Measurement and Analysis - KU-SKE17/Software-Process GitHub Wiki

Measurement Components

  • Measurement Object
  • Measurement Techniques

Measurement Flow

  • (loop)
  • Collect Data
  • Store Data (database, data warehouse)
  • Analysis -> get outcome
  • Update Analysis

Measurement Scales

Nominal Scale

  • choosing between choices (objects)
  • A or B
  • blue, red, or green (Which one you prefer?)

Ordinal Scale

  • choosing non-number scale
  • happy, normal, bad
  • mostly about emotion

Interval Scale

  • choosing number scale
  • (worst) 1, 2, 3, 4, 5 (best)
  • you will know the different between choices

Ratio Scale

  • choosing from range
  • 1-2, 3-4, more then 5
  • found mostly in survey

Types of Data

Quantitative

  • Number
    • Discrete - [1, 2, 3, ...]
    • Continuous - func, line-graph
      • Interval (time)
      • Ratio (can be determined/grouped) - height, weight

Qualitative

  • Non-Number - yes/no, color
    • Nominal - Color of eye
    • Ordinal (can convert to number) - feeling (happy = 1, unhappy = 0)

Types of Measurement/KPI

Quality Measures

  • Code Quality
    • maintainability index
    • depth of inheritance
    • line of code
  • Testing Quality
    • test coverage
    • defects found in UAT

Process Measures

  • (for project manager)
  • Burndown Rate
  • APIs Collaborate Rate

Outcome Measures

  • Solution Availability
    • software performant
    • page load time
  • Solution Security
    • number of unavailability cases
    • number/severity of security incidents
  • User Satisfaction
    • survey, ranking
    • software stability

Top 4 Analysis Techniques

Regression Analysis

  • Quantitative [Number] - hard!
  • relation btw 2 variables

Hypothesis Analysis

  • Quantitative [Number] - hard!
  • hypothesis is true or not

Content Analysis

  • Qualitative [Non-number]
  • see the trends by time series
  • Example:
    • time series analysis
    • parametric tests (normal graph)
    • summary statics

Discourse Analysis

  • Qualitative [Non-number]
  • check social context
  • Example:
    • context analysis

MA 1 Align Measurement and Analysis Activities

  • Design framework to analysis
  • focus on class workshop
  • have 4 steps

MA 1.1 Establish Measurement Objectives

➢ Create KPI

  • Objectives in term of bushiness = KPI
  • KPI: Key Performance Indicator
  • Example
    • Reduce Time to Delivery
    • Reduce Lifecycle Cost
    • Improve Quality
    • Improve Customer Satisfaction

MA 1.2 Specify Measures

➢ Choose 'What are the things that need to measure?'

  • Example
    • Delivered Solution Quality
    • Code Quality
    • Process Quality
    • User Satisfaction

MA 1.3 Specify Data Collection and Storage Procedure

➢ Collection/Store data

  • What to be collected?
  • How to collect?
  • [next week] Where/How to store the data?

MA 1.4 Specify Analysis Procedures

➢ Manage [sort, classify, extract patters, ...] date before analysis

MA 2 Provide Measurement Results

  • How to use what we designed from MA 1
  • have 4 steps

MA 2.1 Obtain Measurement Data

➢ Set Timeframe for data collection

➢ Collection the data

MA 2.2 Analyze Measurement Data

➢ Analyze data (follow your MA 1.2)

➢ using analysis techniques

MA 2.3 Store Data and Results

➢ Store Data

➢ Data Cleaning

➢ Data Transformations

➢ (real) Analysis

➢ Reports/Dashboards

➢ Store all of above items!

  • more [next week]

MA 2.4 Communicate Results

➢ Make Report

  • Normal Report (Excel)
  • Dashboard (Tableau, Power BI)
  • Interactive Data Visualization