Isn’t variability important in our lives? Let me give you an example.

Let’s say two pizza joints advertise a 20 minute average delivery time.

It sounds good!!! When you are hungry, you may be unsure about the ideal way to order your pizza.


Now is the time to analyze both establishments’ variance. Do you know anything about measurements of variability?


Don’t worry; I’ve provided all the specifics you need to know about measuring variability and calculating it.


Also, I’ve explained how to find the ideal restaurant for your favorite pizza.


So, without further ado, let us introduce the new notion of statistics.


What are variance measures?


The statistical summary illustrates the dispersion within the datasets. The standard value is defined by the measure of central tendency.


Statisticians use variability measures to see how far the data points deviate from the assigned central value. That’s why statisticians look at variability to determine value distribution.




The lesser the dispersion, the closer the data points will be grouped.


The higher the dispersion, the further the data points are from the center.


Do variations matter?


Yes, it does!


Lower variability is good for better population projections. The larger variability score is considered less consistent. This will make projections more difficult.


Moreover, the data sets may have a similar central trend yet differ in variability.


Consider either variability or central tendency; you cannot say the same about other aspects. Now, both phrases can assist you understand your data.


What good are variability measures?


Everywhere there is variability. Say you ordered the same dish at a restaurant several times.


The assembly line may appear similar, yet it has varied widths and lengths. This is where you need to use variability to choose the appropriate assembly line for your order.


Aside from that, some degree of variance is unavoidable due to irregularity. How?


If you take longer than average, you may be late for work. You might not order pizza if it tastes very different from the last one. This is how measures of variability work.


Are there 4 types of variability?




It is used to find out the data’s spread from the lowest to the highest value. It is also one of the simplest metrics of variability to calculate.


Subtraction of the least and biggest values in the dataset.


To comprehend it, consider this:


Assume 5 data points:


10 25 5 35 40


Clearly, 40 is the highest and 5 is the lowest. Therefore,


R = H-L => 40-5


The data range is 5 minutes.


As you can see, two integers are used here, thus outliers might affect the range. The range also provides no information on value dispersion.


Add extra measures to achieve accurate findings.




The IQR (interquartile range) represents the distribution’s center. The IQR comprises half of each distribution’s value.




So, third quartile minus first quartile.




Let us use an example:


Consider calculating an IQR for 8 data points. So, get the Q3 and Q1 values first. Then add 0.75 to Q3 and 0.25 to Q1.


0.25*8 Equals 2


0.75*8 Equals 6


Q1 is 110 and Q3 is 287. Now the IQR is:


177 – 287


Like range, IQR requires two numbers to calculate. But IQR has less impact on outliers. IQR also provides uniform variability for normal and skewed distributions.


std de


The SD reveals how far the score is from the average. The higher the SD, the more varied the data set.


Calculate the data set’s standard deviation using the formulas below.








Steps to compute SD


Calculate the average score.


Subtract the average from each score to find the variance.


Square each deviance and add them all.


Subtract N (for the population) or n-1 (for the sample).


To find the standard deviation, take the value and square it.


Let us use an example:


Assume you have 5 data points to compute SD.


Data outliers




Subtract the sum


std dev












70 − 70 = 0


110 – 70 =


50-70 = (-20)


20-70 = (-50)


30 – 70










Square = 5400


This is because we are dealing with a sample.


n+1 = 5+1 = 4


5400/4 = 1350s = 36.74


The data’s standard deviation is 36.74.


It indicates a score variation from 36.74 points.




Squared departure from the mean. Also, variance = SD squared. Notably, variance is more difficult to interpret.


The variance illustrates the distribution of the data. The data spread increases with variance.


Here are the variance formulas.








Let us provide an example to comprehend:


Consider the standard deviation example. The standard deviation squared is:






36 * 36 = 1350


To compute variance, follow the standard deviation steps (excluding the last one).


Let’s get to the bottom of the pizza delivery question!!!


We saw two pizza places claim that they can deliver pizza in 20 minutes. But how to choose the best?


Here, we calculate the measures of variability for each point and compare them. The graph below displays delivery time distribution.


The restaurant variable will have a larger distribution curve due to irregular delivery.








A delivery time of 30 minutes or more is clearly unsatisfactory. We are famished! The graph shows the delivery time proportion shaded.


Almost 16% of deliveries (Restaurant 1) go over 30 minutes. Also, 2% delivery (Restaurant 2) is longer and has less fluctuation. Both eateries deliver in 20 minutes on average. But now I know where to order pizza. That’s 2.


In this case, the central tendency cannot provide complete information. To get a clear response, you need to know the variability around the distribution’s middle.


How can I acquire the best variance measures?


To acquire the best variability, you must examine the measurement distribution and level. So, what are they?


Measuring scale


The IQR (Interquartile Range) and range (explained below) are the only elements that need to be considered to derive the ordinal level of measured data.


But for difficult ratio and interval level, consider variance and SD.




Remember that all metrics are typical. But variance and SD favor the whole data set. However, outliers can easily affect variance and SD.


The IQR measures skewed distribution well. IQR focuses on the middle data spread. That’s why it’s unaffected by extremes.


A recap


Dispersion is another term for variability.


It is the middle half of a distribution.


Difference from average is called variance.


The range is the highest-to-lowest.


The standard deviation is the average variance.




Variability is found in practically every element of life. A statistician must also consider four measures.


Range, IQR, SD, and Variance. We’ve covered all the bases to help you grasp the notion of variability.


I hope you like these long-term details. Besides that, feel free to ask any questions about measurements of variability.


Comment your doubts and get the finest solutions.


“Keep learning with Statanalytica blogs.”


Questions & Answers


What is the most trustworthy variance measure?


Notably, standard deviation uses the original data units that aid in data interpretation. That is why SD is the most widely used measure of variability.


Psychologist’s two measures of variation


It is important to understand the variance and standard deviation in psychology statistics.


How to explain variance?


VARIABILITY: The degree to which a dataset’s points can deviate from a Moreover, it is the degree of variation between the dataset points.