Contextual Determinants of Learning Behaviour Differentials amongst Host-farmers of University-Student Outreach in Uganda

Advances in Agricultural Science 07 (2019), 02: 33-47

Contextual Determinants of Learning Behaviour Differentials amongst Host-farmers of University-Student Outreach in Uganda

Stephen W. Kalule 1,2, Haroon Sseguya 3,1, Duncan Ongeng 2 and Gabriel A. Karubanga 1*

Department of Extension and Innovations Studies, College of Agricultural and Environmental Sciences, Makerere University, P. O. Box 7062, Kampala, Uganda.
Faculty of Agriculture and Environment, Gulu University, P.O. Box 166, Gulu, Uganda.
International Institute of Tropical Agriculture (IITA), Regional Hub for Eastern Africa, Plot 25, Light Industrial Area, Mikocheni B, P.O Box 34441, Dar es Salaam, Tanzania.


In Sub-Saharan Africa, a knowledge void still exists on two contextual questions of university-community outreach: 1) how social demographics differences are related to farmer learning behaviour towards student outreach, and 2) how student outreach models compare with public and non-public extension services. A cross-section survey was used to obtain data from a sample of 283 respondents purposively selected from previous hosts of student outreach of Gulu University in Uganda. Results from Kruskal-Wallis method revealed that there were significant differences among host-farmers with respect to farmstead distance to the university for knowledge sharing (χ2 (2) = 8.5; P < 0.05) and giving feedback (χ2 (2) = 7.6; P < 0.05). Regarding the experience of participating in outreach program, significant differences among host-farmers were found in seeking information (χ2 (2) = 12.3; P < 0.01); knowledge sharing (χ2 (2) = 10.4; P < 0.01); seeking feedback (χ2 (2) = 16.4; P < 0.01) and giving feedback (χ2 (2) = 8.1; P < 0.05). Further, Friedman test results showed that host-farmers perceived the university-student outreach to be superior and significantly different from public and non-public agricultural extension. We conclude that university outreach is a useful service to farmer. However, its success in facilitating farmer learning is dependent on farmstead distance to the university and farmers’ level of experience of participating in university activities. We recommend more logistical support from governments to university outreach programs so that outreach services can efficiently complement public and non-public interventions in delivering community-based training and learning approaches.

KeywordsHigher educationCommunity linkagesFarmer learning

How to Cite: Kalule, S., Sseguya, H., Ongeng, D., & Karubanga, G. (2019). Contextual Determinants of Learning Behaviour Differentials amongst Host-farmers of University-Student Outreach in Uganda. Advances in Agricultural Science7(2), 33-47.    


In developing countries, university-community linkages are largely underdeveloped (Sherrard, 2016). It has been reported that the aims of many universities most especially in Sub-Saharan Africa, are quite disconnected from the local development needs (Blackie, 2016). This has been attributed to the fact that these universities experience diminished public funding for which the fiscal allocation is meant to simultaneously cater for the core functions including teaching, training, research and community outreach (Larsen, 2016). In the face of existing financial squeeze, community outreach has been least prioritized by universities in the planning and budgeting processes. This is well-evidenced in a host of scholarly work that often refer to community outreach as a relegated third mission of the university (Tefera, 2013; Muriisa, 2014). Inevitably, as suggested in Preece (2013), outreach activities in universities have remained marginal, barely contributing to community development. Quite a lot of researched work in terms of knowledge, technologies and innovations have been shelved within universities and are hardly exploited to facilitate learning in communities. Of particular importance, universities need to take into account the contextual factors in their target communities of outreach service provision. Key among these contextual factors are socio-demographics and alternative development interventions.  Such a responsive community outreach is likely to more effectively contribute to bridging the development gap in communities where such universities are located (Cloete and Maassen, 2015).

In recent times, and following good lessons learnt on community outreach from other universities, for instance, EARTH University in Costa Rica, rolling out student-oriented outreach models in African universities has gained prominence (Sherrard, 2016). For example, many South African universities are currently running service learning programs for enhanced community outreach (Preece, 2013). Elsewhere in Africa, agricultural-leaning universities are also implementing student-focused outreach models. A case in point, Egerton University in Kenya runs a farm attachment program (as form of community outreach) for student experiential learning which supports farmer learning (Mungai and Njuguna, 2016). In Uganda, student outreach services are well known in two universities. For example, Makerere University runs a student attachment program, in which undergraduate students train smallholder farmers for purposes of developing their capacities in farming business (Opolot et al., 2018). At Gulu University (the focus of current study), in what is known as the Student-Centered Outreach (SCO), students are attached to farmsteads for both own experiential learning and provision of advisory services to host-farmers (Kalule et al., 2016).

Conversely, recent research on university student outreach has concentrated on the influence of these outreach programs on entrepreneurial competences of host-communities (Opolot et al., 2018; Mukembo, 2017). Other studies have merely explored the role of resistance, stereotyping and gender-bias (Roberts and Edwards, 2017) and description of processes for student attachment (Sherrard and Alvarado, 2017) with limited focus on the underlying contextual determinants of learning behaviour differentials. In the SCO, there has been claims of social environment factors limiting ultimate student-farmer interaction for learning (Roberts and Edwards, 2017). Yet, there is still little attention paid to analyzing how the differences in farmer socio-demographics are related to farm-level learning behaviour, an issue that calls for further investigation. In addition, a research gap still exists on how the student-to-farmer university outreach compares with alternative agricultural development interventions namely public and non-public agricultural extension. This study therefore assesses the role of socio-demographic differentials for learning behaviour among host-farmers of university-student outreach. It also compares farmers’ perceptions of university student outreach to both public and non-public agricultural extension services using Gulu University (GU) in Uganda as a case. The results of the differences in farmers learning behaviour attributable to socio-demographic characteristics are important for segmented targeting of outreach services to host-communities in enhancing more even impact. In addition, comparative results on university outreach with public and non-public extension services can be insightful to policy brokers in lobbying for a more supportive policy environment for university outreach programs.


