Computergesteunde voorlichting : een decisiegericht voorlichtingskundig onderzoek naar Epipre en andere geautomatiseerde informatiesystemen in de landbouwvoorlichting

K.J. Blokker

    Research output: Thesisinternal PhD, WU

    Abstract

    1. Decision-oriented extension research (see chapter 2)

    1.1 Applied and decision-oriented extension research

    Through extension research one tries to contribute to the improvement of extension. Extension researchers try, in other words, to generate results that are utilisable for extension practice. Two types of extension research can be distinguished: applied and decision-oriented extension research. Although through applied extension research researchers try to contribute to the improvement of extension practice, they seldom actively try to stimulate the utilisation of the research results. Most of the time the researcher tries to find answers to problems he thinks are of relevance to extension practice. When the research is finished, he will produce a report, write one or more articles and, maybe, he will give lectures on the topic. The results usually are intented to be adaptable to many situations, in other words he is trying to generate results which are generalisable. Decision-oriented extension research is more closely connected with practice. The main focus is not the production of generalisable results but to contribute to the decisions in a specific situation concerning extension policy, management or the way extension should be put into effect. The request for research is made by extension officers, managers of extension, policy makers or those who receive extension services. We shall restrict ourselves to decision-oriented extension research. A practical reason for this choice is the fact that the research we will report, is a decision-oriented extension research. But attention to this kind of research is justified also because most extension research is in fact decision-oriented. In decision-oriented extension research it is important that the researcher actively tries to further the utilisation of the reults. Our aim is now to formulate recommendations for the researcher that will stimulate the utilisation.

    1.2 Recommendations in furtherance of the utilisation of decision-oriented extension research

    Most of the recommendations are directed toward achieving good communication between the researcher and (representatives of) the utilisers. Since it has been found that good communication between the researcher and (representatives of) the utilisers promotes utilisation, collaboration is recommended throughout the whole research process; from the first stage of formulating the objectives of the research, up to and including the implementation of the recommendations. We recommend that the researcher and (representatives of) the utilisers together take the main decisions concerning what should be investigated and how, the interpretation of the research results, the recommendations that are formulated woth reference to the research and the transformation of these recommendations (the process of adapting the recommendations for use in practice). Because researcher and utilisers have different expertise (the researcher being specialised in extension education and research methods, utilisers having more knowledge about extension practice) and because the participants in the decisionmaking process often will differ in regard to the priorities they set (because these are influenced by different values and positions they occupy within the organisation), the decisionmaking process will sometimes be very much like the process of negotiation. Negotiation is especially evident in the first stage of research when the parties are seeking agreement over what is to be investigated. It is also evident during discussions of the implications of the research results. So far, we have only paid attention to the communcation between researcher and those utilisers who are strongly involved with the research process. Usually it is not possible for all (potential) utilisers to participate to that extent. A good communication between the researcher and this broader circle of utilisers is nevertheless also important, for the furtherance of the utilisation of extension results. Therefore we have also dealt with the communication process between researcher and other utilisers. This communication is intended to inform those utilisers about
    the progress that has been made and to elicit suggestions concerning the continuation of the research. It also implies the process of dissemination of research results and recommendations to this broader circle of utilisers. We not only have formulated. recommendations concerning the communcation process between researcher and utilisers. We have also stressed that the problems we are confronted with in decisionoriented extension research are of a multidisciplinary character. The researcher needs to be aware of this fact and, in collaboration with (the representatives of) the utilisers, carefully to decide which variables are to be included in the research and which are to be excluded. It is also recommended that, when deciding which variables are to be investigated, we try to ensure that we will obtain sufficient insight into where and how interventions might be made, in order to be able to influence the situation under study. The last recommendation is that the researcher be flexible in choosing the most approriate research methods. In decision-oriented extension research the methods used in information-gathering should meet the requirements demanded by practice (i.e. the information that is produced be of relevance and useful) and by science (i.e. the information be reliable and valid). This calls for a careful decision between the use of current and new (adapted) methods of information-gathering.

    It is suggested that the recommendations made here be placed within the broader context of the influence that information produced by (social) research has on decisionmaking processes. Our conclusions are that such information is only one component of the considerations which influence decisionmaking and therefore researcher and utilisers must both recognise the limited nature of the assistance they would gain from decision-oriented research. The recommendations advocated here have been put into practice in the research reported in the following pages.

    2. The objectives of the research (see chapter 3)

    In 1981 the Ministry of Agriculture and Fisheries asked the department of Extension Education of the Agricultural University
    to make a study of Epipre (see 3.1) and other automated information systems. Our study has focussed on the following questions:
    - should Epipre be continued? If so, which improvements are desirable and how can they be implemented?
    - under which conditions can automated information systems be of help to farmers in their decisionmaking?
    First we shall give an account of the research on Epipre (3.1 up to and including 3.4.2) and then of the research on other automated information systems (4.1 up to and including 4.4.4).

