bars can be found here.
If you'd like to plot individual Std Dev error bars, you can first calculate Stdev.s or Stdev.p, depending on your data set. And then, on your chart, click Error Bars > More Options >Custom, and select the range of the calculated Std Dev values. An example can be found here: How to make individual error bars in Excel.
PROF. You are a great lecturer, I am grateful to grab a bit knowledge from you within this few seconds.my regard to u.
Extremely clear and comprehensive. Not everybody that knows this stuff can explain it as straightforwardly as this! Great work!
I'm looking for formula wherein I need to calculate Standard Deviation only for selected criteria. Like I have A B C D and those A B C D are available in my table from where I need to pull respective numbers. So I need to calculate standard dev. for A separately, B separately and so on... Also, need to take care of excluding Zeros 0. Presently I'm using below formula.
=IFERROR(STDEV.P(IF(ISNUMBER(1/D4:D88),IF(D4:D88>0,D4:D88))),"-")
I need to add comparison to it.
Thanks a bunch...
Q. How to calculate the statistic S y/x, which estimates the random errors in the y-direction by used excel