Design of Student-Centered Outreach at Gulu University

In the original design of the SCO, undergraduate students of agriculture were being attached to farmsteads situated within the 10km radius from the University campus (Odongo et al., 2017). However, the reorganization of this SCO following the introduction of graduate programs in the Faculty of Agriculture and Environment led to diversification of community attachment approaches. In the study of Kalule and Ongeng (2016), it was articulated that in addition to the SCO described above, graduate students, upon completion of the first year of their masters study programs, are attached to farmers’ organizations including associations and producer groups for a period of 2 -3 months. The core features of the SCO (as explained by Kalule et al., 2016) include: 1) students interact with farmers to facilitate learning; 2) farmers participate in problem communication and knowledge sharing with students; 3) students transmitting identified farming problems to the faculty; and 4) students transmit researched solutions to the community. In this outreach set up, it can be imagined that long term success and linkages in the SCO would most likely depend on adequate student-farmer engagement in information seeking, knowledge sharing, seeking and giving feedback on experiences, farming practices and technical knowledge.


Theoretical Framing

University outreach is associated with enhanced farmer learning and behavioural change through fostering student-farmer interactions. Farmer learning models therefore provide a strong basis for application to the context of university outreach programs whose main purpose is to support learning in farming communities. This study therefore, adapted and modified the learning behavior model advanced by Edmondson (1999). The model explains that learning behaviour comprises of four tenets: seeking information, knowledge sharing, seeking feedback and speaking about concerns (also known as giving feedback). However, how these occur in the context of university outreach remains unclear. It has been argued that in order to be able to learn and adapt (Karubanga et al., 2016), farmers seek to acquire complete, high quality and timely information related to their crops and livestock throughout the year (Kumar et al., 2018). Learning in relation to university outreach refers to a dynamic behavioural process of interaction and exchange among people (Kozlowski and Bell, 2007). For instance, Karubanga et al. (2017) and Kirkpatrick and Johns (2003) point out that farmers’ learning may involve interpersonal practices of seeking and sharing information either amongst the farmers or between the farmers and other stakeholders such as extension agents. These processes enable learners to compare new information with their experiences, creating new solutions for more efficiency and effectiveness (Carmeli et al., 2009).

In particular, information seeking refers to an individual way and manner of gathering and sourcing for information for personal use, knowledge updating and development (Igwe, 2012). Seeking information through university outreach is useful for farmers to overcome their inadequacies in knowledge of certain basic practices that may include technical, managerial and marketing (Owolade and Kayode, 2012). University outreach is further known for enhancing knowledge and experience sharing among students and farmers which also involve donating and collecting knowledge thereby facilitating farmer learning (Kamarudin et al., 2015). Lastly, the learning loop is only complete, if there is giving and receiving feedback from the source of knowledge (Hamilton, 2010). For instance, Milestad et al. (2010) point out the relevance of feedback for farmer learning where they reasoned that feedback enhances the adaptive capacity of the farmers in a given learning situation. They further explained that feedback construes new or revised interpretations of farmers’ experiences and in effect changes these farmers’ attitudes for decision making and behaviour for production.

Previous research has demonstrated that farmers’ acquisition of information and knowledge is connected to social demographic characteristics of the farmers (Acheampong et al., 2017). For instance, key among the social demographics that are likely to affect farmer learning behaviour are: sex, age and education status. Besides, there are also some factors which are peculiar to the context of student-to-farmer University outreach that are also worth testing. These include among others: participation in farm-labour sharing, experience of hosting students on the farm and farmstead distance to the university. Therefore, this study hypothesized as follows:

H1: Host-farmers of university student outreach are significantly different in learning behaviour for socio-demographic factors

H2: Host-farmers of students of university outreach are significantly different in learning behaviour in terms of a) level of experience of hosting students on their farms; and b) farmstead distance to the university

Another key assumption in this study is that farmers are confronted with different sources of information but with varying levels of efficiency in service delivery. Extant literature has shown that farming households experience underservice from public agricultural extension. Seemingly, alternative development interventions, for instance, university outreach would help to address the service gap. It is presumed that comparative perceptions on facilitating farm-level learning amongst farmers differs between university-student outreach and other development interventions of agricultural extension service and thus, this study hypothesized as shown below:

H3: Farmers’ perceptions of own learning behaviour in the student-to-farmer university outreach is significantly different from that of public and non-public agricultural extension services



Research Design

A cross-sectional survey was conducted on a sample purposively selected from previous host-farmers of the Student-Centered Outreach (SCO) program of Gulu University. In the SCO, students are meant to commute to and from the farmsteads, at least once in every two weeks for a period of not less than a year.  Selected farmers had previously received students onto farms during the period of 2007 to 2016. Study participants were drawn from neighbouring districts of Gulu and Omoro in Northern Uganda. Besides, these two districts are in close proximity to Gulu University. The location of Gulu University is largely rural in nature and as such, there are many farming households in the neighbourhood of the University campus, a factor that has enabled ease of reach of farming households. In this study area, apart from university outreach, common development interventions for farmer agricultural extension service include:  the public agricultural extension service, non-governmental organization (NGO) extension and the farmer-to-farmer extension.

