Documente Academic
Documente Profesional
Documente Cultură
bell curve.
In the bell curve, the highest point is the one that has the highest probability of occurring, and
the probability of occurrences goes down on either side of the curve.
It is often used during employee performance appraisals or during evaluation in exams (ever
heard – “You will be graded on the curve?”).
Now before I jump in on how to create a bell curve in Excel, let’s get a better understanding
of the concept by taking an example.
This means that even if your team is the best team ever and you’re all super heroes, only a
handful of you would get the top rating, most of the people in your team would get an
average rating, and a handful will get the lowest rating.
Fair question!
Suppose you have a class of 100 students that appear for an exam. According to your grading
system, anyone who gets above 80 out of 100 gets an A grade. But since you set a really easy
paper, everyone scored above 80 and got the A grade.
Now there is nothing wrong in this kind of grading system, however, using it, you can not
differentiate between someone who got 81 and someone who got 95 (as both would get the A
grade).
To keep the comparison fair and keep the competitive spirit alive, the bell curve is often used
to evaluate performances (at least that’s how it was when I was in college).
Using the bell curve approach, the marks of students are converted into percentiles that are
then compared with each other. Students getting higher marks are on the right side of the
curve and students getting low marks are on the left of the curve (with most of the students
being in the middle around mean score).
Now to understand bell curve, you need to know about two metrics:
When you have a dataset that is normally distributed, your bell curve will follow the below
rules:
The center of the bell curve is the mean of the data point (also the highest point in the
bell curve).
68.2% of the total data points lie in the range (Mean – Standard Deviation to Mean +
Standard Deviation).
95.5% of the total data points lie in the range (Mean – 2*Standard Deviation to Mean
+ 2*Standard Deviation)
99.7% of the total data points lie in the range (Mean – 3*Standard Deviation to Mean
+ 3*Standard Deviation)
Here are the steps to create a bell curve for this dataset:
In cell A1 enter 35. This value can be calculated using Mean – 3* Standard
Deviation (65-3*10).
In the cell below it enter 36 and create a series from 35 to 95 (where 95 is Mean + 3*
Standard Deviation). You can do this quickly by using the autofill option, or use the
Again use the fill handle to quickly copy and paste the formula for all the cells.
Select the data set and go to Insert tab.
Note that when you have a low standard deviation, you get a packed slim bell curve, and
when you have a high standard deviation, the bell curve is wide and covers more area on the
chart.
This kind of bell curve can be used to identify where a data point lies in the chart. For
example, in case a team is full of high performers, when evaluated on a curve, despite being a
high performer, someone can get an average rating as he/she was in the middle of the curve.
Note: In this blog post, I have discussed the concept of a bell curve and how to create it in
Excel. A statistician would be better suited to talk about the efficacy of the bell curve and
limitations associated with it. I am more of an Excel guy and my involvement with Bell curve
has been limited to the calculations I did when I worked as a Financial Analyst.