Planning Management of Project Quality


Plan qualityRight off the bat, let us try and answer this question.

What is the difference between Quality Assurance and Quality Control?

Assuring that processes and methodologies used on the project are in line with those prescribed by the organization by means of quality audit; and getting feedback for improvements on organizational quality processes is all that you do in Quality Assurance activities.

Controlling quality on the deliverables, by means of running tests, identifying defects and raising change requests – are what you do in Quality Control activities.

Like all other project management domain knowledge areas, you need a plan to ensure quality of project deliverables. This plan talks about quality requirements and standards to be used to assess quality of deliverables as per documented requirements and to show how the project demonstrates this quality compliance. This plan is called Quality Management Plan – and is the primary outcome of this project management activity.

If the performing organization does not contain a quality policy, then project management team can come up with one and assign it to be the organizational policy. This would be then strengthened and formalized through quality assurance processes on various projects.

What do I need?

Project management plan, of course

By now you would have noticed that Project management plan is a required artifact for all planning related project management activities. For this one though specifically the baselines are to be studied –

Scope baseline

  • Project scope document – contains the project description, major deliverables, acceptance criteria and so on.
  • Work breakdown structure and details – scope broken down into smaller packages

Schedule baseline – Start and end dates of activities, their dependencies and resource requirements

Cost baseline – Project cost expenditures and time interval used to measure cost performance

Stakeholder register

Why should stakeholder register be required as an input to Plan Quality process?

Stakeholder register contains information about people, their position in the organization, role in the project and particular interests in the project including expectations on quality of deliverables. These are to be studied to make sure that stakeholders expectations and needs are met through quality management activities.

Risk register

This contains information about threats and opportunities, which may impact quality requirements.

Requirements documentation

Stakeholder’s needs are documented here, which include quality requirements as well. These need to be studied in order to satisfy them in project deliverables.

Others

Any regulations put forth by industry (example: HIPAA in Healthcare) or government (example: environmental clearance for infrastructure projects), organizational policies and procedures related to quality, quality metrics from earlier projects, lessons learnt and tips from earlier projects.

How do I do this?

Cost-benefit analysis

This is about comparing the cost of performing each quality activity and the benefits it provides. At the end of the exercise you could decide which activities to include, as well as priorities of them – so that performing these activities will provide maximum benefit to the project.

The primary advantages of this analysis is reduced rework, higher productivity, lower cost and increased stakeholder satisfaction – which is priceless, actually. You can also see that most of these are indirect benefits unlike the effort on development activities produce. However, all effort related to quality produce indirect benefits. Benefit of not having defects is what you get for the effort.

Cost of quality (COQ)

Cost of activities to ensure conformance to requirements, and cost arising out of non-conformance of produced deliverables to the requirements is the Cost of Quality. Again, Pareto’s 80-20 rule is applicable here as well – 80% of the issues are due to 20% of the reasons. If these 20% reasons are identified and addressed upfront, the returns on investment on quality activities will be realized up front.

Here’s a quick look at the reasons and impact of these causes –

Cost of QualityFigure 1: Cost of Quality – reasons and impact

Seven basic tools of quality

These are also called 7QC tools. Each of these are explained below –

1. Cause-and-effect diagrams

These are also called Ishikawa diagrams or fishbone diagrams, as we saw in Identify Risks process. This is a popular root cause diagram, where a root of a cause of problem can be traced by continuously trying to answer ‘why’ along each of the lines of symptoms.

Refer to the diagram below. Wrong measurements are one of the symptoms of Jogging track defects. And asking “why” has led to identifying that calibration of tools has been the root cause of this issue.

Fishbone diagramFigure 2: Fishbone diagram

2. Flowcharts

are used to show relationship between steps in a processes. Activities, decision points and order of processing are main constituents of flowchart. They can unearth potential problems in quality processes and help plug any gaps.

flow chart
Figure 3: Simple flowchart

3. Checksheets

Checksheets are also called tally sheets. These are good to use as a check list to collect information and group them together. While inspecting deliverables checklists are useful to make sure all attributes are verified.

