Cover Image
close this book Agricultural extension
close this folder Providing agricultural support services
close this folder Direct services
View the document Testing recommendations
View the document Administering credit
View the document Selecting and producing seed
View the document Providing farm inputs
View the document Surveying agricultural land
View the document Providing storage
View the document Marketing agricultural products

Testing recommendations


As the middle link in the Research-Extension-Farming chain, it is the extension agent's responsibility to make sure recommendations to farmers are relevant, benefical and appropriate. Given the gap between researchers and small-scale farmers in most developing countries, recommendations often must be tested in the field by extensionists.

Why is it necessary to test at all? First of all, agriculture is an extremely location- and time-specific endeavor. There are too many major variables in agricultural situations to consider specific practices to be universally applicable. For example, a certain type of maize may be well suited to early rainy season cultivation in one area of Honduras while 100 miles away the local climate, soil characteristics and pest problems may make it a poor choice. Therefore, even practices developed by researchers for one region cannot be described as clearly suited to other local settings until they are tested thoroughly.

There are other reasons why recommendations must be tested: The extensionist herself must have confidence in a practice she is to recommend, so she must see herself the results of its use. This is also true for farmers. Unless a farmer sees a practice work well in his own locale, or hears that it works well from a reliable source, caution holds him back from employing the practice. Lastly, local testing of recommendations helps arouse farmer interest in new practices and enables farmers themselves to take part in the process of testing and formulating new practices.

In Chapter One, there is an outline of the steps in the Research Extension-Farmining chain. They can be summarized as follows:

1. Pure research

2. Basic research

3. Applied research

4. Adaptive research

5. Extension work

6. Farmer use and feedback

Testing recommendations in the field involves a combination of the adaptive research (4) and extension work (5) steps. In developing countries especially, these steps overlap and most often are handled by the extensionist herself. The extensionist plays a dual role (field research/extensionist). Because they rarely have the scientific training to do research without assistance, however, extensionists are urged to conduct tests in collaboration with local or regional research stations and staff. It is absolutely essential that all testing be done within the context of the knowledge, experience and interests of local farmers who are the most important experts in local situations.

Adaptive research


A Closer Look at Adaptive Research: How Yield Improving Practices are Tailored to Local Conditions

a. Small-plot experiments: These should be conducted at both the national station and the regional substations and are designed to test promising leads gained from applied research. In the case of most developing countries, this applied research may well have taken place in another country. The small plot experiments point out the more promising of these leads which may again be subjected to further small plot tests. An example would be the testing of a number of promising crop varieties. The very best of the promising leads become tentatively recommended practices.

b. The field experiment (farm experiments): Those practices showing the best results at the national and/or regional substations are now tested at scattered locations within the region to measure their performance under more varied soil and climatic conditions. They are tested at various levels of application and in varying combinations. Those practices with the best results become general recommended practices for the area involved. In short, these field experiments are the means for localizing general recommendations.

An experiment is a test that compares two or more treatments (e.g. two or more practices or varying rates of an input like fertilizer). Its design and management are based on rigidly standardized scientific procedures for assuring accurate unbiased results that are not influenced by extraneous factors. The results are subjected to a rigorous statistical analysis to determine if there is really any significant difference between treatments.

Because of the highly technical nature of such experiments, extensionists are not encouraged to carry out either small-plot experiments or field experiments themselves, unless absolutely necessary and then with tremendous technical support and conservatism with regard to results. Adaptive research steps like these are the final research efforts which adapt a practice to local conditions. As such, the extensionist role is to make sure they occur, scrutinize their appropriateness, and assist where necessary in their completion.

c. The field test or trial (the result test): This type of test or trial is usually conducted by extension workers in cooperation with participating farmers. Here the practice or "package" of practices is tested under realistic local farm conditions to determine its true range of profitability for the farmer. A new practice cannot claim to be proven until it passes the result test, which provides the ultimate basis for making a specific recommendation.