Literature shows that there are two common paradigms that guide the choice of research design in social sciences and these are positivism and constructivism. In the positivist approach, the research argument is based on widely accepted principles (knowledge or theories about human behaviour) and normally, the research process ends with deductive reasoning guided by rules and laws (Creswell, 2014; Saunders et al., 2012). For the constructivist approach, the research inquiry involves moving from ‘specific to general phenomenon’ using inductive reasoning, allowing to gain a subjective sense of the environment in which study participants live (Trochim 2006). As such, the positivists start with theory and use quantitative data for making inferences while constructivists start with observation through qualitative data collection with concurrent analysis resulting into inductive reasoning (theorizing). In this study, the pragmatism approach as articulated by Creswell (2014) was preferred. Pragmatism transcends the positivist and constructivist approaches and allows the researchers to combine the quantitative and qualitative research methods. The pragmatism approach was found necessary for this study for two reasons: 1) theory testing to deductively provide evidence to either support or refute the research hypotheses derived from studying literature; and 2) gather qualitative information for detailed explanation of issues underlying positive perception of the university-student outreach as compared to public and non-public agricultural extension among program participants.

A sample of 283 respondents was computed following the sample size estimation method suggested by Krejcie and Morgan (1970). Subsequently, a purposive and multi-stage sampling approach was used to arrive at the final study participants. The reason for purposive sampling was to target only the program participants of university-student outreach for making deductions. First, respondents were selected from the sub counties of Paicho and Bungatira (Gulu District) and Koro and Bobi (Omoro District). In each sub county, farmer groups that had received students were identified for sampling. Subsequently, a list of farmers from each group was used to systematically sample respondents. To ensure representativeness, proportional sampling was used to distribute the sample between the selected districts. Up to 65% of SCO program participants come from Gulu District, possibly, because of the closeness of this district to Gulu University. As such, 181 participants representing 64% (Paicho = 94 and Bungatira = 87) were respectively sampled from Gulu District. That meant that 102 representing 36% (Koro = 61 and Bobi = 41) were sampled from Omoro District.


Data and Data Sources

Data were collected using semi-structured questionnaire. The questionnaire captured both Likert scale data as well as qualitative information that allowed respondents freedom of expression. Initially, the questionnaire was pretested to assess its validity and reliability. Pretesting of the questionnaire was done in Lakwana Sub County (Omoro District), which neighbours the study area on a sample of 33 respondents. Results from the pretested questionnaire were then used to reword questions and/or items for improved consistency and clarity of questions to respondents. Construct reliability (CR) was used to check for measurement reliability of the construct of farmer learning behaviour.  The CR index was 0.93, which was above the minimum threshold value of 0.7 and thus, satisfying the condition of measurement validity. A key section in the questionnaire captured responses on farmers’ learning behaviour. This construct comprised of four indicators and these were: seeking information, knowledge sharing, seeking feedback and giving feedback to the service provider.  Each of these four indicators was measured using four items, adapted and modified from Edmondson (1999) to suit the context of the student-to-farmer university outreach. The same indicators and items were repeated for each of the three other sources of farmer agricultural extension service namely public extension service, NGO extension and farmer-to-farmer extension. In all, a total of 64 items were used to measure farmers’ learning behaviour. A sample item from the domain of information seeking read as follows: “I seek information on farming practices from university students”. All items were rated on a 5-point Likert scale (1-5, where 1 = not at all and 5 = always).

Social demographic factors used in this study include: sex of the household head (captured as female = 1 and male = 2) and education status of household head (no-formal education = 1; primary education = 2 and post-primary education = 3). Other variables include: age of household head in years, participation in rotational farm-labour sharing in the community (if participant = 1; otherwise = 2, experience of hosting students on the farm (number of times the farmer had received student since inception of university outreach). Lastly, farmstead distance to the University was captured in km.


Data Analysis

Data were entered, processed and transformed in the computer statistical package of SPSS. Means and standard deviations were computed from Likert scale data on the indicators of farmer learning behaviour. Preceding testing for statistical differences in this study, correlation analysis was done to assess the possibility of multicollinearity. Given that correlates between seeking information, knowledge sharing, seeking feedback and giving feedback ranged from .69 to .77 (see appendix, Table 1), it suggested a risk of multicollinearity as explained by Hamilton (2006). As such parametric methods were unsuitable for making inferences and thus, non-parametric methods namely Kruskal-Wallis, Friedman and Wilcoxon signed-rank tests were preferred for subsequent analysis. Effect sizes were also computed manually following the method suggested by Field (2009). Effect size were calculated as follows:



where Z is the statistic that tests for significance of differences from Wilcoxon signed-rank test and is the square root of sample size. The study also adapted Cohen (1988) criteria for evaluation of effect sizes (ρ). This criteria explains that the magnitudes of such effect sizes are considered small, if ρ≤ 0.1. These effect sizes are medium in size, if ρ≤ .3 and they are large, if ρ≥.5.  Lastly, the statistical findings were corroborated with qualitative information collected from some respondents in order to enhance the validity of study results.


Results and discussion

Socio-demographic differences in farmer learning towards student outreach

Results of differences in farmer learning behavior towards student outreach are presented in Table 1. The results show that respondents were significantly different in learning behaviour with respect to experience of hosting students.

