Publications   KSA-Series   Paper5

KANSAS COUNTY EXTENSION AGENTS AND ATTITUDES
ABOUT ON-FARM RESEARCH (WORK):
RESULTS OF A SURVEY

David Norman, Stan Freyenberger, Bryan Schurle

KANSAS SUSTAINABLE AGRICULTURE SERIES

PAPER NUMBER 5

ABSTRACT

This paper reports the results of a survey designed to ascertain information about on farm research (i.e., on farm work (OFW)) implemented by county extension agents (CEAs) in Kansas. The survey involved collecting information on the types of OFW, who was responsible for it, what data were collected, what was done with the results, and attitudes of CEAs toward OFW. The survey results indicate that OFW associated with CEAs in Kansas is fairly common, that CEAs have positive attitudes for OFW and play important roles in initiating, implementing, analyzing, and distributing the results of such work, but major constraints exist to expanding such work.

(Contribution No. 97-195-D from the Kansas Agricultural Experiment Station. David Norman is a Professor, Stan Freyenberger is a Research Assistant, and Bryan Schurle is a Professor, all in the Department of Agricultural Economics, Kansas State University, Manhattan, Kansas, 66506.)

CONTENTS

Rationale

Approach

Personal Characteristics of the Extension Agents

Contact with Farmers

Trials and Demonstrations -- Frequency

Trials and Demonstrations -- Design and Implementation Issues

Trials and Demonstrations -- Evaluation and Result Dissemination Issues

Perceptions of, and Support System for, CEAs Concerning OFW

Concluding Remarks

References

.

TABLES

Table 1: CEAs, Farmers per County and Methods of Extension Contact by Region
Table 2: Numbers of Trials/Demonstrations Implemented
Table 3: Responsibilities for Design and Implementation of Trials/Demonstrations (Percent)
Table 4: CEAs Ranking of Factors Limiting OFW-Related Work
Table 5: CEAs Perceptions of Solutions to the Factors Limiting OFW
Table 6: CEA Desires for OFW if No Constraints Exist
Table 7: CEAs Perceptions of the Biggest Issues Facing Farmers

Table A1: Ages of CEAs in Relation to Experience in Extension (Percent)
Table A2: Relationship between Age and Previous Jobs of CEAs
Table A3: Relationship between Age and Education of CEAs
Table A4: Degrees and Majors of CEAs (Percent)
Table A5: Correlation Coefficients Between the Different Extension Contact Methods
Used by CEAs
Table A6: Responsibility for Suggesting Trial/Demonstration (Percent)
Table A7: Responsibility for Designing Trial/Demonstration (Percent)
Table A8: Responsibility for Setting Out Trial/Demonstration (Percent)
Table A9: Responsibility for Equipment Needed (Percent)
Table A10: Responsibility for Data Collection (Percent)
Table A11: Responsibility for Monitoring (Percent)
Table A12: Replication of Trials/Demonstrations Within and Between
Farms (Percent Breakdown)
Table A13: Intention of Trial/Demonstration (Percent Breakdown)
Table A14: Details about Field Days with Respect to Trials/Demonstrations
Table A15: Data Collected and Type (Percent)
Table A16: Responsibility for Data Analysis (Percent)
Table A17: Details about Reports Published
Table A18: Types of Reports Produced (Percent)
Table A19: Distribution of Reports (Percent)
Table A20: Composition of Program Development Committees
Table A21: Age Distribution of Program Development Committee Members

[home page] [top of article] [appendix tables]

RATIONALE

In recent years, interest in on-farm research (OFR) has increased, not only in low income countries where it has been associated closely with the farming systems research approach [Norman et al, 1995] but also in the USA where it has been linked with the sustainable agriculture movement. Arguments also have been made that considerable complementarity exists between station-based and on-farm research and that ideally these should be viewed as being on a continuum rather than as substitutes for each other [Norman, Frankerberger, and Hildebrand, 1994]. Certainly, there is considerable support and enthusiasm for OFR on the part of farmers. Indeed, a 1993 survey of farmers in Kansas indicated that both conventional farmers and those with an overt sustainable agriculture orientation were supportive of OFR, whether done by themselves or in cooperation with outsiders [Freyenberger et al., 1994]. However, because of the location specificity of many technologies relating to sustainable agriculture, the support for OFR by farmers with a sustainable agriculture orientation was relatively greater.