The result test is neither an experiment nor a demonstration. Only one variable selected for testing (e.g. a new practice or a specific "package" of practices) and is compared with the present or "traditional" practice. For example, a result trial would be used to determine the profitability of a given rate of fertilizer like pounds of manure and compost per square meter. On the other hand, a field experiment would be used to determine the response to several or more different fertilizer rates. The result tests are designed to obtain information about a practice, not to promote it. They are conducted on farms, but the purpose is to prove the worth of the practice to the extension worker and the extension agency, not to the cooperating farmers. The two treatments (e.g. the "old" and the "new") are not randomized and replicated as with an experiment; rather, the result test is repeated simultaneously on a number of local farms, since the goal is to get an overall idea of the new practice's performance in the area. Extensionists with good training in the reference crop or animal and the relevant practices can competently supervise result tests, and this can be one of their most valuable activities.

Extension begins when the emphasis shifts from testing to promotion, although there are areas of overlap between adaptive research and extension, especially in situations where resources are limited. Extensionists often visit national experiment stations in search of worthwhile practices to promote, which have undergone adaptive research. It is then up to the extensionist to promote these practices through result tests and demonstrations. The result test can be viewed as either the last step of adaptive research or the first step of the extension process.

d. The demonstration (result demonstration): This is not a testing procedure like the result test. Its objective is to demonstrate the profitability of a proven (locally tested) practice under actual farm conditions. If the new practice requires a change in traditional procedures, a good amount of in-thefield instruction may be needed along with the demonstration. When groups of visiting farmers receive such instruction as well as view the plot, the demonstration becomes a method-result demonstration. The extension worker should approach farmers from this standpoint: "We have good reasons for believing this practice is profitable and would like to help you prove it to yourself". A demonstration plot like this can be set up by an extensionist soon after arrival in a new locality. It should be laid out in a prominent place in order to arouse interest.

Method/result demonstrations, workshops during which farmers actually practice specific techniques or innovations, are the major extension tool for promotional purposes. See Chapter Four, FARMER TRAINING, for a detailed description of method-result demonstrations.

In all of these techniques for testing the suitability of recommendations to farmers, the extensionist is asked to consider the limits of her technical skill and the limitation of her point of view. Conservative, adequately-tested recommendations serve farmers best.


Lay out: a result test

A result test




1. Factors which Decrease the Need for Extensive Local Testing of a General Recommendation before Promoting It

a. Local growing conditions are uniform over the work area (soils, climate, management level, etc.)

b. The adaptive research on which the general recommendation is based was technically adequate and was subjected to a rigorous analysis (including statistical analysis).

c. The adaptive research took place under growing conditions similar to those of the work area.

d. The new practice represents a single factor change (only one new input or change).

e. Capital requirements are low.

f. The potential benefits are high and not subject to great variation.

g. No changes in growing practices are needed.

h. The extension agent has had prior experience with the new practice.

i. The new practice reduces costs or shortages by replacing a higher priced or less available input with a lower-priced or more available one. Examples: Substituting an animal-drawn cultivator for laborious hand weeding.



2. Some New Practices and their Relative Need for Extensive Local Testing Prior to Promotion

a. A new (improved) crop variety: Very extensive local testing needed along with detailed analysis of the results.

b. Fertilizer use: Low to moderate rates can be recommended on the basis of limited local testing, lab soil tests, and diagnosis of obvious visual hunger signs. Higher rates should be based on local farm experiments, result trials, and individual farm soil testing. Considering the many local variations in soil fertility, general recommendations often result in the application of too much or too little fertilizer or the wrong nutrient combination. For example, applying only N + O soil also deficient in P (phosphorus) may give the farmer only 25% of the yield response obtained when both are applied.

c. Mechanization: Depends on the model and type of equipment. Appropriate small scale equipment usually has a wider adaptation than most new practices, although soil and weather can affect performance.

d. Change in crop rotation or a new cropping system: Very extensive testing is needed (at least several years).

e. Disease and Insect Control: Chemical and cultural methods have much wider adaptation than biological ones. At least some limited testing should be done with specific chemical and cultural controls before promoting them to farmers.

f. Chemical weed control: Effectiveness varies greatly with different soils and weather conditions; local testing should always precede promotion.

g. Irrigation practices: Feasibility studies conducted by experienced technicians should always precede the installation of a new irrigation system; possible negative environmental effects must also be examined (i.e. salinization, ground water depletion, malaria, bilharzia, etc.)

h. Introduction of a new crop: Very extensive testing is required.