The differences were detected in seeking information (χ2 (2) = 12.3; P < 0.01); knowledge sharing (χ2 (2) = 10.4; P < 0.01); seeking feedback (χ2 (2) = 16.4; P < 0.01) and giving feedback (χ2 (2) = 8.1; P < 0.05). These results supported the hypothesis (H2a) that farmers’ learning behaviour was significantly different for the level of farmers’ experience of hosting University outreach students. There was also significant differences in farmstead distance to the university for knowledge sharing (χ2 (2) = 8.5; P < 0.05) and giving feedback (χ2 (2) = 7.6; P < 0.05). These findings supported the hypothesis (H2b) on existence of significant differences in farmstead distance to the University for the behaviours of knowledge sharing and giving feedback.

However, hypothesis (H2b) was not supported in respect of seeking information and seeking feedback, for which no significant differences were found. Participation in farm-labour sharing was only significantly different for knowledge sharing (χ2 (1) = 4.7; P < 0.05) and was not statistically significant in any of seeking information, seeking feedback and giving feedback. However, for the variables of sex, age and education status of the household head, there was no significant differences across the four indicators of farmers’ learning behaviour, and thus, hypothesis (H1) was not supported. This implied that the three variables were independent of farmers’ learning behaviour.

A comparison of short-distance and medium-distance farmers revealed that significant differences were only in giving feedback (χ2 (1) = 3.9; P <0.05). For the other three indicators of farmers’ learning behaviour, no significant differences were found. Still, there was no significant differences across the four indicators of learning behaviour between medium and long-distance farmers. Thus, the effect size in the comparison of short and medium-distance as well as the medium to long-distance farmers, which ranged

Table 1. Socio-demographic differences for farmer learning behaviour

Variables N Kruskal-Wallis Test for Group Differences
Seeking Information Knowledge Sharing Seeking feedback Giving feedback
MR χ2 MR χ2 MR χ2 MR χ2
Sex of household head
Male 154 139.2 141.1 147.0 141.0
Female 129 145.3 0.4 143.1 0.0 136.0 1.3 143.2 0.0
Farm-labour sharing
Participant farmer 195 139.4 135.0 135.9 136.2
Non-participant farmer 88 147.9 0.7 157.6 4.7* 155.5 3.5 155.0 3.3
Age of household head
≤ 30 years 69 148.2 142.5 153.2 141.2
> 30 – 50 years 123 139.5 0.5 142.2 0.1 136.1 2.0 144.7 0.3
> 50 years 91 140.7 141.3 141.6 139.0
Education of household head
Pre-primary education 27 129.5 133.6 124.1 116.8
Primary education 160 146.1 3.8 143.2 1.8 146.2 0.3 148.4 1.2
Post-primary education 96 138.7 142.3 140.1 138.4
Farmstead dist. to the University
Short distance (≤ 10 km) 142 153.4 154.4 149.9 153.3
Medium dist. (>10 – ≤ 20 km) 81 130.8 5.6 137.6 8.5* 142.8 4.9 139.0 7.6*
Long distance (> 20 km) 60 130.2 118.7 122.3 119.2
Experience of hosting students
Low (≤ 3 times) 203 132.3 133.9 131.5 133.9
Medium (4 – 6 times) 57 158.4 12.3** 152.3 10.4** 156.5 16.4** 157.0 8.1*
High (> 6 times) 23 187.2 188.2 198.5 176.3

* & ** means significant at P < 0.01 & P < 0.05 respectively


Table 2. Post hoc tests for pairwise comparisons

Pairwise Comparisons Farmers’ learning Behaviour (N = 283)
Seeking Information Knowledge Sharing Seeking feedback Giving feedback
χ2 ρ χ2 ρ χ2 ρ χ2 ρ
Farmstead distance to the University  
Short (≤ 10 km) & Medium (>10 – ≤ 20 km) 1.6 0.1 0.5 0.0 2.2 0.1 3.9* 0.3
Medium (>10 – ≤ 20 km) & Long (> 20 km) 0.0 0.0 1.8 0.2 2.6 0.2 2.1 0.2
Short (≤ 10 km) & Long (> 20 km) 3.5 0.2 8.3** 0.6 4.5* 0.3 7.4** 0.5
Experience of hosting university students  
Low (≤ 3 times) & Medium (4 – 6 times) 3.6 0.2 4.4* 0.3 2.3 0.1 4.8** 0.3
Medium (4 – 6 times) & High (> 6 times) 1.0 0.1 6.0* 0.7 3.5 0.4 3.1 0.3
Low (≤ 2 times) & High (> 4 times) 5.5* 0.4 13.3** 0.9 9.1** 0.6 8.8** 0.6

* & ** means significant at P < 0.01 & P < 0.05 respectively


Table 3. Comparison of university student outreach to public and non-public agricultural extension services

Farmers’ Information Sources Means (SD); n= 283
Information seeking Knowledge Sharing Feedback seeking Giving feedback
Student Outreach 3.91 (0.99) 3.99 (0.92) 3.94 (1.01) 4.02 (0.92)
Public Ext. Service 2.49 (1.15) 2.54 (1.14) 2.46 (1.16) 2.44 (1.14)
NGO Ext. Service 3.30 (1.04) 3.36 (1.02) 3.31 (1.06) 3.39 (1.08)
F2F Ext. Service 3.11 (1.23) 3.21 (1.24) 3.12 (1.23) 3.15 (1.21)


from 0.0 to 0.3, were all small as affirmed by Cohen (1988). The effect sizes between short-distance and long-distance were above small criterion; seeking information (ρ = 0.2), seeking feedback (ρ = 0.3). Large effect sizes were detected in   giving feedback (ρ = 0.5) and knowledge sharing (ρ = 0.6). In the case of knowledge sharing, the result implied that the differences in short and long-distances from the university amongst farmers accounted for 60% variation in farmers’ knowledge sharing with students.

