7  Statistical and graphical methods and the E5 framework for problem solving

This chapter introduces the data-based tools used to define, measure and diagnose problems — Voice of the Customer, Critical to Quality measures, line graphs, control charts, Pareto charts, the fishbone diagram and the 5 Whys — together with the development of countermeasures, the Balanced Scorecard, personality-based problem-solving preferences and the E5 framework for organisational problem solving.

7.1 Why these tools are introduced here

The statistical and graphical tools in this chapter — control charts, Pareto charts and fishbone diagrams — originate from quality improvement traditions, particularly Total Quality Management (TQM) and Six Sigma (Deming, 1986; Juran, 1951). They are included here not to import Six Sigma as a separate subject, but because these tools are the most widely validated methods for the measurement and root-cause-analysis phases of problem solving — essential disciplines for any organisation, including agricultural extension. The Six Sigma DMAIC cycle (Define → Measure → Analyse → Improve → Control) provides a logical organising framework showing which tool belongs to which phase of problem diagnosis (Table 7.1, Figure 7.1).

Table 7.1: The DMAIC cycle and the tools covered in this chapter

Phase	Core question	Tools covered
Define	What is the problem? What do customers need?	Voice of the Customer; CTQ measures
Measure	How bad is the current situation? What is the baseline?	Line graphs; control charts
Analyse	What are the root causes?	Pareto charts; fishbone diagram; 5 Whys
Improve	What countermeasures eliminate the root causes?	Countermeasure development
Control	How do we sustain the improvement?	Control charts; Balanced Scorecard

flowchart LR
  D[Define] --> M[Measure] --> A[Analyse] --> I[Improve] --> C[Control]
  C -. new problem .-> D

Figure 7.1: The DMAIC cycle.

7.2 Voice of the Customer (VOC)

The Voice of the Customer (VOC) is a structured process of capturing the expressed and latent needs, expectations and complaints of those served. In agricultural extension, the direct customers are the farmers who receive advisory services; indirect customers include government departments, funders and communities. VOC data are gathered through interviews, focus groups, surveys, direct observation and complaint analysis. Raw data are processed to extract specific customer requirements — statements of what is needed, free of solution language — which are then translated into measurable specifications through CTQ analysis.

7.3 Critical to Quality (CTQ) measures

CTQ measures translate qualitative customer language — ‘I want faster advice’ — into quantitative operational targets — ‘Response within 48 hours in 95% of cases.’ There are three levels:

1. Customer requirement — what the farmer expresses;
2. Drivers — the key dimensions that underlie the requirement; and
3. CTQ measure — a specific, measurable standard for each driver.

Without CTQ measures, quality improvement efforts lack direction and measurable targets.

7.4 Line graphs

A line graph plots data points collected at regular intervals and connects them sequentially, making trends, patterns and turning points immediately visible. In problem analysis, line graphs display baseline performance, monitor the trajectory of a problem and visualise the effect of an intervention. Upward or downward trends, flat lines, sudden spikes and cyclical patterns each carry a distinct diagnostic meaning. In agricultural extension, many problems have seasonal signatures that a line graph immediately reveals.

7.5 Control charts (statistical process control)

Invented by Walter A. Shewhart at Bell Laboratories in the 1920s (Shewhart, 1931), the control chart is the most analytically powerful graphical tool in this chapter. Its value lies in distinguishing two fundamentally different types of variation:

• Common cause variation: predictable, inherent variation that is always present in any process — the normal output of the existing system. It does not require corrective action; it requires system-level redesign if it is deemed unsatisfactory. Treating common cause variation as special cause variation creates instability and wastes effort (Deming, 1986).
• Special cause variation: unusual, sporadic variation triggered by a specific, identifiable external factor, which appears as data points outside the control limits. It demands investigation and correction.

Control limit calculation

\[ \text{CL} = \bar{x}, \qquad \text{UCL} = \bar{x} + 3s, \qquad \text{LCL} = \bar{x} - 3s \]

where \(\bar{x}\) is the mean of all data points (the Centre Line) and \(s\) is the standard deviation.

Worked example: average agent visit time = 24.9 min; standard deviation = 6.9 min.

• UCL = 24.9 + (3 × 6.9) = 24.9 + 20.7 = 45.6 min
• LCL = 24.9 − (3 × 6.9) = 24.9 − 20.7 = 4.2 min

Any day outside these limits is a special cause requiring investigation.