    3. The research on Epipre

    3.1 Epipre (see chapter 3)

    Epipre is a computer-based pest and disease management system for wheat. Wheat is a crop commonly grown in the Netherlands. In order to protect it, farmers spray the crop from one to possibly four times during the growing season. Normally, they do so on the basis of their own observations and their own judgement or for instance on those of the local extension officers service. In the Epipre-project, however, farmers are expected to make a number of standardised observations and to post these to a computer centre. There the data are processed field by field and a spraying advice is obtained for each field for which the farmer has sent an observation. The advice is then posted to the farmer. Epipre and its underlying computer programme were developed by the Department of Phytopathology of the Agricultural University in collaboration with the Department of Theoretical Production Ecology. It has been implemented on an experimental basis by the state extension service since 1982. Some 500-600 farmers are presently involved in the project. The research reported here was carried out in 1981, 1982.

    3.2 The interviews (see chapter 5)

    In the first months of 1982 a total of 357 farmers were interviewed. They belonged to one of the following categories:
    - users of Epipre (n = 156)
    - ex-users of Epipre (n = 76)
    - non-users (n = 125).
    Those users who did not plan to use Epipre in 1982 again, we have
    for some of the analyses combined with the ex-users.

    3.3 The main results (see chapter 6)

    3.3.1 Users often deviated from the Epipre advice

    Only 17.5% of the users report that they have in all cases strictly followed the recommendations that Epipre gave to them. The great majority therefore has at least once not followed the Epipre recommendation. Most farmers at least once:
    - spray more often than Epipre advises them to do (46.5%);
    - spray before they have received the recommendation through the mail (41.3%);
    - use more pesticides per treatment than stipulated in the Epipre recommendation (32.3%).

    3.3.2 Farmers' critique of Epipre

    Their critique comprises the following points:
    - some 50% of the users mention that they have to wait too long (2 to 3 days) before the Epipre advice reaches them;
    - some 50% of the users also mention that Epipre is too modest in its recommendations: the number of times that it recommends spraying is lower than they would have liked;
    - some 25% of the ex-users of Epipre report that they have with drawn from the project because it took them too much time to do their field observations properly. It should be mentioned, also, that some 50% of the ex-users report that they withdrew because they felt that they had learned enough.

    3.3.3 Did Epipre affect the net yield?

    We did not find that Epipre had any statistically significant influence on the net yield. It appeared that when farmers strictly follow the Epipre advice, they are likely to spray less often and to use less pesticides per treatment than other farmers.
    In so doing this will mostly lower their yields but it will also reduce their costs, so at the end they do not, at least, lose any money.

    3.3.4 What did farmers learn?

    The great majority (75 to 90%) of both the users and the ex-users of Epipre report that, thanks to the project:
    - they are now making their observations closer to the right time;
    - they have increased their ability to diagnose pests and diseases in their wheat;
    - they now know better when to spray.

    3.4 The decisions (see chapter 7)

    3.4.1 Epipre is to be continued

    At the end of 1982 the Ministry decided to continue Epipre. Two factors were of importance in the decision. The first is related to extension policy. The Ministry holds the opinion that computerised information systems will come to play an important role in agricultural extension. Secondly, the Ministry is of the opinion that the project had had interesting teaching-and-learning effects on participating farmers and therefore merited continuation, even though Epipre did not necessarily increase farmers' net income from wheat (3.3.3).

    3.4.2 Epipre is to be changed

    It was, however, also proposed that Epipre be modified in a number of ways:
    - Epipre itself be expanded into a more comprehensive crop management system for wheat;
    - the time that Epipre required of farmers in making observations be reduced and a number of procedures therefore simplified;
    - various measures be taken to improve Epipre's ability to function as a helpful device that can assist farmers in their decision making. For instance, the spraying recommendation be presented more convincingly by giving the farmer the reason for a particular recommendation rather than sending him only the recommendation itself.

    4. Other automated information systems (see chapter 8)

    4.1. What did we study?

    As was pointed out earlier (see 2), one of the research questions concerned the conditions under which automated information systems can be of use in agricultural extension. Our study of the literature has focussed on systems like videotex (e.g. Prestel, Green Thumb Box, Viditel) and computer network systems that are mainly in use within the extension service itself (e.g. Facts, Agnet, CMIN, Telplan. AIS). We also paid attention to farmers ' use of microcomputers and programmable calculators. The various systems differ from each other in the information that they provide to the users. They also differ with respect to the amount of direct communication between the farmer and the system that they require. For instance, videotex systems usually require direct communication with the farmers whereas in computer network systems the communication lines are extended and run through the extension officer.

    4.2 Some evaluation results

    Restricting ourselves to those systems that require direct contacts with the farmers, we must conclude that at the average those
    systems are only moderately used. Farmers do not appreciate most systems very much and they generally report they had only marginal influence on their decisionmaking It may well be that the designers of these systems have tried to do too many things too fast and have produced systems that do not to a satisfactory level fulfil farmers' requirements. We consider that a number of
    factors have contributed to the lack of satisfaction and shall discuss further below the one factor which to our mind has been
    of greatest importance.

