Description of action research by the interviewees
As action research tends to be defined differently based on the discipline of the individual (Reason and Bradbury 2008), the individual interview started by asking the respondent’s understanding of action research. From the discussions with BRIDGE project staff and partners, action research was described as follows:
i. A research conducted involving the end-users with purpose of providing practical information for implementation of project activities.
The emphasis from this description is on participation of the end-users in implementation of research activities. As participation of different stakeholders is a critical element of action research, this description is consistent with definitions in the literature by McNiff (2013), and Lebesky and Benders (2020). Participation of different stakeholders is necessary in giving voice to the end-users and to enhance ownership of both the process and the outcomes (Loo 2014; Schulz et al. 2021). In stressing the importance of participation in action research, Brydon-Miller et al. (2003) observed that research conducted without a collaborative relationship with the relevant stakeholders is likely to be incompetent. Another key point from this definition by the project participants is action, that is the research is action-oriented to address practical issues of importance to the stakeholders. This is consistent with one of the tenets of action research, that is, it is research that leads to action (practice) and effects change or leads to innovation (Brydon-Miller et al. 2003). The “action” in “action research” is critical to testing knowledge in action or putting theory into practice (Brydon-Miller et al. 2003). The action to be undertaken by the relevant stakeholders may be both remedial and developmental (Kenefick and Kirrane 2022).
ii. Action research is a learning platform to inform implementation or specific action by the stakeholders.
The emphasis of this description of action research is on learning that takes place along the action research cycle. This learning is by all stakeholders including researchers. The emphasis on learning in this description by the BRIDGE participants is consistent with observations in the literature on action research that learning is a fundamental element (McNiff 2013; Kindon et al. 2007). With involvement of different participants in a project, diverse opportunities for learning exist through interactions among the stakeholders, based on the premise that multiple knowledges are available. This implies that researchers are not the only source of knowledge in action research. As there are different kinds of knowledge, the emphasis of action research is on transformation knowledge which produces practical solutions to the pressing concern of the people (Brydon-Miller et al. 2003). For co-learning to take place, it is therefore necessary that issues to address are jointly identified and co-implemented by all stakeholders, rather than being imposed by the researchers as in conventional research. In support of the “action” in action research, one of the respondents observed that: “There is a research fatigue among farmers, where researchers just come to collect data and disappear. Hence, there is need for research that translates into practical action, thereby contributing to improvement of the livelihood of farmers.” (Respondent PS 5).
From the above descriptions, it is obvious that many of the BRIDGE staff and partners have a good understanding of the action research approach, although the emphasis may differ from joint planning to participatory learning. This is not surprising as development assistance agencies and non-governmental organizations have been good at applying the values and practices of action research to development projects, as observed by Mead (2002). One important element of action research that was missing from above descriptions is that of reflection, which may suggest that the reflection part of action research was weak in the action research activities of the project. From the discussions on the general understanding of the action research approach by the respondents, the most frequent keywords or phrases used were “learning platform”, “co-implementation”, and “technology demonstration” (Fig. 2). Action research was also described as a linkage between research and extension. Again in the keywords/phrases used, there was nothing on reflection or participatory evaluation. This is not unexpected for a development project, where project monitoring and evaluation is often top-down and extractive. In this regard, participatory evaluation may not be included as element of action research activities.
General impression of action research approach by the participants
The general impression of the respondents regarding the action research approach of the project in our study was that it is relevant and important for implementation of project activities. It was also observed by most respondents that some of the action research activities are visible and widely adopted by farmers, for example, forage cultivation and silage making. Another impression of the action research approach was that it has brought different disciplines together to work on project activities. One of the respondents put this clearly: “Action research has removed working in silos and has bridged the disciplinary gap in the project” (Respondent PS 2).
Another common observation by the respondents was that action research has provided a good platform for learning by BRIDGE project staff, university partners, extension services and farmers. This observation is consistent with one of the descriptions of action research presented above. Nearly all the external partners of BRIDGE interviewed acknowledged that they learnt about an action research approach for the first time through the project. One of the project partners said that: “Action research has put focus on research that addresses practical issues of the dairy farmers” (Respondent PP 5). This observation again re-emphasizes the “action” pillar of action research, as it is not a basic theoretical research, but applied, with focus on practical issues. However, there is the danger that action research may become “all action” and “no research”, especially when applied to development projects. So, it is important that action research practice should be informed by theory (Reason and Bradbury 2008).