7.6 Pareto charts

The Pareto principle — that approximately 80% of effects arise from 20% of causes — guides practitioners to focus limited resources on the ‘vital few’ causes rather than dispersing effort across all possible causes (Juran, 1951). A Pareto chart displays categories of causes in descending order of frequency (bars on the left Y-axis) with a cumulative percentage line (right Y-axis) that rises from left to right. Categories to the left of the 80% threshold on the cumulative line are the vital few — the highest-leverage improvement targets. Stratification using a second Pareto chart then breaks the biggest category into sub-causes, connecting directly to the fishbone diagram.

Figure 7.2 applies the method to the problems reported by county extension agents in West Virginia (Table 5.2).

Figure 7.2: Pareto chart of the problems reported by county extension agents in West Virginia (1965), drawn from Table 5.2. Bars show the share of each problem area; the line shows the cumulative percentage, with the 80% threshold dashed.

7.7 Fishbone (Ishikawa) diagram

The fishbone diagram systematically maps all potential causes of a problem, with the problem (the ‘effect’) at the right and categories of potential causes branching off a central spine (Ishikawa, 1990). Kaoru Ishikawa formalised the technique at the Kawasaki shipyards; Mazda later used it proactively during the design of the Miata (MX-5) to prevent problems before they occurred. The most widely used framework is the 6Ms: Manpower, Methods, Machines, Materials, Measurement and Mother Nature (environment) (Figure 7.3). The fishbone is built by:

1. defining the effect;
2. selecting a category framework;
3. identifying the biggest contributor from the Pareto chart;
4. brainstorming sub-causes using the 5 Whys; and
5. prioritising candidate causes for data verification.

Figure 7.3: Skeleton of a fishbone diagram using the 6M framework. Sub-causes are added along each bone during brainstorming.

7.8 The 5 Whys technique

Developed by Sakichi Toyoda and central to the Toyota Production System (Ohno, 1988), the 5 Whys technique peels back layers of causation by asking ‘Why?’ repeatedly until an actionable root cause — not merely a symptom — is reached.

5 Whys example

Problem: ran through a red light.

1. Why? → Late for work.
2. Why? → Woke up late.
3. Why? → The alarm did not work.
4. Why? → The battery was exhausted.
5. Why? → Forgot to check it.

Root cause: no routine for maintaining personal equipment. Countermeasure: a monthly calendar reminder to check the alarm battery.

Lesson: what appeared to be a traffic violation was a failure of a personal maintenance habit. Issues that seem technical frequently turn out to be human and process problems.

Unvalidated root causes lead teams to pursue worthless countermeasures, wasting time and resources. Root causes identified through diagramming and questioning must therefore be verified with actual data (correlation analysis, designed experiments or stratified data analysis) before countermeasures are developed.

7.9 Developing and evaluating countermeasures

A countermeasure is a targeted action designed to eliminate a confirmed root cause. Each candidate countermeasure is evaluated against two criteria:

1. Effectiveness — how completely and permanently does it prevent the root cause from recurring?
2. Feasibility — how practical is it given the available resources, time and capacity?

Countermeasures are plotted on a two-axis prioritisation matrix; those scoring high on both axes are immediate priorities. After implementation, control charts and line graphs confirm whether improvement has occurred. If it has, the new process is standardised — documented, trained and embedded in operating procedures to prevent regression.

7.10 The Balanced Scorecard (BSC)

Developed by Kaplan & Norton (1992), the Balanced Scorecard translates organisational vision and strategy into performance measures across four perspectives:

• Financial: how do we appear to funders?
• Customer/farmer: how do we appear to those we serve?
• Internal process: what processes must we excel at?
• Learning and growth: how will we sustain our capacity?

The BSC ensures that operational improvements from TQM tools are aligned with strategic intent — diagnosing what is broken and fixing it in a way that advances organisational goals.

7.11 Personality types and problem-solving preferences

Research using the Myers–Briggs Type Indicator (MBTI) demonstrates that different personality dimensions align with distinct problem-solving strategies (Huitt, 1992). Extroverts prefer brainstorming and thinking aloud; introverts prefer private reflection and incubation. Sensing types favour inductive reasoning from facts; intuitive types prefer deductive reasoning and visualisation. Thinking types gravitate toward analysis and task breakdown; feeling types toward values clarification. Judging types prefer evaluation and backward planning; perceiving types prefer brainstorming and taking another’s perspective. For extension facilitators, this means structuring groups for cognitive diversity, allowing individual idea generation before group discussion, and recognising that different phases of problem solving benefit from different personality orientations.

7.12 The E5 framework: a comprehensive organisational problem-solving model

The E5 framework provides a structured, five-phase approach to solving complex organisational problems. It is distinguished by its emphasis on deep problem understanding before solution development — the most common failure point in organisational problem solving (Table 7.2).