4. Pareto diagrams

are also called Pareto charts, named after Wilfred Pareto. This type of chart contains both bars and line. While bars represent frequency of occurrences (or some unit of measure), line graph represents cumulative values. Remember that bars are arranged in descending order of their heights (values) and line graph will always represent a concave curve. Here is a sample Pareto chart for Kathy’s Landscaping project from Procurement management knowledge area

Remember Pareto chart as his famous 80-20 rule, which states that roughly 80% of effects are caused by 20% of reasons. When a Pareto chart is plotted representing reasons for a particular problem, if you take care of top 20% of causes, they will address about 80% of the problems!

pareto chart

Figure 4: Pareto Chart For Kathy’s Landscaping project

5. Histogram

This is a graphical representation of distribution of data, shown as vertical bars representing characteristics of problem. The y-axis can show probability of the characteristic happening during problem situation, or simply the number of times this characteristic materializes during the problem situation.

6. Control charts

These are used to monitor repetitive quality processes for deviation from the mean. Control charts are also known as Shewhart charts (named after Walter A Shewhart) and are useful to verify whether a business process is in a state of control (or ‘stable’).

Each process under observation is defined upper and lower control limits and specification limits. If a process point falls outside of the control limit then action needs to be taken to bring it back the process within the control limits. Control limits indicate the maximum acceptable results without taking any corrective action on the process.

control-chartFigure 5: Control Chart

If seven consecutive points are found on any one side of the mean, or when a single data point is out of any one of the control limits the process is said to be out of control. This necessitates quick action by the project manager to see whether there is an issue in the product development process and get things back on track.

7. Scatter diagram

Also called Scatter Plot or Scattergraph is a type of mathematical diagram using Cartesian coordinates to display relation between two variables. One would be dependent variable and the other independent variable. Y-axis is used to plot the variable to be predicted (dependent variable) and X-axis is used for the variable to make prediction (independent variable).

Correlation refers to the degree of relationship between two variables. Degree of correlation can be found out looking at the plot. If the points are fairly random then correlation between variables is less (or ‘null’). If pattern of dots slope from lower left to upper right then it is called ‘positive’ correlation, and if pattern of dots slope from upper left to lower right then it is called ‘negative’ correlation.

An example of Scatter diagram is plot of two sets of variables, one representing height of an athlete (independent variable, on X-axis) and another representing weight (dependent variable, on Y-axis) – to find whether height has any correlation with weight of an athlete.

scatter diagramFigure 6: Scatter Diagram

As you can see, there is a ‘positive’ correlation between these variables. It appears that as height of an athlete increase she tends to weigh more.

These are the 7 basic tools of quality.

Benchmarking

This indicates the practice of comparing another project of similar type and nature from the past to the current one to identify processes or practices, or even to find better ways to do things.

This method is typically used in industries where testing is destructive in nature and/or leads to additional cost – such as inspecting nails, or inspecting ready-made shirts.
Sample sizes and frequency of tests are determined during Plan Quality process so that cost of quality will consider amount of scrap and number of tests.

Design of experiments (DOE)

DOE is used to figure out the number and types of tests and impact of their output on cost of quality. This is a statistical method. Each of the several identified tests are performed many number of times and the results are statistically analyzed to identify which of the tests help measure quality of deliverables most accurately.

Statistical sampling

Statistical sampling is a sampling technique where a representational sample is picked from the population and this sample is thoroughly tested using the designed tests. If number of defects found cross the acceptable limit in this sample, entire population is rejected.

For instance, out of 10000 shirts (population) a batch of 100 shirts is randomly selected. Each of these shirts are tested for manufacturing defects. If the threshold is say 3 shirts, then 3 defective shirts at the most are allowed in this batch of 100 shirts. If more than 3 defective shirts are found, then entire population of 10000 shirts are rejected.

Random samplingRandom samplingFigure 7: Statistical sampling flow

This method is typically used in industries where testing is destructive in nature and/or leads to additional cost – such as inspecting nails, or inspecting ready-made shirts.