In spite of the above points, the support for OFR amongst station-based researchers generally can be termed as lukewarm at best. This lack of support stems from a number of factors, including a lack of appreciation concerning the value of the continuum between station-based and OFR, concerns about the loss of ceteris paribus conditions in OFR, the mistaken impression that OFR is relatively more expensive than station-based research, a lack of appreciation of the constructive role that farmers can play in the research process, and a reluctance to change the analytical approach to handle the special challenges of OFR.

As a result, although OFR (whether induced by outsiders or by farmers themselves) appears to be quite common, the payoff from efforts and resources allocated to it is potentially lower than it could be, in terms of sharing the results with interested outsiders (e.g., station-based researchers) and other farmers. Thus, we undertook a survey, to ascertain the degree and type of OFR-related work implemented by the front line extension workers (i.e., County Extension Agents (CEAs)) in Kansas; who was responsible for it; what information was collected and what was done with the results; and what were the attitudes of the CEAs themselves about it. In this paper the term on-farm work (OFW) is used to denote both trials (i.e., which is really OFR) and demonstrations with which CEAs are also involved.

APPROACH

In 1993, a questionnaire was mailed, through the Associate Director of Cooperative Extension, to all 105 CEAs in Kansas. Eighty-two completed questionnaires were returned (i.e., a 72% response rate). Following some personal details (Part A), other parts of the questionnaire involved collecting information about the trials/demonstrations they had coordinated on farms (Part B), their perceptions of the constraints to on-farm work (OFW) (Part C), and some attitudinal information (Part D). In this report, presentation of the results approximates the format of the survey form. Some of the more detailed tables are placed in the appendix.

PERSONAL CHARACTERISTICS OF THE EXTENSION AGENTS

The average age of the CEAs was about 35, and the average time they had spent in their current post was about 3.5 years (Table A1). However, 55% had held other extension jobs, and 75% had held jobs outside extension. Of these other jobs, agribusiness (28.3%) and teaching (20.7%) were most common (Table A2). This implies a fair degree of job mobility with respect to CEAs in Kansas. Not surprisingly, however, a significant (i.e., at 5% level) positive correlation (i.e., r = 0.4670) occurred between age and years in their current post, indicating that job mobility decreases as CEAs become older.

In terms of qualifications, the average level of formal education included a BS degree plus some extra courses (Table A3). Consequently, 65% had BS degrees, and 35% had MS degrees. For those with an MS degree, the average age tended to be in the late 30s. In terms of majors the most dominant one at the BS level was animal science (41.5%), with agricultural economics and education being distant second and third (i.e., 24.5 and 20.8%, respectively) (Table A4). Animal science was also most popular at the MS level, which is not altogether surprising, because there was little evidence of changing majors between the BS and MS levels.

CONTACT WITH FARMERS

Table 1 indicates the average number of farmers per CEA for the different parts of the state. As would be expected, the average is lower for the drier western part of the state where farms are larger. As Table 1 indicates, various methods are used to contact farmers, but, apart from one, the percentage of farmers contacted by each method decreases as the number of farmers in the county increases. Superficially, the one exception appears to the percentage of farmers visited. However, a more rigorous examination with the help of correlation coefficients (Table A5) indicated a significant negative correlation between total farmers in the county and the percent of farmers visited. Negative relationships also existed with respect to the other variables, but only one was statistically significant (i.e., percent of farmers who attend tours/field days). However, a number of statistically positive correlations occurred between the percent of farmers contacted and among the methods of contact themselves (Table A5). Direct contact with farmers through visits by CEAs appears to have a positive relationship with the percent of farmers who attend field days/tours and seminars/training schools. The significance of field tours/days as a method of contacting farmers is obvious from the data in Tables 1 and A5. This is important if trials/demonstrations are to be used effectively in collecting information from, and disseminating information to, farmers. No statistically significant relationships were detected between the age of the CEA and percent of farmers contacted or the methods of contact used.