Table 7.2: Phases of the E5 framework

Phase	Name	Core activity	Key tools
1	Expand	Framestorming — explore the problem from every angle before committing to any definition; diverse team; ‘What if…?’ and ‘How might we…?’ questions	Six Thinking Hats; brainstorming; VOC
2	Examine	Identify root causes via the Iceberg Model: surface events → patterns → systemic structures → mental models	Fishbone diagram; 5 Whys; Pareto charts
3	Empathize	Map all stakeholders; create empathy maps (Say/Think/Feel/Do); conduct interviews to gather real perspectives	Empathy mapping; stakeholder analysis
4	Elevate	Apply the Four-Frame Model (structural, human resources, political, symbolic) to connect the problem to organisational dynamics	Bolman and Deal’s Four-Frame Model
5	Envision	Design the future through backcasting: ideal future state → required milestones → prioritised action plan	Backcasting; Balanced Scorecard

The Iceberg Model used in the Examine phase draws on systems thinking (Section 5.3.1), in which visible events are traced to underlying patterns, structures and mental models (Senge, 1990).

7.12.1 The Four-Frame Model in the Elevate phase

The Four-Frame Model (Bolman & Deal, 2017) diagnoses organisational problems from four complementary perspectives:

• Structural: systems, roles, hierarchies and coordination mechanisms. Problems are diagnosed as misaligned structures or unclear roles.
• Human resources: people’s needs, motivations and relationships. Problems are diagnosed as unmet needs, skill gaps or leadership failures.
• Political: power, interests, coalitions and competition for resources. Problems are diagnosed as conflicts of interest or power imbalances.
• Symbolic: culture, values, rituals and shared meanings. Problems are diagnosed as cultural misalignment or erosion of trust.

7.12.2 Case study: Omega Soundscapes

Case study: Omega Soundscapes

Omega Soundscapes, a premium headphone manufacturer, experienced a significant decline in sales. Leadership assumed that a recent price increase was the cause. Applying the E5 framework revealed the following:

• Expand: framestorming generated alternative explanations, including quality decline, competition and brand perception.
• Examine: the Iceberg Model uncovered that a cost-cutting decision had substituted cheaper components and that brand overconfidence had reduced quality monitoring.
• Empathize: empathy maps showed customers, distributors and investors all experiencing different manifestations of declining perceived value.
• Elevate: the Four-Frame analysis revealed structural coordination failures between R&D and marketing, declining sales-team morale, rigid financial policies overriding quality concerns, and a gap between the brand’s self-image and customer experience.
• Envision: the team set a goal of 10% sales recovery within 12 months and backcasted a phased plan — quality audits and restoration (months 1–3), brand campaigns and R&D–marketing integration (months 4–8), and new product launches and distributor rebuilding (months 9–12).

Key reminder

Problem solving is not a straight line. New information at any phase may require returning to an earlier phase. The value of the E5 framework lies not in rigid sequence compliance but in the discipline of comprehensive problem understanding before commitment to a solution.

7.13 Summary: problem-solving models at a glance

The models and concepts covered in Chapters 5 to 7 are summarised in Table 7.3.

Table 7.3: Problem-solving models and concepts at a glance

Section	Model / concept	Key author(s)	Core contribution
Section 5.1	Problematic situation	Linhart (1976)	Static versus dynamic problem contexts; agricultural problems are inherently dynamic
Section 5.2	Five-step problem-solving process	Bormann & Bormann (1988)	Sequential framework; highlights people-problems and the danger of mental set
Section 5.3	Group problem solving and systems thinking	Checkland (1981); Adams & Galanes (2009)	Collective, participatory process within complex agricultural systems
Section 5.4	Problem solving versus decision making	Simon (1960); Kahneman (2011)	Clarifies the relationship; distinguishes programmed from non-programmed decisions
Section 6.1	Reflective thinking model	Dewey (1933)	Seven-step cyclical inquiry; evidence-based analysis and follow-up monitoring
Section 6.2	Socratic / dialectical model	Socrates	Questioning, contradiction and synthesis as routes to creative understanding
Section 6.3	Problem space model	Newell & Simon (1972)	Problem solving as heuristic search; expert chunking; limited working memory
Section 6.4.1	Six Thinking Hats	de Bono (1985)	Parallel thinking in six modes; reduces conflict; enriches group analysis
Section 6.4.2	Brainstorming	Osborn (1953)	Quantity-first idea generation; psychological safety; three approaches
Section 6.5	Lateral thinking and cognitive biases	de Bono (1970); Kahneman (2011)	Deliberate departure from vertical thinking; awareness and correction of biases
Section 7.2 – Section 7.9	VOC, CTQ, control charts, Pareto, fishbone	Shewhart (1931); Juran (1951); Ishikawa (1990)	Data-based tools for defining, measuring and diagnosing problems
Section 7.10 – Section 7.12	Balanced Scorecard and E5 framework	Kaplan & Norton (1992); Bolman & Deal (2017)	Strategic alignment and comprehensive organisational problem solving

7.14 Chapter Summary

Fill in the blanks

Answers are given at the end of the chapter.