Sample sizes and frequency of tests are determined during Plan Quality process so that cost of quality will consider amount of scrap and number of tests.

Additional quality planning tools

Again, which one of these are to be used is mentioned in Quality Management Plan.

a)   Brainstorming – is a technique where several people discuss a problem and potential solutions. They discuss the pros and cons of each and choose the one that suits them the best.
b)   Force field analysis is used to measure the forces for and against a change.
c)    Nominal group techniques – we looked at this in Collect Requirements process.
In short, this is a technique where a group of stakeholders are allowed to generate ideas silently. Then listed ideas are discussed as a group, voted and prioritized. The top ideas are then considered for action. This is a facilitator driven group decision making technique.
d)   Quality management and control tools  – these represent the tools used in this process as well as Control Quality process, and we shall look at them in detail in Control Quality process.

What is produced here?

Quality management plan

This describes how the project implements performing organization’s quality policies and methodologies. This plan defines,

  • how quality is measured on the project,
  • what quality control and assurance activities are conducted and when they are conducted,
  • what are the acceptable quality standards.

Process improvement plan

This is a way to contribute back to organizational quality processes, based on the lessons learnt on the project. This plan indicates how you can critically look at your processes, identify areas of improvements, define targets, metrics to measure improvements, and identify the person on the team responsible for process improvement.

Process improvement plan is a subsidiary plan of project plan.

Quality metrics

Robin Sharma in his book The Greatness Guide says, ‘what gets measured gets improved’. Quality metrics is that WHAT of the product that you need to measure in order to improve its performance.

For instance, let us say that in a statistical sampling exercise you figure that 12% of shirts on an average are defective. This is a great measure, because it sets you thinking to further dig and find the source of the issue. Your quality control team tells you that 80% of these defects are due to improper stitching of buttons or button holes. You have drilled down the reason to be around the buttons. Now you dig into the automated process that stitches buttons on the shirts. This way you can drastically reduce the defects, and corresponding costs of nonconformance that we saw in cost of quality.

Then again, what gets measured gets improved.

Quality checklists

Checklists can be for prevention as well as inspection. You can look at earlier similar projects, lessons learned from them and come up with initial checklists. As project progresses this can be updated whenever a new error patterns are discovered. The objective of checklists is to prevent defects from occurring, and when they do occur, how quickly they can be discovered.

Updating some of documents

This may be required whenever you find that some impacted document needs updates due to this quality plan exercise.

Plan Quality is quite an exhaustive process and usually done along with QA lead, and with help from performing organization’s quality assurance team.

If you have worked in smaller companies or start-ups like I do, you might have found that you need to start on some of the activities on your own without much help from organizational process assets. You may even end up preparing the quality plan that sort of becomes the organizational quality policy for other projects to follow.

Recommended by Google

{ 2 comments… add one }
  • Tina January 6, 2015, 5:19 am

    Hi Shiv,
    GREAT article on Plan Cost Management. Your discussion of the 7QC tools was great and I found the examples you provided really helped drive home the concepts much better than the PMBOK.

    In one example, is there a typo? It is the one about Statistical Sampling:
    “For instance, out of 10000 shirts (population) a lot of 100 shirts is randomly selected. Each of these shirts are tested for manufacturing defects. If the threshold is say 3 shirts, then 3 defective shirts at the most are allowed in this lot of 50 shirts. If more than 3 defective shirts are found, then entire population of 10000 shirts are rejected.”

    I think that it is the same lot of shirts – either a lot of 100 shirts, or a lot of 50 shirts. So, if I am reading this right, the 100 needs to be changed to 50, or vice versa.

    Do you still find these comments helpful?

    And thank you for connecting via Google and LinkedIn!
    Best,
    Tina

    Reply
    • Shiv Shenoy January 6, 2015, 9:35 am

      Hi Tina,
      Yes that was a typo, I’ve corrected it now. Thanks for pointing it out! 🙂
      Regards,
      Shiv

      Reply

Leave a Comment