Results of post hoc analysis for the variables, which had three comparison groups and exhibited significant differences in at least two indicators of learning behaviour are presented in Table 2. For example, farmstead distance to the university, the results show that farmers within short-distance radius from the university (≤ 10km) performed better in three indicators of learning behaviour compared to those located in long-distance to the university (>20km). Thus, farmsteads within short distance to the University were significantly different from long-distance farmsteads in knowledge sharing (χ2 (1) = 8.3; P<0.01), feedback seeking (χ2 (1) = 4.5; P<0.05) and giving feedback (χ2 (1) = 7.4; P<0.01).

Regarding the experience of hosting university outreach students, significant differences were found across the four indicators of farmers’ learning behaviour between low and high number of times of hosting students. The results were as follows: seeking information (χ2 (1) = 5.5; P < 0.05), knowledge sharing (χ2 (1) = 13.3; P < 0.01), seeking feedback (χ2 (1) = 9.1; P < 0.01) and giving feedback (χ2 (1) = 8.8; P < 0.01).

There was also significant differences between low and medium experience of hosting students for knowledge sharing (χ2 (1) = 4.4; P < 0.05) and giving feedback (χ2 (1) = 4.8; P < 0.01). The comparison between medium and high, the significant differences were only found in knowledge sharing (χ2 (1) = 6.0; P < 0.05). The effect sizes were biggest in the comparison between low experience and high experience. These effect sizes ranged from medium (ρ = 0.4 in information seeking) to large (ρ = 0.6, 0.6 and 0.9 in seeking feedback, giving feedback and knowledge sharing respectively). For knowledge sharing, as a case, the results meant that the difference in low and high experience of hosting students amongst farmers accounts for 90% of the variation in the learning behaviour of knowledge sharing.

Elsewhere, studies have shown that farmers in close proximity to knowledge institutions, for instance, universities and research institutions, and knowledge sharing platforms have a relative advantage in accessing information and knowledge, and tend to exhibit better learning (Mashavave et al., 2013). It is likely that farmsteads that are closer to the university are regularly reached by students because of ease of commuting to and from these farmsteads. It is also imaginable that long-experience farmers could have realized the value of participating in student outreach activities. In turn, these farmers could have been more encouraged to exhibit positive learning behaviour towards student outreach. These results support earlier findings that have suggested that host-communities tend to have positive perceptions towards student outreach because these communities assume that students possess valuable knowledge (Shuda and Kearns-Sixsmith, 2009).


Comparison of student-to-farmer university outreach to other farmer extension services

Results in Table 3 reveal that the student-to-farmer university outreach was rated highest amongst the four domains of farmer learning behaviour. Results from Likert scale data indicate that with respect to university student outreach, farmers rated least information seeking (M = 3.91; SD = 0.99) while the highest rated indicator was giving feedback (M = 4.02; SD = 0.92). Farmers rated lowest the four domains of learning behaviour with respect to public extension service, ranging from giving feedback (M = 2.44; SD = 1.14) to knowledge sharing (M = 2.54; SD = 1.14). NGO extension service was rated second highest from information seeking (M = 3.30; SD = 1.04) to giving feedback (M = 3.39; SD = 1.08). The second lowest rated service provider was farmer extension whose rating ranged from information seeking (M = 3.11; SD = 1.23) to giving feedback (M = 3.15; SD = 1.21).

A follow-up with tests on differences in farmer learning behaviour between the student outreach and other service providers are presented in Table 4. Friedman test results showed that information seeking (χ2(3) = 180.38; P < 0.01); knowledge sharing (χ2(3) = 170.91; P < 0.01), feedback seeking (χ2(3) = 186.62; P < 0.01) and giving feedback (χ2(3) = 190.68; P < 0.01) were significantly different across the four sources of farmers’ information. Post hoc analysis, in which Wilcoxon signed-rank tests were conducted with a Bonferroni correction applied, resulted in a significance level set at p = 0.0125. Accordingly, there was a statistically significant reduction in information seeking from student outreach to public extension service (Z = -12.81; P< 0.0125), NGO extension service (Z = -7.12; P<0.0125) and farmer extension (Z = -7.38; P<0.0125). Accordingly, these results supported the hypothesis (H3) that farmers’ perceptions of own learning behaviour in the student-to-farmer university outreach is significantly different from that of public and non-public agricultural extension services.

Regarding the knowledge sharing, there was significant differences in favour of the student outreach when compared to public extension (Z = -12.60; P< 0.0125), NGO extension (Z = -7.24; P<0.0125), and farmer extension (Z = -7.40; P<0.0125). A similar pattern of significant differences was also found for seeking feedback and giving feedback when student outreach was compared with other three service providers examined in this study. There was also significant differences in favour of NGO extension compared to public extension for seeking information (Z = -8.75; P<0.0125), knowledge sharing (Z = -8.71; P<0.0125), seeking feedback (Z = -9.14; P<0.0125) and giving feedback (Z = -9.54; P<0.0125). Surprisingly, when NGO extension was compared to farmers’ extension, significant differences were only found in giving feedback (Z = -2.88; P<0.0125) and for other three indicators, there was no statistical differences. Lastly, the respondents rated farmer extension better than public extension with significant differences found in seeking information (Z = -5.80; P<0.0125), knowledge sharing (Z = -5.93; P<0.0125), seeking feedback (Z = -5.84; P<0.0125) and giving feedback (Z = -6.35; P<0.0125).