TRIALS AND DEMONSTRATIONS -- FREQUENCY

The CEAs are involved with both trials and demonstrations, which constitute OFW. In the survey, we tried to distinguish between the two, considering trials as experimental in nature and demonstrations as OFW that involves recommended technologies or final stage testing of a technology. Thus, the dividing line between the two becomes somewhat blurred at times. On average, the number of trials/demonstrations implemented per CEA in 1993 was almost 6 (median 4) with a range between 0 (3.7% of the CEAs) and more than 15 (2.4% of the CEAs). More than 65% of the CEAs implemented between 2 and 6 trials/demonstrations in 1993.

Table 1: CEAs, Farmers per County and Methods of Extension Contact by Region

a. Some double counting could have occurred (e.g., farmers be visited and attending tours, attending more than one field day, etc.).
b. Includes distributing publications, telephone calls, etc.

Information on some of the types of trials/demonstrations implemented by CEAs is given in Table 2. As expected, by far the majority of the trials/demonstrations could be considered as demonstration (about 72%). As anticipated, the most common trial/demonstration, for which details were available, was crop related (66%), with those relating to livestock being relatively unpopular (10%). This finding is consistent with results obtained elsewhere. For example, farmers in Kansas, in their own initiated research, tend to concentrate much more on crop- related issues [Freyenberger et al., 1994). One major reason is that OFW with respect to livestock is methodologically more complex than that with crops [Norman et al., 1995). This is particularly unfortunate given the fact that the most common formal training of CEAs in Kansas is in animal science.

In terms of crop-related OFW, about 54% of the trials/demonstrations on which detailed information was collected involved evaluation of crop varieties and 88% of these were implemented in a demonstration format. In contrast, those that required more of a systems perspective or were livestock related tended to be implemented in the form of trials. The reason for this relatively high emphasis on the trial format presumably partly relates to the difficulty of implementing OFW with respect to livestock.

Table 2: Numbers of Trials/Demonstrations Implemented

a. The CEAs were asked to give details on up to five trials/demonstrations with which they were associated.

TRIALS AND DEMONSTRATIONS -- DESIGN AND IMPLEMENTATION ISSUES

A number of questions were asked concerning the dynamics and responsibilities in suggesting, designing, and implementing the trials/demonstrations. A consolidated summary of some of the major issues is given in Table 3, and breakdowns by trial/demonstration and by subject area are given in Tables A6 to A11.

The results clearly show that much of the initiative for suggesting and designing trials/demonstrations with which the CEAs are associated comes from the agents themselves. However, implementation in terms of not only laying out of the trial requires the cooperation of the farmers, who also play a major role in providing the necessary equipment. Farmers also play important roles in collaborating with CEAs in monitoring the trials/demonstrations and in collecting data. The results in Tables A6 to A11 indicate that these general observations apply to both trials and demonstrations and are also independent of whether the focus is on crops, livestock, or some other subject area.

Although extension staff play such an important role in suggesting and initiating trials/demonstrations, at least a third of them are not replicated either within or across farms (Table A12), thus severely limiting the submission of the results to any kind of formal statistical analysis. This is likely to curtail the interest of researchers, who obviously are interested in making generalizations and thereby extrapolating the results from OFW. However, this may not be important for farmers, because they are interested in evaluating possible changes in terms of their own specific situation (i.e., moving from the general to the particular) [Anderson and Lockeretz, 1991]. When replication did occur, it was within or between farms or included some combination of both. The average number of replications ranged from 1.4 to 2.7. Not surprisingly, replication was a little more frequent in trials than in demonstrations and replication within farms tended to receive greater emphasis. This is not surprising given that experiment station-based researchers tend to emphasize replication within site-specific trials. Therefore, they are likely to feel most comfortable extrapolating their methodology to OFW. However, for those more intimately involved in OFW, one of the compelling arguments of such trials/demonstrations is to evaluate the robustness of potential technologies across a range of production environments that include not only variations in the technical (e.g., soil) but also the socioeconomic (e.g., management) environments. Although within-farm replication can help farmers on whose farms such trials/demonstrations are located evaluate the technology specific to their own situation, resources available for such OFW are likely to preclude repeating it on many farms. In any case, results of the survey of farmers in Kansas indicated that when farmers initiate their own "experimental trials," they rarely replicate their work, so their evaluation is based on criteria not related to formal statistical verification [Freyenberger et al., 1994]. Thus, replication is not an issue that influences farmers interest in using the results of trials/demonstrations. However, replicating trials/demonstrations across farms makes it easier to target recommendations, and also elicits the opinions of more farmers. Fortunately, techniques such as adaptability (i.e., modified stability) analysis have become available in recent years [Hildebrand and Russell, 1996] to handle these types of replications.