    4.3 Many systems are insufficiently adapted to the intended users

    Various authors have emphasized that the two main advantages of automated information system are the following:
    - they make available a lot of information;
    - they do so very fast.
    Apparently the designers of these systems view farmers as managers who use a great quantity of external information that, moreover. needs to be up to date. Several investigations, however, have demonstrated that this image is incorrect: most farmers use relatively little (external) information and in only a few cases do they want it quickly (e.g. information on the weather or on market prices).

    4.4 Some recommendations for the development of automated information systems

    Obviously, its initial design is of crucial importance for the future functioning of an automated information system. In what follows we will discuss some decisions that have to be taken when designing such systems.

    4.4.1 Defining the exact nature of the information that the system is to provide

    Before the information to be provided by the system can be determined, it is necessary to define the issues raised by the follo-
    wing questions:
    - what are the decisions that farmers actually make?
    - what problems do they face in making these decisions?
    - which of these problems are actually caused by shortcomings in the present information system?
    - can an automated information system help to overcome these shortcomings?
    An answer to questions such as the above is of course only the beginning, there is for instance also the matter of whether or
    not the new system can be expected to increase the efficiency of the extension service involved.

    4.4.2 The quality of the information

    The information should be scientifically sound and in some cases - e.g. weather information, market prices - be highly up-to-date.

    4.4.3 The decisionmaking process of farmers

    It has been mentioned already that farmers often need and use information in ways that differ from what designers of information systems assume (see 4.3). It would appear that some systems even require farmers to change the content and nature of their decisionmaking. The Epipre spraying advice, for example, was based on a dynamic simulation model that included a fairly large number of parameters. The extension service, however, has demonstrated to farmers how a decision could be reached based on a much smaller number of variables. Using simple criteria farmers can, therefore, reach their own conclusions. Epipre uses a highly sophisticated model and farmers have difficulties in tracing the origins of an advice Epipre gives them. At the same time however often the advices of Epipre and those of the extension service are different. In much cases Epipre advises the farmer to spray less often and to use less pesticides per treatment than the extension service would have recommended. Both factors (i.e. the Epipre model is not fully understood by farmers and the Epipre advice often deviates of what is usual) we assume at times had led to a disagreement between users and Epipre, as a result of which users often deviated from the Epipre advice (see 3.3.1). This is not to suggest that the criteria that farmers use should provide the only basis for the computer model that constitutes the backbone of an information system like Epipre. What we want to suggest, however, is that it would be fruitful to look for an optimum that includes both the scientific and economic considerations of the specialists and the practical considerations of the farmers, an obvious rule-of-thumb for most extension work, yet one that in many cases appears not to have been followed. Too often, it seems, those who design such information systems allow themselves to be guided by the level of sophistication and the potential of their equipment, especially their computer, which leaves the farmers with the unnecessarily cumbersome and timeconsuming task of providing a considerable amount of data that is both reliable and sufficiently precise. The lessons of experience strongly suggest that automated systems should not demand from the user more time and effort than is most strictly necessary.
    Furthermore, the system should be kept as simple as possible, not only with regard to the data collection aspect of it, but also as far as its output in terms of information or advice to farmers is concerned. Further, one could add that it would be desirable if the systems' users had a basic understanding of the way in which the programme operates in reaching a decision. In the absence of this understanding it would be difficult for users to decide whether the recommendation that results from the system is optimal, considering their circumstances and their goals (see also 3.4.2 on the third suggested improvement for Epipre).

    4.4.4 Collaboration between designers, extension staff and farmers

    In the foregoing we have stressed the importance of adapting an information system to the needs and circumstances of its target group. i.e. its potential users. Several suggestions have been made as to how this could be done even at the design stage of an information system. We conclude with the suggestion that scientists alone should not be involved in designing the system. Farmers and extension workers should participate in the design as well. Farmers because they have a lot of practical information to offer, extension staff because they are by profession familiar with the intricacies of bridging the gap between science and practice through the transformation and dissemination of information.

    5. Epilogue (see chapter 9)

    As we have already mentioned, we have tried to further, by several measures, the utilisation of the research on Epipre and other automated information systems. We might conclude that there are several indications that these measures have been rather effective. We must add to that, that a more substantial participation during the research on Epipre by the regional extension service, would probably have enhanced the utilisation. This seems especially true for the dissemination of Epipre by regional extension service personnel.

    Original languageDutch
    QualificationDoctor of Philosophy
    Awarding Institution
    Supervisors/Advisors
    • van den Ban, A.W., Promotor
    Award date20 Jun 1984
    Place of PublicationWageningen
    Publisher
    Publication statusPublished - 1984

    Keywords

    • information services
    • extension
    • computers
    • minicomputers
    • microcomputers
    • data processing
    • mechanization
    • automation
    • agricultural extension
    • machines

    Cite this