Common action research methods used by BRIDGE
The most common methods or techniques used in action research activities by BRIDGE were Farmers’ Field Days (FFD), on-farm demonstration, and training (Fig. 3). FFD and demonstration were commonly used for cultivation of improved forage varieties and silage making in all the four project clusters. These techniques have been reported as common elements of participatory action research (Kindon et al. 2007) and they may enhance co-production of knowledge according to Omondi (2020). Other action research methods used by the project included training on different project related subjects, farmer to farmer learning, and campaigns, for example on milk safety. The least common techniques used in action research activities were laboratory analysis for feed and milk, rapid field tests for milk adulteration, and on-farm experimentation. Many of these methods are not the conventional approaches where the outside researcher sets the agenda, defines the questions to be asked, and administers questionnaires for later analysis. Most of the action research techniques under BRIDGE are participatory, which emphasizes shared knowledge. Some of these methods are context-specific, such as a campaign on milk safety, rapid field testing of milk quality, messaging to farmers, and sharing of extension packages, while others included adaptation of traditional social science methods like interviewing and surveying. The action research techniques used essentially mirrored the degree of participation in action research activities by the participants.
Participation in BRIDGE action research activities by stakeholders
Participation at all stages is fundamental for any action research project, as this is essential for fostering learning and achieving social transformation (Omondi 2020). Using the participation continuum proposed by Pretty et al. (1995), different action research activities by BRIDGE project were mapped (Fig. 4). Another conceptual framework for participation that could be used in this study is the “ladder of stakeholder participation” developed by Arnstein (1969) based on examples from three US Federal social programmes for citizens. We used Pretty et al.’s (1995) participation continuum as it is more suitable for action research in development projects. Expectedly, the level of participation by different stakeholders varied for different action research activities. The level of participation ranged from passive participation for on-farm experimentation on crop residue treatment to self-mobilization for improved forage cultivation and silage making, which dairy farmers adopted and carried out on their own. The high degree of participation in forage cultivation and silage making shows that when the benefit to stakeholders, in this case dairy farmers, is obvious and immediate, the level of participation in action research can be high.
For meaningful and effective participation in action research activities, Kindon et al. (2007) proposed three core ethical principles to be adhered to, namely respect for participants, beneficence and justice or equity. Respect for participants implies that people are treated as autonomous agents and that vulnerable participants are protected. Participation should not lead to domination by the community elites or to reinforcement of the pre-existing power hierarchies, which has not been beneficial to the community (Kindon et al. 2007). For many development projects, this is always a challenge, i.e., how to handle the existing power hierarchies in the community to prevent domination by the elites (Aragón and Glenzer 2017). The principle of beneficence, according to Kindon et al. (2007) should go beyond ‘doing no harm’ but maximize beneficial outcomes for participants and the community. Action research activities by BRIDGE, such as forage cultivation and silage making, were beneficial to dairy farmers in the project sites, as these led to significant increase in milk yield and consequently household income. The respondents also observed that the level of participation varied at different stages of the action research cycle for the project activities. For example, co-implementation of the action research activities generally tended to involve more and diverse participants, whereas design of specific interventions from jointly identified problem(s) tended to have fewer participants. The danger inherent in a low level of participation at any stage of the action research cycle is retention of researchers’ control, which can adversely affect ownership of the process and the outcomes (Brydon-Miller et al. 2003).
To ensure effective participation of stakeholders in environmental management, Reed (2008) proposed eight best practices, which are also valid for stakeholder participation in action research. These best practices include having an underpinning philosophy of participation that emphasizes empowerment, equity, trust and learning; embedding participation in the action research activities right from the onset, adequate representation of different groups of relevant stakeholders, and having clear objectives for the participatory process. Other best practices according to Reed (2008) are appropriate methods tailored to the local context, effective facilitation, integration of local and scientific knowledge, and institutionalization of participation. These best practices were applied to BRIDGE action research activities, though to a varying degree. For example, all relevant groups of stakeholders were adequately represented in BRIDGE action research activities and the methods were tailored to the local context. However, the institutionalization of participation needs to be strengthened.