A further examination of the triggers of superior rating of student outreach for farmer learning behaviour revealed more frequent student-farmer contacts compared to public and non-public extension service. These farmers also reported that they expected better knowledge from students compared to their peers.  In most cases, the respondents expressed preference of student-to-farmer outreach to public extension service, as some farmers articulated below:

“… the students are more frequent and are easier to access compared to government workers…”, farmer, Boke Village, Bungatira Sub county, November 2017.

“… I think students are studying and always reach us with adequate knowledge, and are easier to access too. Government workers, provide free inputs but do not come down to individual farmers to guide us on the use of such inputs…”, farmer, Holly-Rosary village, Laroo Sub County, November 2017

The lower rating of NGO extension was attributed to downscaling of services to farmers. That aside, much as these farmers recognized that peer farmers (farmer-to-farmer extension) were more accessible, they still preferred university outreach as articulated by some respondents below:

“… NGOs give information on managing village saving and loan associations. However, of late they have not been coming to us…”, farmer, Kalam-Omida, Bobi Sub County. December, 2017.

Table 4. Friedman and Wilcoxon signed-rank tests for differences

Tests/ Comparisons Information Seeking Knowledge Sharing Feedback Seeking Giving Feedback
Friedman Test        
Chi-Square/ χ2(3) 180.38** 170.91** 186.62** 190.68**
Wilcoxon Signed-Rank Test Z-Val. ρ Z-Val. ρ Z-Val. ρ Z-Val. ρ
Student Outreach – Public Extension -12.81* -.76 -12.60* -.75 -13.04* -.77 -12.94* -.77
Student Outreach – NGO Extension -7.12* -.42 -7.24* -.43 -7.23* -.43 -7.02* -.42
Student Outreach – Farmer Extension -7.38* -.44 -7.40* -.44 -7.67* -.46  -8.31* -.49
NGO Extension. – Public Extension -8.75* -.52 -8.71* -.52 -9.14* -.54 -9.54* -.57
Public Extension – Farmer Extension -5.80* -.34 -5.93* -.35 -5.84* -.35 -6.35* -.38
NGO Extension – Farmer Extension -2.08 -.12 -1.67 -.10 -2.18 -.13 -2.88* -.17

* Significantly different at P = 0.0125 (after Bonferroni adjustment for multiple comparisons); ** Difference within

the farmers’ learning behaviour was significant (P=0.01)


“… peer farmers only give knowledge learnt from other sources. But university students have been taught and have more new knowledge to share compared to peer farmers”, farmer, Boke Village, Bungatira Sub county, November 2017.

“… peer farmers are more close to me and provide knowledge on a daily basis …. University students, however, have always given us first-hand knowledge on farming”, farmer, Holly-Rosary, Laroo Division, November 2017.

Conversely, effect sizes were equally interesting. In the comparison of student outreach to public extension service, the effect sizes were as follows: knowledge sharing (-.75), information seeking (-.76), feedback seeking (-.77) and giving feedback (-.77). This meant that the differences between student outreach and public extension service explained the variation in farmers’ learning behaviour ranging from 75% up to 77%. In the comparisons between student outreach and NGO extension, the effect sizes ranged from .42 – .43 while between student outreach and farmer-to-farmer extension, the effects ranged from .44 to .49. Surprising findings were detected in the comparisons of NGO extension service and farmer-to-farmer extension, in which effect sizes were small (.10 – .17). However, the effects in the comparison of NGO and public extension were larger than medium sizes (.52 – .57).

The findings corroborated with previous studies, which have reported inadequacies in linkages between farmers on one hand, and on the other hand, public extension systems and research (Sewell et al., 2017; Klerx et al., 2017). Surprisingly, NGO agricultural extension service did not significantly differ from the farmer-to-farmer extension. This could be because most structures of farmer-to-farmer extension have been established by the NGOs themselves, and as such, farmers find little differences between the two forms of service. The high rating of student outreach had connection to frequent student visits to farms and expected knowledge quality. This finding is in support of the studies which have reported that frequent contacts between extension agents and farmers’ influences farmers’ information seeking and learning (Goud and Ram, 2018; Dambazau et al. 2015). In relation to farmer extension, respondents acknowledged that their peers were equally easily accessible and were useful in the learning process, a finding in support of results in Karubanga et al. (2017).


Conclusion and Recommendations

It has been established that farmstead distance to the university and the length of experience of hosting university students significantly influenced farmer learning behaviour towards student outreach. Farmers within short-distance from the university are better at seeking information, knowledge sharing, seeking and giving feedback in the student-to-farmer university outreach.  Similarly, farmers with long experience of participating in student outreach are better than their counterparts at interacting with students for seeking information, knowledge sharing, seeking feedback and giving feedback. This study also demonstrated that the student-to-farmer university outreach is a valuable service to host communities. It has been shown that student outreach services have a relative advantage at influencing farmer learning behaviour in areas that are proximal to the university compared to public agricultural extension, NGO extension and farmer-to-farmer extension. However, it is also evident that the comparative advantage of student outreach services over public extension service is because these farmers experience more contacts with students, a fact that can be attributed to the rural nature of Gulu University location.

 This study contributes by enriching the literature on farmers’ learning behaviour. The four domains of farmers’ learning namely information seeking, knowledge sharing, feedback seeking and giving feedback have been applied to assess how social demographic differences are related to farmer learning behaviour towards student outreach. Further, these concepts have been used to compare farmers’ perceptions of the student-to-farmer university outreach with public and non-public extension service. From a managerial perspective, it has been shown that university student outreach to farmers brings more meaningful impact and lasting learning relationships with the community when three conditions are meet. The conditions are: 1) frequent student visits to the farming households; 2) students exhibiting sound knowledge quality; and 3) nearness of farmsteads to the university. Thus, student outreach models can be exploited to complement public and non-public extension approaches in supporting farm-level learning and innovation. This is more likely to succeed with rural-based universities whose campuses tend to be in close proximity to farming households.

