Software Quality CIS 375 Bruce R Maxim UMDearborn

Software Quality CIS 375 Bruce R. Maxim UM-Dearborn 1

Software Quality Principles • Conformance to software requirements is the foundation from which quality is measured. • Specified standards define a set of development criteria that guide the manner in which software is engineered. • Software quality is suspect when a software product conforms to its explicitly stated requirements and fails to conform to the customer's implicit requirements (e. g. ease of use). 2

Mc. Call’s Quality Factors • Product Operation – – – Correctness Efficiency Integrity Reliability Usability • Product Revision – Flexibility – Maintainability – Testability • Product Transition – Interoperability – Portability – Reusability 3

Mc. Call’s Quality Factors 4

Mc. Call’s Software Metrics • Auditability • Accuracy • Communication commonality • Completeness • Consistency • Data commonality • Error tolerance • Execution efficiency • Expandability • Generality • • Hardware independence Instrumentation Modularity Operability Security Self-documentation Simplicity Software system independence • Traceability • Training 5

FURPS Quality Factors • • • Functionality Usability Reliability Performance Supportability 6

ISO 9126 Quality Factors • • • Functionality Reliability Usability Efficiency Maintainability Portability 7

Measurement Process - 1 • Formulation – derivation of software measures and metrics appropriate for software representation being considered • Collection – mechanism used to accumulate the date used to derive the software metrics • Analysis – computation of metrics 8

Measurement Process - 2 • Interpretation – evaluation of metrics that results in gaining insight into quality of the work product • Feedback – recommendations derived from interpretation of the metrics is transmitted to the software development team 9

Technical Metric Formulation • The objectives of measurement should be established before collecting any data. • Each metric is defined in an unambiguous manner. • Metrics should be based on a theory that is valid for the application domain. • Metrics should be tailored to accommodate specific products and processes 10

Software Metric Attributes • • • Simple and computable Empirically and intuitively persuasive Consistent and objective Consistent in use of units and measures Programming language independent Provides an effective mechanism for quality feedback 11

Representative Analysis Metrics • Function-based metrics • Bang metric – function strong or data strong • Davis specification quality metrics 12

Specification Quality Metrics - 1 nn = nf + nnf nn = requirements & specification. nf = functional. nnf = non-functional. Specificity, Q 1 = nai/qr nai = # of requirements with reviewer agreement. 13

Specification Quality Metrics - 2 Completeness, Q 2 = nu / (ni * ns) nu = unique functions. ni = # of inputs. ns = # of states. Overall completeness, Q 3 = nc / (nc + nnv) nc = # validated & correct. nnv = # not validated. 14

Representative Design Metrics - 1 • Architectural design metrics – Structural complexity (based on module fanout) – Data complexity (based on module interface inputs and outputs) – System complexity (sum of structural and data complexity) – Morphology (number of nodes and arcs in program graph) – Design structure quality index (DSQI) 15

Representative Design Metrics - 2 • Component-level design metrics – Cohesion metrics (data slice, data tokens, glue tokens, superglue tokens, stickiness) – Coupling metrics (data and control flow, global, environmental) – Complexity metrics (e. g. cyclomatic complexity) • Interface design metrics (e. g. layout appropriateness) 16

Halstead’s Software Science Source Code Metrics • Overall program length • Potential minimum algorithm volume • Actual algorithm volume – number of bits used to specify program • Program level – software complexity • Language level – constant for given language 17

Testing Metrics • Metrics that predict the likely number of tests required during various testing phases • Metrics that focus on test coverage for a given component 18

Estimating Number of Errors Error Seeding - 1 (s / S) = (n / N) S = # of seeded errors s = # seeded errors found N = # of actual errors n = # of actual errors found so far 19

Estimating Number of Errors Error Seeding – 2 E(1) = (x / n) = (# of real errors found by 1/ total # of real errors) = q/y= (# errors found by both / # real errors found by 2) E(2) = (y / n) = q / x = (# real errors found by both / # found by 1) n = q/(E(1) * E(2)) 20

Estimating Number of Errors Error Seeding – 3 Assume x = 25 y = 30 q = 15 E(1) = (15 / 30) =. 5 E(2) = (15 / 25) =. 6 n = [15 / (. 5)(. 6)] = 50 errors 21

Software Confidence S = # of seeded errors. N = # of actual errors. C (confidence level) = 1 if n > N C (confidence level) = [S / (S – N + 1)] if n <= N Example: N = 0 and S = 10 C = 10/(10 – 0 + 1) = 10/11 91% 22

Confidence Example How many seeded errors need to be used and found to have a 90% confidence a program is bug free? N=0 C = S/(S – 0 + 1) = 98/100 Solving for S S = 49 23

Failure Intensity • Suppose that intensity is proportional to # of faults or errors present at the start of testing. – function A has 90% of duty time – function B has 10% of duty time • Suppose there are 100 total errors 50 in A, 50 in B (. 9)50 K + (. 1)50 K = 50 K (. 1)50 K = 5 K or (. 9)50 k = 45 K 24

Maintenance Metrics Software Maturity Index SMI = [Mt = (Fa + Fc + Fd)]/Mt Mt = number of modules in current release. Fa = modules added. Fc = modules changed. Fd = modules deleted. • SMI approaches 1. 0 as product begins to stabilize 25