SWOT analysis of BRIDGE action research activities
The action research activities covered in this SWOT analysis included forage cultivation, silage making, crop residue treatment, design of milk quality and safety programme, assessment of milk cooling facilities, assessment of dairy farm benchmarking, assessment of forage seed smart subsidy model, school milk programme, and consumer insight study on the potential of probiotic yoghurt (Table 5). Again, the common strength of the action research approach as mentioned by the interviewees for these activities was that of providing a platform for learning by different stakeholders. Learning by all participants in action research activities is essential for knowledge development, which could lead to purposeful individual and community action for socio-economic transformation (McNiff 2013). Lebesby and Benders (2020) observed that the primary purpose of action research is to produce practical knowledge for everyday life of the participants. The importance of learning by participants through action research was also reported by Omondi (2020) as essential for co-production of knowledge on climate change adaptation in the Mara River Basin in Kenya. The other reported strengths of BRIDGE action research activities were activity-specific. For example, strong participation was reported as one of the strengths of action research activities for forage cultivation and silage making (Table 5). This suggests that the degree of participation in an action research project or activity depends on its type or nature, potential benefit and cost, and socio-cultural contexts (Kindon et al. 2007).
The common weakness of BRIDGE' action research activities mentioned by the respondents was the absence of systematic monitoring and documentation of learning by participants (Table 5). The challenge with such monitoring and documentation is a common issue when action research activities are led by researchers whose preoccupation about data collection is on technical parameters at the expense of collecting data on processes. Omondi (2020) observed this challenge in her study that most researchers may not be familiar with the elements of self-critique and reflection, which are fundamental for monitoring and documentation of learning while collecting data in an action research project. In addressing this challenge of the competence of researchers in conducting action research, Brydon-Miller et al. (2003) suggested that there should be changes in researchers’ practices. For example, the traditional epistemological methods of research, which tend to be extractive in nature, may be inadequate to capture the processes of learning by participants in an action research project. Other reported weaknesses of BRIDGE action research activities tended to be activity-specific.
Some opportunities of using an action research approach for the BRIDGE project activities, as reported by the respondents, included enhancement of farmers’ innovation, awareness raising, and strong interest in the action research approach by public extension services and research partners. The project research partners acknowledged during the interview that they learnt about action research through the project and that their capacity in participatory action research techniques has been enhanced. This response by project partners on enhanced capacity is consistent with the observation by McNiff (2013) that action research projects often leave behind enhanced capacities of the participants in view of the emphasis on collaboration and learning. The enhancement of farmers’ innovation through participatory action research reported by the respondents has also been reported by Mponela et al. (2023) for a project on soil fertility management in Tanzania. For example in BRIDGE, farmers’ innovations in forage cultivation included planting of fruit trees and vegetables along with improved forage cultivars on the demonstration plot. According to the farmers, these innovations were to optimize land use in view of land shortage. In general, opportunities of action research activities were also activity-specific.
The main threat to action research activities by the project as reported by the respondents is government policy, particularly on land use, which generally tends to favour crop farming. This can undermine forage cultivation. One of the participants put it clearly: “Government officials see the grazing land for livestock as a waste, especially near Addis Ababa, the capital, and would rather prefer that it is used for cropping” (Respondent PS 6). The increasing competition for land is driven partly by demographic pressure, leading to expansion of crop fields into grazing areas in order to produce more food to feed the rapidly growing population (Balehegn et al. 2020). Other threats mentioned by the respondents were specific to different action research activities. For example, the reported potential threats to the school milk programme were volatility of price and quality of milk supplied to the school.
