4  Methodologies for decision making in extension organisations

This chapter describes the main methodologies used to structure decisions in agriculture and extension organisations: decision analysis and decision trees, the Delphi technique, brainstorming, the cause-and-effect (fishbone) diagram, data-driven decision making and decision support systems.

4.1 Decision analysis and decision trees

4.1.1 Decision analysis

Decision analysis is the systematic evaluation of alternatives using quantitative and qualitative tools. It involves defining objectives, identifying alternatives, assessing outcomes and selecting the best option (Hardaker et al., 2015). Decision analysis is applied both at farm level — deciding which crop to grow — and at organisational level — determining which extension programme to fund.

4.1.2 The decision tree

A decision tree is a visual, tree-shaped diagram that maps decisions, chance events and their possible outcomes. Its key components are:

• Decision nodes (squares) — where the decision maker chooses among alternatives;
• Chance nodes (circles) — where outcomes depend on probability;
• Branches — representing decisions or chance outcomes; and
• Terminal nodes — end points showing the final payoff.

4.1.3 Method

1. Define the decision problem clearly.
2. Identify all possible decision alternatives.
3. Identify chance events and assign probabilities.
4. Assign monetary values or utility scores to each terminal node.
5. Calculate the Expected Monetary Value (EMV): multiply each probability by its outcome value and sum the products.
6. Select the decision branch with the highest EMV.

Agricultural example: to irrigate or not

A farmer decides whether to irrigate paddy or rely on rainfall (Figure 4.1).

• Irrigation: cost Rs. 5,000; guaranteed yield of 4 t/ha regardless of rain.
• Rainfall: 60% chance of adequate rain (yield 3.5 t/ha; revenue Rs. 35,000) and 40% chance of drought (yield 1.5 t/ha; revenue Rs. 15,000).

EMV of relying on rainfall = (0.6 × 35,000) + (0.4 × 15,000) = Rs. 21,000 + Rs. 6,000 = Rs. 27,000.

At the same price (Rs. 10,000 per tonne), irrigation gives a certain revenue of Rs. 40,000, or Rs. 35,000 net of its cost. The decision tree allows the farmer to compare the two branches directly; in this example irrigation is the better choice.

flowchart LR
  D[Irrigate paddy?] -->|Irrigate<br>cost Rs. 5,000| T1(["4 t/ha: Rs. 40,000<br>Net Rs. 35,000"])
  D -->|Rely on rainfall| C((Rain))
  C -->|Adequate, p = 0.6| T2(["3.5 t/ha: Rs. 35,000"])
  C -->|Drought, p = 0.4| T3(["1.5 t/ha: Rs. 15,000"])
  C -.- E["EMV = Rs. 27,000"]

Figure 4.1: Decision tree for the irrigation example. The square is the decision node, the circle a chance node, and the rounded boxes are terminal payoffs.

4.2 The Delphi technique

The Delphi technique is a structured expert consultation method used to achieve consensus on complex issues through multiple rounds of anonymous questionnaires (Linstone & Turoff, 1975). It is particularly useful when there is no single correct answer and when expert judgement is required to address a question on which objective data are scarce.

4.2.1 Method

1. Define the problem or question clearly.
2. Select a panel of relevant experts (farmers, scientists, extension specialists, policymakers).
3. Distribute the Round 1 questionnaire anonymously.
4. Analyse responses and identify areas of agreement and disagreement.
5. Prepare a feedback summary and distribute the Round 2 questionnaire.
6. Repeat rounds until consensus emerges or positions stabilise.
7. Compile a final report reflecting consensus or documented divergence.

4.2.2 Advantages and limitations

The strengths and weaknesses of the technique are summarised in Table 4.1.

Table 4.1: Advantages and limitations of the Delphi technique

Advantages	Limitations
Anonymity reduces social pressure and dominant-voice effects	Time-consuming across multiple rounds
Structured feedback fosters reflection and refinement	Requires skilled facilitation
Useful when no objective data exist	Consensus may mask genuine diversity of opinion

4.3 Brainstorming in decision making

Brainstorming — covered in detail as a problem-solving technique in Section 6.4.2 — is also used as a decision-making technique when generating decision alternatives, identifying barriers to adoption and developing extension programme ideas. Its core rules are to defer judgement, encourage wild ideas, build on others’ ideas and go for quantity (Osborn, 1953). In extension it is used for identifying new programme approaches, solving field problems encountered by agents, staff planning and annual review sessions, and identifying barriers to technology adoption.