A potential limitation in this study is that all data were obtained from previous hosts of agricultural students of Gulu University during university outreach. This gave no room for comparison of participants and non-participant farmers of university outreach. The implication of this scenario is that the insights of non-program participants were not available in the comparison of the various farmers’ information sources. Nevertheless, the results are still relevant for application in contexts of student oriented outreach services.

We therefore recommend better logistical support to university outreach programs so as to improve delivery of community outreach and complement existing public extension systems. For more impact, it is also recommended that universities take a lead in facilitating farmers’ learning spaces that bring together extension service providers, private sector actors and financial service providers which ensures complementarity of services. Future research should evaluate the role of human and financial resources in supporting student-centered outreach to facilitate farmers’ learning for innovation in Uganda and elsewhere in the world. In addition, studies should further examine how farmers’ motivational and social factors influence farmers’ active participation and gaining from the student-centered outreach.



This article resulted from the PhD research supported by the Regional Universities Forum for Capacity Building in Agriculture – RUFORUM (Grant Number: RU 2014 NG 15). Support for research was also provided by Carnegie Corporation of New York through RUFORUM Post-Doctoral programs.


Declaration of no Conflict of Interest

The authors have no conflict of interest to disclose



Acheampong, L.D., Frimpong, B.N., Adu‑Appiah, A., Asante, B.O., and Asante, M. D. (2017). Assessing the information seeking behaviour and utilization of rice farmers in the Ejisu‑Juaben municipality of Ashanti Region of Ghana. Agriculture and Food Security 6 – 38.

Blackie, M. (2016). Higher education and rural development in Africa: Building a new Institutional framework. African Journal of Rural Development 1(2):115 – 125.

Carmeli, A., Brueller, D., and Dutton, J.E. (2009). Learning Behaviours in the Workplace: The Role of High-quality Interpersonal Relationships and Psychological Safety. Systems Research and Behavioral Science 26:81-98.

Cloete, N., and Maassen, P. (2015). Roles of Universities and the African Context. Knowledge Production and Contradictory Functions in African Higher Education. Cape Town, 2015 African Minds.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.

Dambazau, S.A., Srivastava, J.P., and Tajudeen, A. (2015). Communication Behaviour of Potato Growers in Allahabad district of Uttar Pradesh. Journal of Communication Studies 33(2): 48-55.

Edmondson. A.C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2):350-383.

Field, A. (2009). Discovering Statistics Using SPSS. 3rd Edition, Sage Publications Ltd., London.

Franz, N., Piercy, F., Donaldson, J., Richard, R., and Westbrook, J. (2010). How Farmers Learn: Implications for Agricultural Educators. Journal of Rural Sciences 25:37–59.

Gyenfie, J.N. (2014).  Linking Programme Planning Approaches to Livelihood Outcomes of Farmers: A Case Study of Agricultural Extension Services in Ghana. PhD Thesis. University of Ghana.

Goud, E. R., and Ram, D. (2018). Comparative profile of Communication Behaviour among the Rice growers in Imphal West District of Manipur, India. International Journal of Current Microbiology and Applied Sciences 75(5):2273-2279.

Hamilton, N. A. (2010). Learning to Learn with Farmers. A case study of an adult learning extension project conducted in Queensland, Australia 1990 – 1995. Thesis Landbouw Universiteit Wageningen. ISBN 90-5485-476-6.

Hamilton, L.C. (2006). Statistics with Stata: Updated for Version 9; Thomson-Brooks/Cole: Belmont, CA, USA.

Igwe, K.N. (2012) Introduction to information science. Offa: Department of library and information science, Federal Polytechnic, Offa.

Kamarudin, H.D., Aziz, N. E.A., Zaini, M.K., Ariff, Z.M.N.Z. (2015). Exploring Knowledge Sharing Practices among Paddy Farmers towards Building a Foundation for Knowledge Creation. International Journal of Social Science and Humanity 5(1):112 – 115.

Kalule, W.S., Odongo, W., Kule, E., Ndyomugyenyi, K.E., Omara, P., and Ongeng, D. (2016). Conceptualizing the Student-Centered Outreach Model for Experiential Learning and Community Transformation. African Journal of Rural Development 1(3): 219 – 227.

Kalule, W.S., and Ongeng, D. (2016). Embedding community engagement in University Training Curricula: Experiences from Gulu University in Uganda. RUFORUM Working Document Series 14 (1): 409-413.

Karubanga, G., Kibwika, P., Okry, F., and Sseguya, H. (2016). Empowering farmers to learn and innovate through integration of video and face-to-face extension approaches: The case of rice farmers in Uganda. Cogent Food Agric. 2(1): 1-12.

Karubanga, G., Kibwika, P., Okry, F., and Sseguya, H. (2017). How farmer videos trigger social learning to enhance innovation among smallholder rice farmers in Uganda enhance innovation among smallholder rice. Cogent Food and Agriculture 9:1–16.

Kilpatrick, S., and Johns, S. (2003). How Farmers Learn: Different Approaches to Change. The Journal of Agricultural Education and Extension 9(4):151 – 164.

Klerkx, L., Seuneke, P., De Wolf, P., and Rossing, W.A.H. (2017). Replication and Translation of Co-innovation: The influence of Institution Context in Large International Participatory Research Projects. Land Use Policy 61: 276–292.