Table 5 Strength, Weakness, Opportunity and Threat analysis of the project action research activities
Activity
|
Strength
|
Weakness
|
Opportunity
|
Threat
|
Forage cultivation
|
- Demonstration plots managed by farmers
- Strong participation in Farmers’ Field Day (FFD)
- Provides platform for learning
- Facilitates adoption by farmers
- Support by extension services
|
- Over-reliance on FFD as platform for learning
- Poor monitoring and quality of data collected
- Slow response to emerging issue
|
- Enhances farmers’ innovations
- Awareness raising and strong interest in action research approach by public extension services and research partners
|
- Demonstration / FFD fatigue as it takes time
- Government land use policy that favours use of land for crops
|
Silage making
|
- Provides platform for learning
- Good linkage to forage cultivation
- Enhances adoption
- Strong participation at demonstration during FFD
- Support by extension services
|
- Limited technical knowledge
- Over-reliance on FFD as platform for learning
- Poor monitoring and quality of data collected
- Weak documentation of learning
|
- Availability of input service by agro-input dealers
- Can be a viable business
- Strong interest by commercial dairy farms
|
- Demonstration / FFD fatigue as it takes time
- Government land use policy
|
Crop residue treatment
|
- Good monitoring and data collection
- Building capacity of young researchers
- Provides platform for learning, to a limited extent
|
- Limited in scale
- Limited participation by farmers
- Often researcher-managed
- Weak documentation of learning
|
- Networking with relevant stakeholders
- Collaboration with research partners
|
- Relevance for practical application
- Funding of the on-farm experimentation
|
Assessment of milk cooling facilities
|
- Focused and in-depth
- Informs practical recommendations on specific issue
- Can provide platform for learning though to a limited extent
|
- May lack holistic perspective
- Tends to be extractive – collecting data from respondents
- Absence of documentation of learning
- Slow feedback to actors
|
- Can facilitate policy dialogue
- Can help identification of priority areas for implementation
- Opportunity for networking with experts on issue being addressed
|
- On the shelf reports, which may not be used
- Respondents’ fatigue with answering questions
|
Milk quality and safety programme
|
- Provides platform for learning
- Effective campaign on milk quality on awareness of milk adulteration
- Addresses public health issue
|
- Regularity of milk sample collection for lab analysis
- Absence of documentation of learning
|
- Alignment with regulations by regional government on food safety
- Strong interest by the government on food safety
- Increasing consumer awareness
|
- Influence of milk traders, which can undermine milk safety campaign
- Implementation of government policy on food safety
|
Dairy farm benchmarking
|
- Provides platform for learning
- Facilitates professionalization of dairy farm operations
|
- Limited technical knowledge
- No systematic monitoring and data collection
- Weak documentation of learning
|
- Openness of commercial dairy farmers to innovations
- Availability of feed cost reducing technologies, e.g., silage making
- Commercialization of dairy farm
|
- Government land use policy, which favours crop cultivation
- Increasing competition for land
- Profitability of dairy farm enterprise
|
Forage seed smart subsidy model
|
- Can provide platform for learning to a limited extent
- Kick-starting forage seed marketing
- Incentive for cultivation of improved forage
|
- Absence of systematic monitoring and data collection
- Weak documentation of learning
|
- Incentive for private sector engagement in forage seed market
- Scalability of the model
- Commercialization of dairy enterprise
|
- Weak national forage seed systems
- Government policy on seed systems
|
School milk programme
|
- Provides platform for learning and for timely feedback
- Good linkage of different actors
- Strong interest by students, parents and teachers
|
- School coverage is small as it is still at pilot stage
- No systematic monitoring and data collection
- Weak documentation of learning
|
- Well aligned with government policy to improve child nutrition and health
- Complements government School Feeding programme
- Can be funded through Corporate Social Responsibility
|
- Volatility in milk price
- Quality of milk supplied for School Milk programme
|
Consumer insight study on potential of probiotic yoghurt
|
- Focused and in-depth
- Can provide platform for learning
- Informs practical recommendations on consumption of dairy products
|
- Tends to be extractive
- Absence of documentation of learning
- Slow feedback to actors
|
- Increasing consumer awareness
- Alignment with government nutrition policy
|
- Implementation of recommendations from study
- On the shelf reports which may not be used
|
The results of the SWOT analysis of the action research activities of BRIDGE are generally as expected for a development project. Maximizing the strengths, minimizing the weaknesses and utilizing the opportunities of these action research activities are critical to social transformation and developmental changes in the project target communities. Generally, these results on strengths and weaknesses of, and opportunities and threats for action research as applied by BRIDGE are context-specific, depending on the activity and location. These findings are consistent with the observation of Brydon-Miller (2003) that one of the weaknesses of action research is its localism, which makes it difficult in intervening in large-scale social change efforts. This implies that action research may produce a great good in a local situation but may sometimes be difficult to extend beyond that local context. To enhance scaling of action research, a good documentation of the processes and the outcomes of action research activities is indispensable. Martin (2008) opined that two key challenges to be addressed to apply action research on a larger scale are sensemaking of the suitability to the project or project activities, and project design and implementation processes. For sensemaking, Martin (2008) proposed a number of questions to address such as: Who are the players? Where is the power? What will motivate the larger public to take interest in any change? The author further elaborated on the second challenge that the design and implementation processes should allow for engagement of multiple perspectives and support inquiry and learning. For large development projects applying action research, these two challenges enumerated by Martin (2008) must be adequately addressed.