4.4 Cause-and-effect (fishbone) diagram in decision making

The fishbone diagram — also covered as a problem-solving tool in Section 7.7 — serves a complementary role in decision making by structuring the diagnosis of why a decision problem exists (Ishikawa, 1990). Using the 6M framework (Man, Machine, Method, Material, Measurement, Mother Nature), it organises potential causes systematically before decision alternatives are generated. In extension it is used for diagnosing why a technology adoption programme underperformed, investigating causes of low crop yield, analysing reasons for poor attendance at training programmes, and improving the quality of service delivery.

4.5 Data-driven decision making (DDDM)

Data-driven decision making (DDDM) refers to basing decisions on the analysis of data rather than on intuition alone (Jones et al., 2017). In agricultural extension, DDDM enables evidence-based programme design, resource allocation and performance monitoring. The main data sources in agricultural extension are:

• farm surveys and household data;
• remote sensing and satellite imagery;
• weather and climate data from meteorological agencies;
• market price data and value chain analysis; and
• extension programme monitoring and evaluation reports.

Decision support built on these sources may be data-driven or document-driven (Table 4.2).

Table 4.2: Data-driven and document-driven decision support systems

Feature	Data-driven DSS	Document-driven DSS
Primary input	Quantitative data; databases	Reports; policy documents; research publications
Analysis type	Statistical; computational	Qualitative; interpretive
Strengths	Precision; scalability; real-time updates	Contextual richness; policy relevance
Example tools	DSSAT; AquaCrop; remote sensing platforms	Knowledge management systems; digital libraries

4.6 Decision support systems (DSS) in agriculture

A decision support system (DSS) is an information system that integrates databases, simulation models and user interfaces to assist farmers, extension agents and policymakers in making informed decisions (Jones et al., 2017). Some widely used systems are listed in Table 4.3.

Table 4.3: Selected decision support systems used in agriculture

System	Developer	Primary function
DSSAT	University of Florida / ICASA	Crop simulation; yield forecasting; agronomic management
AquaCrop	FAO	Water productivity and irrigation management
SARRA-H	CIRAD	Cereal crop yield forecasting in semi-arid areas
iSAT	IFDC	Integrated soil and agronomic technology recommendations
Digital advisory platforms (e.g., farmer.CHAT)	Various	AI-based personalised extension advisory

4.7 Chapter Summary

Fill in the blanks

Answers are given at the end of the chapter.

1. In a decision tree, decision nodes are drawn as __________ and chance nodes as circles.
2. EMV stands for Expected __________ Value.
3. The Delphi technique achieves consensus through multiple rounds of __________ questionnaires.
4. The fishbone diagram commonly uses the __________ framework to group causes.
5. DSSAT, a crop simulation system, was developed by the University of Florida and __________.
6. AquaCrop, used for water productivity and irrigation management, was developed by __________.

Short-answer questions

1. What is decision analysis? List the components of a decision tree.
2. Using the irrigation example, show how the EMV of each branch is calculated.
3. Describe the steps of the Delphi technique. What are its advantages and limitations?
4. How is the fishbone diagram used in decision making in extension?
5. What are the main data sources for data-driven decision making in extension?
6. Distinguish between data-driven and document-driven decision support systems.

Quick revision

• Decision analysis → objectives, alternatives, outcomes, best option.
• Decision tree → squares (decisions), circles (chance), branches, terminal payoffs; choose the highest EMV.
• EMV → Σ (probability × outcome value).
• Delphi (Linstone and Turoff) → anonymous expert rounds with feedback until consensus.
• Brainstorming (Osborn) → defer judgement, wild ideas, build on others, quantity.
• Fishbone (Ishikawa) → 6Ms; diagnose why before choosing.
• DDDM → decisions from data; DSS → databases + models + interfaces (DSSAT, AquaCrop, SARRA-H, iSAT).

Answers to fill in the blanks

1. squares 2. Monetary 3. anonymous 4. 6M 5. ICASA 6. FAO

Hardaker, J. B., Lien, G., Anderson, J. R., & Huirne, R. B. M. (2015). Coping with risk in agriculture: Applied decision analysis (3rd ed.). CABI.

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

Jones, J. W., Antle, J. M., Basso, B., Boote, K. J., Conant, R. T., Foster, I., Godfray, H. C. J., Herrero, M., Howitt, R. E., Janssen, S., Keating, B. A., Munoz-Carpena, R., Porter, C. H., Rosenzweig, C., & Wheeler, T. R. (2017). Brief history of agricultural systems modeling. Agricultural Systems, 155, 240–254. https://doi.org/10.1016/j.agsy.2016.05.014

Linstone, H. A., & Turoff, M. (Eds.). (1975). The delphi method: Techniques and applications. Addison-Wesley.

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