INTERVAL QUALITY
Vasantha Gundeti
OUTLINE
 Introduction
 Problem
 IQ-Definition
 Goal
 Approaches
 Comparisons
 Interval Quality
 Major vs Minor releases
 Benefits
 Methodology
 Conclusion
Introduction
 Now-a-days software organizations are
concentrating more on quality rather than
time.
 Product with lower quality may lead to more
expense to improve the quality of the
product.
 A key software engineering objective is to
improve quality via practices and tools.
Software development organizations had
raised question
 How to evaluate the customers perception
of the quality of the software product is?
 How customer perceived quality is related
to process quality?
 Every organization makes changes in its
development process to improve the quality
of software as per the requirements of the
customer.
Problem
 Comparing process improvement from the
customers perception has become difficult
problem
Solutions:
• Lead to success - Satisfies the customer
during the product release.
• Need to find the way of improving its
process.
Goal
 To find a way to improve the process so
software quality lead to better customer
satisfaction.
 To find a practical approach to measure
customer’s perception of software quality.
Interval Quality
 The measure that calculates the number of
failures observed by the customer in certain
intervals.
Approaches to Measure Quality
•
•
•
Theoretical models
Indirect Observation
 Test runs, load tests, stress tests, SW
defects and failures
Direct Observation
Comparison
 Compare interval quality with other
measures
Defect density
Mean time between failures (MTBF)
Defect Density
 Defect density is the number of defects
detected in software during a defined period
of development divided by the size of the
software.
Quantity
0.0
0.005
0.010
0.0
0.015
0.0
Defect Density
F3
DL
DL
DM
F1
F3
F3
DefPerKLOC/100
DefPerPreGaMR*10
Probability 1 m .
Probability 3 m .
DM
F1
DM
F3
DM
F 3L
D
D
F 3L
F1
DM
DL
0.000
0.0
D
DM
L
r1.1
F1
r1.2
r1.3
D
FM
1L
D
F3
r2.0
F1
F1
r2.1
r2.2
Releases
Figure 1. The trend of the defect density and quality measures over releases [5]
Mean Time Between Failures(MTBF)
 MTBF is the predicted elapsed time between
inherent failures of a system during
operation.
0.0
Estimated Hazard
Rate
0.2
0.4
0.6
0.8
MTBF
0.0
0.2
0.4
0.6
0.8
1.0
years
 Probability density of observing a failure
conditional on the absence of earlier failures
Interval Quality (IQ)
 Estimation of three probabilities
 P1:The fraction of systems to have a failure
in the first month of usage.
 P2-3:The fraction of systems to have a failure
in months two and three of usage.
 P4-6:The fraction of systems to have a
failure in months four through six of usage.
Major versus Minor releases
 Defect density is lowest for major releases and
highest for minor releases.
 Defect density calculates defects per line of code
or per change and in larger releases the
denominator creates the trend as the number of
defects is relatively constant.
 In the interval quality measure, the denominator
represents the number of customers and the IQ is
larger in minor releases that are deployed more
broadly.
Benefits
 IQ clearly does not capture feature-richness,
often a strong customer satisfier.
 The overall quality may stay constant.
 The use of IQ in the production
environment suggests its value for internal
quality management.
Methodology: Validation
 Interview a sample of individuals operating
and maintaining relevant systems
 Validate and clean retrieved and modeled
data
 Iterate
Methodology: Existing Models
 Predicting the quality of a patch
 Work coordination
 Effort: estimate MR effort
Conclusion
 Software quality is usually measured by
defect density, MTBF, and other measures.
 Interval quality refers to failures of the
software observed in certain intervals by the
customer.
 This presentation is concerned with
different types of software quality measures
and the practical approach to measure
customer perceived quality by interval
quality.
References
[1] A. Mockus, P. Zhang, and P. Li. Drivers for customer
perceived software quality. In ICSE 2005, pages 225–
233, St Louis, Missouri, May 2005. ACM Press.
[2] M. Buckley and R. Chillarege. Discovering
relationships between service and customer
satisfaction. Proceedings of the International
Conference on Software Maintenance, pages 192 – 201,
1995.
[3] S. Chulani, P.Santhanam, D. Moore, and G. Davidson.
Deriving a software quality view from customer
satisfaction and service data. European Conference on
Metrics and Measurement, 2001.
[4] A. Mockus, David M. Weiss, and Ping Zhang.
Understanding and predicting effort in software
projects. In 2003 International Conference on Software
Engineering, pages 274–284, Portland, Oregon, May 310 2003. ACM Press.
[5] A. Mockus., D. Weiss., Interval quality: relating
customer-perceived quality to process quality, ICSE ’08
Proceedings of the 30th International Conference on
Software Engineering, New York, NY, USA, October
2008,pp 723-732.