1. The five phases of the Six Sigma DMAIC cycle are Define, Measure, Analyse, Improve and __________.
2. CTQ stands for Critical to __________.
3. The control chart was invented by Walter A. __________.
4. Variation that appears as points outside the control limits is called __________ cause variation.
5. If the mean is 24.9 min and the standard deviation 6.9 min, the upper control limit is __________ min.
6. The Pareto principle states that about 80% of effects arise from __________% of causes.
7. The 5 Whys technique is central to the __________ Production System.
8. The Balanced Scorecard was developed by Kaplan and __________.
9. The fourth phase of the E5 framework, which applies the Four-Frame Model, is called __________.

Short-answer questions

1. Explain the DMAIC cycle and the tools used in each phase.
2. How are Voice of the Customer data translated into CTQ measures? Give an extension example.
3. Distinguish between common cause and special cause variation. How are control limits calculated?
4. Explain how a Pareto chart is constructed and interpreted.
5. Describe the steps in constructing a fishbone diagram using the 6M framework.
6. Illustrate the 5 Whys technique with an example from extension work. Why must root causes be verified with data?
7. Explain the four perspectives of the Balanced Scorecard.
8. Describe the five phases of the E5 framework, using the Omega Soundscapes case.

Quick revision

• DMAIC → Define (VOC, CTQ), Measure (line graphs, control charts), Analyse (Pareto, fishbone, 5 Whys), Improve (countermeasures), Control (control charts, BSC).
• CTQ → customer requirement → drivers → measurable standard.
• Control chart (Shewhart) → CL = mean; UCL/LCL = mean ± 3 SD; common vs special cause.
• Pareto (Juran) → 80/20; vital few left of the 80% cumulative line.
• Fishbone (Ishikawa) → 6Ms: Manpower, Methods, Machines, Materials, Measurement, Mother Nature.
• 5 Whys (Toyoda; Toyota) → ask why until an actionable root cause; verify with data.
• Countermeasures → effectiveness × feasibility; standardise what works.
• BSC (Kaplan and Norton) → financial, customer, internal process, learning and growth.
• E5 → Expand, Examine, Empathize, Elevate, Envision.

Answers to fill in the blanks

1. Control 2. Quality 3. Shewhart 4. special 5. 45.6 6. 20 7. Toyota 8. Norton 9. Elevate

Adams, K., & Galanes, G. (2009). Communicating in groups: Applications and skills (7th ed.). McGraw-Hill.

Bolman, L. G., & Deal, T. E. (2017). Reframing organizations: Artistry, choice, and leadership (6th ed.). Jossey-Bass.

Bormann, E. G., & Bormann, N. C. (1988). Effective small group communication (4th ed.). Burgess.

Checkland, P. (1981). Systems thinking, systems practice. John Wiley & Sons.

de Bono, E. (1970). Lateral thinking: Creativity step by step. Harper & Row.

de Bono, E. (1985). Six thinking hats. Little, Brown; Company.

Deming, W. E. (1986). Out of the crisis. MIT Press.

Dewey, J. (1933). How we think: A restatement of the relation of reflective thinking to the educative process. D. C. Heath.

Huitt, W. G. (1992). Problem solving and decision making: Consideration of individual differences using the Myers–Briggs type indicator. Journal of Psychological Type, 24, 33–44.

Ishikawa, K. (1990). Introduction to quality control. Productivity Press.

Juran, J. M. (1951). Quality control handbook. McGraw-Hill.

Kahneman, D. (2011). Thinking, fast and slow. Farrar, Straus; Giroux.

Kaplan, R. S., & Norton, D. P. (1992). The balanced scorecard: Measures that drive performance. Harvard Business Review, 70(1), 71–79.

Linhart, J. (1976). Psychologie učenı́ [psychology of learning]. Academia.

Newell, A., & Simon, H. A. (1972). Human problem solving. Prentice-Hall.

Ohno, T. (1988). Toyota production system: Beyond large-scale production. Productivity Press.

Osborn, A. F. (1953). Applied imagination: Principles and procedures of creative thinking. Scribner.

Senge, P. M. (1990). The fifth discipline: The art and practice of the learning organization. Doubleday.

Shewhart, W. A. (1931). Economic control of quality of manufactured product. Van Nostrand.

Simon, H. A. (1960). The new science of management decision. Harper & Row.