Knill, C., and Tosun, J. (2011). Policy-making. Oxford University Press. ISBN 978-0-19-957497-1

Kozlowski, S.W.J., and Bell, B.S. (2007). Team learning, development, and adaptation. In Work Group Learning: Understanding, Improving and Assessing how Groups Learn in Organizations, Sessa VI, London M (eds). Erlbaum: Mahwah NJ; 15–44.

Kumar, S., Sangeetha, V., Singh, P., Burman, R.R., Bhowmik, A., and Meera, S.N. (2018). Stakeholders’ Information needs, Information Searching and Sharing Behaviour about Rice related Information through Rice Knowledge Management Portal. International Journal of Current Microbiology and Applied Sciences 7(1): 3001-3015.

Krejcie, R.V., and Morgan, D.W. (1970). Determining sample size for research activities. Educational and Psychological Measurement 30: 607-610.

Larsen, C.E.S. (2016). Reflections on the incentive structures at African universities and their impact on the direction and performance of the continent’s development. African Journal of Rural Development 1(2): 127 – 137.

Mashavave, T., Mapfumo, P., Mtambanengwe, F., Gwandu, T., and Siziba, S. (2013). Interaction patterns determining improved information and knowledge sharing among smallholder farmers. African Journal of Agricultural and Resource Economics 8(1): 1-12.

Milestad, R., Westberg, L., Geber, U., and Björklund J (2010). Enhancing adaptive capacity in food systems: learning at farmers’ markets in Sweden. Ecology and Society 15(3): 29.

Mukembo, S. C. (2017). Equipping Youth with Agripreneurship and Other Valuable Life Skills by Linking Secondary Agricultural Education to Communities for Improved Livelihoods: A Comparative Analysis of Project-Based Learning in Uganda. PhD Dissertation, Oklahoma State University.

Mungai, N.W., and Njuguna, V.W. (2016). Linking universities to communities: The case of  Egerton University. RUFORUM Working Document Series 14 (1): 415-420.

Muriisa, R.K. (2015). Rethinking the Role of Universities in Africa: Leadership as a Missing Link in Explaining University Performance in Uganda. JHEA/RESA, 12(1): 69-92.

Odongo. W., Kalule ,W.S., Kule, K.E., Ndyomugyenyi, E.K., Omara, P., and Ongeng D. (2017). Responsiveness of Agricultural Training Curricula in African Universities to Labour Market Needs: the case of Gulu University in Uganda. African Journal of Rural Development, 2(1): 67-76.

Opolot, H.N, Isubikalu, P., Obaa, B.B., and Ebanyat, P. (2018 in Press). Influence of university Entrepreneurship training on farmers competences for improved productivity and market access in Uganda. Cogent Food and Agriculture.

Owolade, E.O., and Kayode, A. (2012). Information-seeking behavior and utilization among snail farmers in Oyo State, Nigeria: Implications for sustainable animal production. Journal of International Agricultural and Extension Education 19(3):39-49.

Preece, J. (2013). Service Learning and Community Engagement in South African Universities: Towards an ‘Adaptive Engagement’ Approach. Alternation Special Edition, 9: 265 – 291.

Roberts, R., and Edwards, M.C. (2017). Challenges to Sustaining University-Community Partnerships in War-Torn, Northern Uganda: Investigating Resistance, Negative Stereotyping, and Gender Bias in Agricultural Students’ Attachments. Journal of International Agricultural and Extension Education 24(2): 4 – 21.

Saunders, M. N., Lewis, P. and Tornhill, A. (2012). Research methods for business students. Harlow: Pearson Education.

Sewell, A.M., Hartnett, M.K., Gray, D.I., Blair, H.T., Kemp, P.D., Kenyon, P.R., Morris, S.T., and Wood, B.A. (2017). Using educational theory and research to refine agricultural extension: affordances and barriers for farmers’ learning and practice change, The Journal of Agricultural Education and Extension 23(4): 313-333.

Shuda, J., and Kearns-Sixsmith, D. (2009). Outreach: empowering students and teachers to fish outside the box. Zebrafish 6(2):133-138.

Sherrard, D. (2016). Linking Universities to Communities: The case of EARTH University. RUFORUM Working Document Series 14(1): 403-407.

Sherrard, D., and Alvarado, I. (2017). Entrepreneurship education in agriculture: The EARTH University approach. African Journal of Rural Development, 2 (2): 153-160.

Teferra, D. (2013). Funding Higher Education in Africa: State, Trends and Perspectives. JHEA/RESA 11 (1and2): 19–51.

Trochim, W. M. K. (2006). Research methods knowledge base. Available at:



 Table 1. Correlations between farmers’ learning behaviour and social demographics

Variable Correlates (N = 283)
1 2 3 4 5 6 7 8 9 10
1.Seeking information
2. Knowledge sharing  .73**
3. Seeking feedback  .75**  .76**
4. Giving feedback  .69**  .77**  .75**
5. Sex  .02  .01 -.06  .01
6. Farm-labour sharing  .07  .14*  .11  .12 -.01
7. Age -.04 -.05 -.06 -.05  .12  .09
8. Education  .03  .05   .04  .04 -.36**  .16** -.28**
9. Farmstead distance -.11 -.13* -.10 -.14* -.05 -.22** -.02 -.07
10. Experience of hosting students  .19**  .16**  .22**  .15*  .01  .00  .11  .04 -.12

* & ** means significant at P < 0.01 & P < 0.05 respectively