3. How to Set Up A Result Demo Plot

I. Select an appropriate practice or "package" to demonstrate

a. Given a lack of local experience and time to adequately test practices, rely on the local extension service to provide appropriate choices; always check to see if adequate local testing has been done; the amount needed will vary with the practice, i.e. selecting an adapted improved crop variety requires much more lengthly local testing than the use of mulch or insecticides.

b. One practice vs. a package: Although a package may be more complex and cost more, it may be the only way to achieve good enough results to interest farmers. A well designed package actually reduces farmer risk.

c. The practice(s) chosen should be afforable, adapted, and profitable for the majority of farmers. Some extension services may divide the region's target farmers into several recommendation domains, each with an adjusted package to reflect variations in soils, topography, economic circumstances, etc.

d. Gestation period: At least in areas where extension efforts are relatively new, practices that produce results in weeks or months are more likely to be readily accepted than those requiring longer periods.

II. Select a cooperating farmer (or organization)

a. Choose key farmers who are influential but not necessarily the best or most progressive, since they may be regarded as eccentric or as favored pets of the extension service.

b. Group demos on rented land are OK, but the group should be a pre-existing one (like a co-op) rather than one specially organized for the demo.

c. Since the farmer or farmer's group should do most of the actual work (this makes the demo credible), be sure that this is understood. This brings up the question of whether the inputs should be donated or charged for.

III. Choose a suitable location and


a. Site criteria: A conspicuous location with good exposure like near a road or trial. The land and soil shouldn't be atypically favorable or unfavorable but representative of target farmers' situations.

b. Size: Large enough to be realistic but not so big that it's difficult to visually compare the traditional and improved practices plot side by side. A couple hundred sq. meters per plot would be adequate for a maize demo, with less area sufficing for more intensive crops like veggies. It's also easier to find cooperative farmers if they know that the plots will be small, thus minimizing any perceived risk.

c. Plot layout: Decide the best way to show off the demo to an audience, plus consider audience size. A side by side layout is better than a front and back layout.

Side by side

It's usually best if the rows run front to back rather than parallel to the viewing area. This makes it easier for farmers to enter the plots.

IV. Provide adequate supervision of the demo

a. Both the extension worker and the farmer need to be thoroughly familiarized with the what, why, when, and how of the various operations involved.

b. Make sure the needed inputs are on hand.

c. Make sure that the inputs and other practices are correctly applied on schedule.

d. Avoid the tendency to favor the improved practice plot by giving it an unrealistic amount of care.

e. Keep accurate records, including rainfall, which will help analyze the success (or failure) of the demo.

V. Promotion and Followup after demo

a. At what stages will the demo produce visible results worthy of farmer attention- (i.e. only at harvest or by color and plant size difference at earlier stages).

b. Arranging farmer visits

• Supervised, scheduled visits conducted periodically at key stages are best. Since new practices usually require a combination of explanation and instruction, a combined result-method demo may be appropriate. However, do not count on farmers being convinced enough to try the new practices even by the time a successful demo reaches harvest stage. Also, farmers may not be able to apply the practices until next cropping season.

• Any demonstration should provide farmers with realistic cost-return data for the practices. Researchers and extension workers tend to exaggerate the claims and benefits of a new practice. Be conservative, yet realistic. A typical result demo is done on a small scale when farmers implement the methods on larger fields.

What About The "Spontaneous" Demo?

A Spontaneous Demo is a very effective type of demo using farmer's field that already demonstrates the benefits of what you're trying to promote. One advantage of the spontaneous demo is that it escapes the possible stigma of appearing contrived like a purposely organized result demo.