Lessons for applying action research approach to agricultural development projects
Generally, the use of an action research approach by BRIDGE has contributed significantly to implementation of some project activities and has produced visible results, which are widely appreciated and adopted. These include for example cultivation of improved forage varieties and silage making. However, there was the impression from some of the respondents that the action research activities by the project are rather many and that there should be focus on fewer key issues or activities that can produce social transformation in the target communities. Given that the action research approach can be applied to a wide array of issues and fields, the danger is always that it may be applied even where other research methods may be more suitable. The romanticization of action research as a participatory approach may lead to de-legitimization of other research methods that are not participatory, which is one of the criticisms of action research (Kindon et al. 2007). Besides, action research approach may become a “fashion”, particularly in community development projects, and can be used unreflectively. On the issue of suitability of an action research approach to different project issues, one of the interviewees said: “Action research is suitable in addressing practical issues at farm and community levels but at higher scale (regional, national) another approach is needed” (Respondent PS 7). For community development projects where stakeholders’ participation is paramount, action research is not only relevant but necessary to achieve the desired goal of community empowerment and social change but when it comes to issues such as policy dialogue and influencing, and development planning, other approaches may be necessary, as these require data aggregation at higher levels.
Action research requires patience from the researchers and other participants as it often takes significant time (Kindon et al. 2007) and this was pointed out by some of the respondents during the interview. One of them said: “Action research activities tend to take too much time as the pace of implementation is slow” (Respondent PS 3). One of the interviewees asked rhetorically: “When do we exit the action research cycle as we cannot continue in the cycle perpetually?” (Respondent PP 4). This observation underscores the importance of a clear exit strategy for action research activities. Though the need for participation of all relevant stakeholders in action research activities is obvious, there should be some guiding rules or principles for stakeholders’ participation to avoid a tedious travelling through a winding alley during the action research process. This raises the need for necessary skills by the facilitators of the action research activities to achieve the jointly pre-defined objectives within the stipulated time. The length of time for action research activities should also be aligned with the project duration.
Another lesson from this study is the necessity of keeping focus on the bigger picture of the project. Following an action research approach may sometimes lead to new cycles of activities, which in principle is good but has the danger that multiplied activities can lead to loss of focus on the bigger picture of the project. For example in our study, action research activities on farm-level demonstrations of improved forage cultivation may be difficult to communicate in terms of the bigger picture of dairy sector transformation in Ethiopia for the BRIDGE project. The need for focus on the bigger picture of the project necessitates prioritization of action research activities, as this will facilitate communication with policy makers.
As part of lesson learning from this study, it is important to emphasize the need for adequate planning for monitoring and documentation of learning from the action research activities. It should be clarified that monitoring of learning by the action research participants is not the same as monitoring or tracking progress of the project activities in relation to the defined milestones, which is often well-planned by the project management. Generally, the development agencies and non-governmental organizations are good in the conventional monitoring and evaluation in the context of project performance-based accountability in response to growing demand by donors for demonstrated success of development projects (Estrella and Gaventa 1998). The focus here is on monitoring and systematic documentation of learning at different phases of action research cycle, for example farmers’ innovations to introduced technology. This monitoring and documentation of learning is one of the major weaknesses reported by the respondents in this study. Therefore, there should be a clear plan right from the onset of the action research on how to monitor and document learning by the participants, as this is necessary for critical reflection and participatory evaluation of the activities being carried out. Lack of documentation of learning during the different phases of action research cycle can be an obstacle to innovative and wider use of all that action research can offer (Kindon et al. 2007). To aid planning of monitoring and documentation of learning, a few guide questions should be addressed. First, what monitoring and learning activities should be done? Second, who should do what? Third, how should it be done? Fourth, when should it be done? Fifth, how much will it cost per activity? Besides, participatory monitoring and evaluation of action research activities is necessary to enhance participation of stakeholders, to share experience among the stakeholders through systematic documentation of processes and outcomes, and to empower the local people to initiate, control and take collective action (Estrella and Gaventa 1998; Dodd et al. 2023). Dodd et al. (2023) suggested that participatory monitoring and evaluation should be built on the foundational elements of local cultures and trust-based relationships among the stakeholders.
For many development projects in sub-Saharan Africa, action research activities are donor-driven and often face funding problems at the expiration of the project (Isobell et al. 2016). Therefore, building capacity of key stakeholders in participatory action research approaches is necessary to be able to continue with essential activities in the target communities after the end of the project, and it is a form of empowering them which will facilitate institutionalization of the action research approach. The capacity building should include both the theory and practice of action research; it is important that the trainees have a conceptual understanding of action research, so that activities are not carried out haphazardly (Khan et al. 2013).