Histogram for Sample Data - Critical Defect No. 1
Cost Data Integrity Project
Once Trend Charts have been completed, enter the same observation data into a Histogram to check for
Normality - may need to revise and change sample plan before moving to Statistical Process Control.
Additional Instructions > Making Histograms
Observation Value Describe the Sample Test and Defect Measured:
1 27 Nine observations were made over a 9 month period regarding the posting of
2 22 vendor invoices into the Project Management System. Compare the invoices
3 20 that were processed and paid per the General Ledger with invoices that were
4 26 posted in the Project Management System. Count the number of invoices that
5 21 were not posted in the Project Management System.
6 28
7 25
8 22
9 28
Maximum 28.0
Minimum 20.0
Range 8.0
Number of Intervals based on Square Root of Observations > 3 5 Making Histograms
Observation Value Describe the Sample Test and Defect Measured:
1 41 For all active employees assigned to projects, how many employees did
2 37 not enter a timesheet into the Project Management System for the
3 35 two week pay period.
4 33
5 36
6 40
7 38
8 36
9 34
Maximum 41.0
Minimum 33.0
Range 8.0
Number of Intervals based on Square Root of Observations > 3 5 < Minimum Intervals
Width of each interval based on Range / (No of Intervals - 1) 2
Histogram Table based on above data: Simple Side Ways Plot
Interval Freq
1 Less than or Equal to 33 1
2 Between 34 and 35 1
3 Between 36 and 37 3 Need additional samples
4 Between 38 and 39 1
5 Between 40 and 41 2
More Formal Histogram Bar Chart:
Frequency Distribution
3.5
3
2.5
Frequency
2
Freq
1.5
1
0.5
0
1 2 3 4 5
Intervals
Histogram Results
Cost Data Integrity Project
Summarize the results of the histograms for each of the three critical defects:
Critical Defect No. 1:
The frequency distribution for vendor invoices is bi-modal indicating that stratified sampling needs to be performed. This is most likely due to the
fact that each IT investment is managed seperately and as a result, each IT project has its own unique defect rates; i.e. some projects do a good
job of processing and posting their invoices, but others do a very poor job. Therefore, each project should represent a sub-group and observations
should be taken from each sub group to arrive at proper measurement of defects.
Critical Defect No. 2:
Similar to defect 1, probably need to conduct subgroup samples on a project by project basis to reach normality.