HISTOGRAM
- MAKING & INTERPRETATION
Faculty : Er Arun Kumar Sharma
Author & Corporate Trainer
SubMastery
E-Mail – support@[Link], Phone - 9084510186
Histogram
• Histogram is a type of bar chart representing the frequency
distribution of the data from a process
Rationale and Benefits
• Displays large amount of data, that are difficult to interpret in
tabular & graphical form
• Shows the relative frequency of occurrence of the various data
values
• Provides useful information for understanding the present
performance and predict future performance of the process
Rationale and Benefits
• Reveals the centering, variation & shape of the data
• Helps evaluate process capability & the question “is the process
capable of meeting customers requirements?”
• Enables corrective action to keep the process within specification
limits
• Excellent planning and forecasting tool
When to Use a Histogram
• When the data are numerical
• When you want to see the shape of the data’s distribution,
especially when determining whether the output of a process is
distributed approximately normally
• When analyzing whether a process can meet the customer’s
requirements
• When analyzing what the output from a supplier’s process looks
like
When to Use a Histogram
• When seeing whether a process change has occurred from one
time period to another
• When determining whether the outputs of two or more processes
are different
• When you wish to communicate the distribution of data quickly
and easily to others
Example
Steps in construction of Histogram
• Decide on the process measure
• Gather data
• Prepare of frequency table
• Calculate range, class intervals, class width and boundaries
• Draw the axes
• Draw the bars
• Interpret the histogram
[Link] on the process measure
• Only one parameter can be used for constructing the histogram
• The data should be variable data, i.e., Measured on a continuous
scale (weight, time, temperature, dimensions, speed, etc.)
• Example : weight of a rice bag
[Link] and Record data
• For accurate analysis of mean, dispersion and patterns,
collect at least 50 to 100 data points
98.7 100.6 100.2 100.6 100.3 100.3 98.6 99.6
100.0 100.3 99.9 99.9 99.9 100.8 100.5 99.5
100.4 99.7 100.2 99.8 100.1 99.7 100.4 99.8
99.6 100.3 99.4 100.4 99.0 99.9 99.5 99.2
101.5 99.3 101.3 99.2 100.9 100.1 101.6 100.2
100.5 101.1 100.0 101.2 100.1 99.4 99.1 100.1
99.2 100.0 99.6 100.0 100.2 101.4 101.0 100.2
101.0 98.8 100.7 100.3 100.8 100.1 100.0 100.9
[Link] a frequency table from the data
• Calculate the number of data points, N
• Determine the range, R
• Determine the number of class intervals, K (SQRT of N)
• Determine the class width (bin width), H
• Construct the frequency table
[Link] a frequency table from the data
• Calculate the number of data points, N
• Count the number of data points “n” in the sample
N = 64 data points
[Link] a frequency table from the data
• Calculate the number of data points, N
• Determine the range, R
• Determine the range “R” in the sample. Range is the smallest
value in the set of data subtracted from the largest value
R = Xmax – Xmin
R = 101.6 – 98.6 = 3.0
[Link] a frequency table from the data
• Calculate the number of data points, n
• Determine the range, R
• Determine the number of class intervals, k as SQRT N
like here SQRT N i.e 64 = 8
[Link] a frequency table from the data
Calculate the number of data points, n
Determine the range, R
Determine the number of class intervals, k
Determine the class width, H
Determine the class width “H”. The formula for this is
H = R/k
This should be rounded off to a suitable value depending upon the
data collected
H = 3.0 / 8
H = 0.375 ~ 0.5
[Link] a frequency table from the data
Calculate the number of data points, n
Determine the range, R
Determine the number of class intervals, k
Determine the class width, H
Establish the class boundaries
Use the smallest individual measurement in the sample or round
off to the next appropriate lowest round number. This will be the
lower end point for the first class interval
Do the same for the highest number. This will be the higher end
point for the last class interval
Class Boundaries – 98.5 & 102.0
[Link] a frequency table from the data
• Establish class intervals
• Add the class width “H” to the lower end point. This will be the
lower end point for the next class interval
Each class interval must be mutually exclusive
Add the class width to the lower class boundary until the “K” class
intervals and/or the range of all the numbers are obtained
98.50 – 98.99 100.00 – 100.49
99.00 – 99.49 100.50 – 100.99
99.50 – 99.99 101.00 – 101.49
101.50 – 101.99
[Link] a frequency table from the data
• Construct the frequency table
Class Class
Frequency Total
# Boundaries
1 98.50 – 98.99 3
2 99.00 – 99.49 8
3 99.50 – 99.99 13
4 100.00 – 100.49 23
5 100.50 – 100.99 9
6 101.00 – 101.49 6
7 101.50 – 101.99 2
[Link] the Axes
[Link] the Bars
Interpretation of the Histogram
Process Centering
Process Variation (Dispersion)
Histogram Shape
Process Comparison with specification(Process
Capability)
Interpretation of the Histogram
Process Centering
Interpretation of the Histogram
Process Centering
Interpretation of the Histogram
Process Centering
Interpretation of the Histogram
Process Variation (Dispersion)
Process
within
What is the
requirements
variation or
spread of
the data? Is
Process
it too
too
Variable?
variable
Interpretation of the Histogram
Histogram Shape
Normal distribution Bi-modal distribution
Distribution with isolated
Cliff type island
Interpretation of the Histogram
Process Comparison with specification(Process
Capability)
Process capable Process capable
and centered but not centered
Maintain Process Re-set to centre
the Process
Interpretation of the Histogram
Process Comparison with specification(Process
Capability)
Process just Process just
capable and capable but not
centered centered
Reduce variation Centre process and
to avoid defects reduce variations
Interpretation of the Histogram
Process Comparison with specification(Process
Capability)
Process not capable
but centered
Reduce process
variation to avoid
defects
Check points for Histogram
Choose critical process parameters
Gather as much as data as possible
Choose appropriate class intervals, width, boundaries
and scale
Specify targets and specification limits
Take corrective action and verify
Histogram Analysis
Before drawing any conclusions from your histogram,
satisfy yourself that the process was operating normally
during the time period being studied. If any unusual events
affected the process during the time period of the
histogram, your analysis of the histogram shape probably
cannot be generalized to all time periods
Analyze the meaning of your histogram’s shape
Thank you