4. How to Make Conservative Recommendations

Researchers and extension workers tend to exaggerate the benefits of a new practice, while farmers usually have a more objective attitude. Here are some rule-of-thumb adjustments for arriving at realist_ claims.

I. Discount the amount of yield increase claimed for the new practice:

a. An experienced extensionist with lengthy local experience can discount claims pragramtically.

b. Result test data is representative: Discount expectations of increased yield by an amount equal to the standard deviation (a measure of statistical variance that you can calculate using the PC/ICE Traditional Field Crops manual.)

c. If the recommendation is based on outside data, discount the yield increase by at least 25%.

d. If test results occurred under a better than average crop year, discount increased yield claims by a least twice the standard deviation or by 35%.

II. Make an additional discount for less than optimum employment of the new practice by farmers.

For example, the effectiveness of recommended fertilizer rates depends greatly on proper timing and placement. There are no rules of thumb here. Usually, the more complex the practice and the more new skills involved, the greater the discount should be.

III. Make conservative economic estimates on returns

A new practice's potential economic returns make fertile ground for exaggeration:

a. Base all estimated costs and returns on "farm gate" costs and prices so that items like transport and commission are included.

b. Use harvest time prices when estimating returns.

c. After calculating the additional costs for the new practice(s), add on an extra 10% as a safety factor.

NOTE: Obviously an extensionist might end up discounting claims to a ridiculous extreme if he applied all the above guidelines. Remember that the purpose is to make claims realistically conservative, so it is wise to use judgement.


5. Ways of Reducing Risks Associated with New Practices

Aside from assuring that new practices are adequately tested locally before promoting them, there are several other ways of reducing farmer risk:

a. It is usually better to encourage farmers to try a new practice on only a of their land. This reduced both risk and hesitation and also enables the farmer to make comparisons.

b. In the case of purchased inputs, recommendations should be geared to providing the low budget small farmer with the maximum return per dollar spent rather than maximum profit per hectare. This is especially true for a high cost item like fertilizer; since yield responses begin to drop off as rates increase; low to moderate dosages will give the best return per dollar. (Bigger farmers can work on the principle of high volume, low return per dollar which give maximum profit per hectare).

c. Make sure that farmers thoroughly understand the how, what, when, and why of the new inputs or practice(s).

d. Small farmers in any area will vary in skills, capital, and management ability. The extension service should make sure that its recommendations are tailored to the needs of the majority but should also make provison for the special needs of more advanced farmers.

e. A "package" of practices that addresses the major yield-limiting factors simultaneously may also reduce risk under certain conditions.


6. Ideal Conditions for Promoting Improved Crop Production Practices to Small Farmers

• The new practice does not increase farmer risks.

• It does not depart radically from current practices or require considerable re-training.

• The potential gains exceed the added costs by at least two to one (This is the benefit/cost ratio.)

• The needed commercial inputs and associated services involved with the practice are readily obtainable at reasonable terms. For example, the spread of improved maize varieties in Kenya was greatly aided by the establishment of a network of 1000 small seed suppliers.

• The practice has been thoroughly tested in the area where it is to be introduced.

• The pay-off from the new practice occurs in the same crop cycle in which it is applied.

• The costs of the new practice are within the farmer's means. This usually implies access to reasonable credit.


7. Cost/Benefit and Net Return Analysis

This is a tool for measuring the economic benefit or loss attributable to a specific practice. It can be used to measure existing practices or new ones.


• List all inputs used in a practice.

• List cost of each input, in money terms. (Give effort to gather them a money value for locally-available inputs).

• Record and summarize costs of inputs.

• determine and record yield attributable to a practice.

• Calculate the money value of that yield.

• Set up a cost/benefit ratio.

• Determine the economic benefit of a practice.

Example :

Inputs and Cost





5 pounds at 15¢/lb.






100 pounds at 10¢/lb.








Estimated Yield


10 pounds onions/square meter

30 pounds onions


3 square meters


Yield Value


Market price $2.00/pound

$60.00 value


30 pounds








Economic Benefit

$49.75 total