Table 3: Responsibilities for Design and Implementation of Trials/Demonstrations (Percent)

a. Includes those coming from researchers.

Table A13 indicates the CEAs’ perceptions of the trials/demonstrations with which they were associated. Illustrating higher yields and to provide focus for field days were the most important stated intentions of such OFW, although other criteria were more important for trials or experiments. The fact that other intentions also commonly underlay livestock-associated OFW was associated with the greater dominance of trials in such types of OFW. The intention of OFW relating to providing a focus for field days has merit, given the finding of the farmer survey that one of the most important sources of information for farmers was other farmers [Freyenberger et al., 1994]. Farm tours and field days provide opportunities not only for contact with CEAs but also for contact with other farmers.

Table A14 indicates that field days were held for nearly 70% of the trials/demonstrations but, as expected, were more common for demonstrations (77%) than for trials (49%). Given the predominance of demonstrations with respect to crops, field days were held for 80% of the crop-related OFW compared with only 32% for livestock-related work. Average attendance at field days was about 55 persons, and the CEA-expressed purpose of such field days was to provide farmers with a visual perception of the technologies being evaluated. This is consistent with the intention underlying OFW discussed earlier (Table A13). However, the results in Table A14 indicate that another important reason for such field days is to provide an opportunity for meetings, something that seems to be important particularly with respect to livestock-related work. Meetings can provide a very important forum for obtaining farmers’ opinions not only on technologies that are shortly to be, or are being, recommended (i.e., demonstrations) but also for those that are still in the developmental stage (i.e., trials). OFW researchers see the latter as an increasingly important purpose of OFW, particularly with respect to incorporating a systems perspective and ensuring consideration of the multiple evaluation criteria underlying farmers’ decisions whether to adopt changes in their farming system [Norman et al., 1994, 1995].

TRIALS AND DEMONSTRATIONS -- EVALUATION AND RESULT DISSEMINATION ISSUES

Tables A15 and A16 indicate the types of data collected from trials/demonstrations and responsibility for its analysis. For 21% of the trials/demonstrations, no data were collected, which is consistent with the notion that the basic intention of many of the trials/demonstrations is to provide a focus for field days (Table A13). However, such trials/demonstrations represent a lost opportunity in terms of transmitting information on the results to others who were physically present at the field days and also those who had not had the opportunity to personally see them. Also much of the data collection emphasized only technical data (i.e., 41% of the trials/demonstrations) and only 14% involved the collection of both technical and economic information. Once again, this represents a missed opportunity, because OFW, in contrast to the more artificial environment on experiment stations, provides an operational milieu in which results can be interpreted from both the technical and socio-economic viewpoints, which are both critically important in the practical farm production environment. The reason that such combined analysis is not undertaken more commonly possibly relates to the fact that most CEAs are formally qualified in the technical disciplines, and that not much opportunity seems to exist for engaging in interdisciplinary team efforts with reference to OFW. Table A16 indicates that the major responsibility for analyzing data from trials/demonstrations falls on the CEAs themselves. Farmers appear to have little involvement in this exercise. This is perhaps inevitable but nevertheless is unfortunate, given the importance mentioned earlier of farmer-farmer interaction and the recognition on the part of OFW practitioners that true collegiate interaction between farmers and outsiders or "researchers" (in this case, CEAs) is needed to maximize the payoff from OFW [Norman et al., 1995].

Tables A17 to A19 provide details about dissemination of the results of the trials/demonstrations with which CEAs are associated. Tables A17 and A18 indicate that reports were published/produced on about 23% of the trials/demonstrations, with the county and KSU being the most popular media. Not surprisingly, KSU appeared to be relatively more popular with reference to trials (50%), whereas the county outlet (54%) was relatively more popular with respect to demonstrations. Newspapers also were used as a way of disseminating findings. The average numbers of copies of the reports produced amounted to about 730. Table A18 indicates that by far the most important type of report produced consisted of informational fliers or summary reports (50% of the total reports produced). More than one type of report was produced in about 20% of the cases published, whereas would be expected, statistical reports and detailed descriptive reports were relatively more common with respect to trials than demonstrations (i.e., about 19% compared with 5%). Table A19 indicates that the major recipient of the reports were farmers (58%) with little difference between trials (61%) and demonstrations (58%). As expected, the KSU experiment station was a more significant recipient of reports based on trials (i.e., 16% compared with 9%), whereas agribusiness was a relatively more significant recipient of results with respect to demonstrations (i.e., 23% and 15%). This is not surprising, because crop varieties were major components of demonstrations.

PERCEPTIONS OF, AND SUPPORT SYSTEM FOR CEAs CONCERNING OFW

Because the Program Development Committees (PDCs) have a major influence on the priorities of the different activities of CEAs, information was obtained about them (Tables A20 and A21). The average size at the time of the survey was about six persons, with farmers plus spouses constituting about 84% of the membership. The average age of the committee members was the early 40's, with very little representation from those less than 30 or more than 60 years old. Not surprisingly, given results from the farmer survey noted earlier [Freyenberger et al., 1994], the PDCs enthusiastically supported OFW (i.e., 90% compared with only 3% who were not supportive).

These results indicate that the vast majority of CEAs felt they had a mandate to be supportive of OFW. However, the perception among CEAs was that the situation with respect to OFW at the time of the survey was not optimal. They recognize a number of limitations with respect to implementing OFW. A list of the major constraints that were perceived by CEAs to be limiting their participation in OFW is given in Table 4. Too many other duties was perceived to be the most important constraint, and together with the lack of adequate equipment, was statistically more important than other constraints. According to Table 4, CEAs had some insecurity about their expertise in designing and analyzing the results of OFW, but in aggregate, this was statistically no more limiting than the lack of a strategic plan for OFW or farmers not being interested in OFW.

Table 4: CEAs Ranking of Factors Limiting OFW-Related Work^a

a. CEAs ranked the limitations from 1 to 6 with 1 being the most important.
b. Different letters indicate significant difference at the 5% level, using the Tukey test.

CEAs also were asked to indicate potential solutions to the factors they had ranked as being constraints relating to OFR. A percentage breakdown of the suggested solutions to each of the perceived constraints is given in Table 5. Although, the importance of the specific solutions for each of the perceived constraints does differ, but they can be combined into three major groups of perceived solutions to encouraging greater OFW:

-The need for KSU assistance and the related need for training (35%).
-The need for budgetary increases and the related need for hiring extra staff and purchasing necessary equipment (34%).
-The need for better planning of OFW activities (16%).

Table 5: CEAs Perceptions of Solutions to the Factors Limiting OFW^a

a. Apart from the last column and last line, the figures in the cells represent the percentages for the totals in each column.

b. The number of responses in each column is weighted by 1 minus the average figure given in Table 4 and then expressed as a percentage of the total. This weights the specific solution according to the severity of the problem.

The obvious implications of these findings are that the CEAs support greater OFW, but that its increased implementation will be very dependent on factors that are outside the control of the CEAs themselves. These relate to the need for it to be of greater priority in the activities of CEAs and the need for this to be complemented by appropriate support systems in the form of finance and training initiatives. Unless these are addressed, the present system relating to planning, implementing, and dissemination of results of OFW probably will not change.

An open-ended question to CEAs on what they would like to see done in OFW yielded the results given in Table 6. The relative greater emphasis on crop in relation to livestock work would be analogous to what is currently occurring (compare with Table 2), but relatively greater emphasis within the crops-related area would be placed on topics related to agronomy and crop protection (e.g., soil quality, weeds, and insects). Such topics often require more of systems perspective in designing appropriate solutions and often manifest themselves in a number of different ways under practical farming conditions. Thus, OFW often can provide a more realistic environment than would exist under experiment station conditions in which to examine such issues and evaluate appropriate solutions.

Table 6: CEA Desires for OFW if No Constraints Exist

a. Percent of total responses (n=106).
b. It is likely most of this relates to crop variety testing/evaluation.

A final question in the survey related to CEAs perceptions of the biggest issues facing farmers. Their responses are summarized in Table 7. Note that at least 40% of them require approaches other than OFW for their resolution (i.e., are not technical in nature). However, many of the issues (i.e., about 28%) also are not related directly to specific commodities but are problems whose resolution would likely best involve a systems perspective. OFW is better able to incorporate a systems perspective than experiment station- based research. Thus, OFW potentially has some sort of role in addressing at least 68% of the issues or problems facing farmers.

Table 7: CEAs Perceptions of the Biggest Issues Facing Farmers

a. Percent of total responses.

CONCLUDING REMARKS

The survey results indicate that, in Kansas, OFW activities associated with CEAs are fairly common, and that CEAs play important roles in initiating, implementing, analyzing, and distributing the results of such work. Demonstrations tend to be more common than trials, and more emphasis is placed on crop-rather than livestock-related work. CEAs appear to have positive attitudes with respect to OFW, but major constraints exist in terms of expanding such work. This is unlikely to occur unless OFW receives greater priority and supportive systems are put in place in terms of the necessary budgets and training in design and analysis of such work. Justification for such efforts would involve ensuring that the payoff from OFW is maximized. This does not appear to be the case at the present time.

REFERENCES

Anderson, M.D., and W. Lockeretz, 1991. "On farm research techniques". Report on a Workshop, St. Paul, Minnesota, 15-16th Nov, 1990. Greevelt: Institute of Alternative Agriculture.

Freyenberger, S., L. Bloomquist, D. Norman, D. Regehr and B. Schurle, 1994. "On-farm research in Kansas, 1993: A survey of farmers' opinions". Report of Progress 720. Manhattan: KAES, Kansas State University.

Hildebrand, P. and J. Russell, 1996. Adaptability Analysis: A Method for the Design, Analysis and Interpretation of On-Farm Research-Extension. Ames: Iowa State University Press.

Norman, D., T. Frankenberger and P. Hildebrand, 1994. "Agricultural research in developed countries: past, present and future of farming systems research and extension". Journal of Production Agriculture 7(1): 124-131

Norman, D., F. Worman, J. Siebert, and E. Modiakgotla, 1995. "The farming systems approach to development and appropriate technology generation". Farming Systems Management Series Number 10. Rome: AGSP, Food and Agricultural Organization of the United Nations.

APPENDIX TABLES

Table A1: Ages of CEAs in Relation to Experience in Extension (Percent)

Table A2: Relationship between Age and Previous Jobs of CEAs

a. The breakdown of other jobs reported was as follows:

Agribusiness = 28.3%
Teacher = 20.7%
Farming/Farm Management = 17.0%
Research Assistant/Trainee = 13.2%
Other = 20.8%

 

Table A3: Relationship between Age and Education of CEAs

Table A4: Degrees and Majors of CEAs (Percent)^a

a. Figure in brackets represents the number of CEAs that changed majors from the BS to MS level

Table A5: Correlation Coefficients Between the Different Extension Contact Methods Used by CEAs

a. Apart from the total number of farmers, the rest of the variables were expressed in percent terms before the correlation coefficients were estimated.
b. * is significantly different from 0 at the 5% level and ** is significantly different from 0 at the 1% level.

Table A6: Responsibility for Suggesting Trial/Demonstration (Percent)

Table A7: Responsibility for Designing Trial/Demonstration (Percent)

Table A8: Responsibility for Setting Out Trial/Demonstration (Percent)

Table A9: Responsibility for Equipment Needed (Percent)

Table A10: Responsibility for Data Collection (Percent)

Table A11: Responsibility for Monitoring (Percent)

Table A12: Replication of Trials/Demonstrations Within and Between Farms (Percent Breakdown)^a

a. Figures in brackets represent average number of replications. Under "both," the first figure represents within farms and the second across farms. "na" means not available.

Table A13: Intention of Trial/Demonstration (Percent Breakdown)^a

a. Because CEAs could have more than one intention, the most important was ranked weighted 3, the second 2, and the third 1.

Table A14: Details about Field Days with Respect to Trials/Demonstrations

Table A15: Data Collected and Type (Percent)

Table A16: Responsibility for Data Analysis (Percent)

Table A17: Details about Reports Published

Table A18: Types of Reports Produced (Percent)

Table A19: Distribution of Reports (Percent)^a

a. Figures in brackets represent average number distributed.