The Effectiveness of Audio Media in Enhancing Farmers’ Knowledge: The Case of Smallholder Banana Farmers in Western Uganda

Although audio media has been presented by previous research as a suitable medium for conveying agricultural information, there is limited research in assessing its effectiveness in assisting farmers’ gain agricultural knowledge. This study thus undertook to carry out that investigation by involving 1000 smallholder banana farmers in the study which involved a baseline survey, an intervention, and an end line survey. The preliminary survey results showed that 456 farmers had relevant knowledge of farming practices and were removed from the study. However, 541 farmers were lacking the required knowledge regarding proper banana farming practices. This group was assigned to an intervention of radio broadcasts for one month. All broadcasted content was recorded on CDs and circulated to the participants for continued revision. This was followed by the end line survey and the two datasets were compared to establish the change in knowledge levels of the farmers. The tests of variances before the intervention and after the intervention indicated that the two groups being compared are assumed to be approximately equal (p >0.05). Results showed a significant difference in the mean knowledge scores before and after the intervention. Furthermore, independent t-test results reveal a statistically significant difference in the preand post-knowledge tests of farmers. In addition, results from Pearson Correlations show that farmers’ characteristics such as age, gender, and level of education do not influence the uptake of knowledge on farming practices by farmers. However, results reveal a statistically significant association between the source of information (Radio Broadcast & audio CDs) and knowledge gain regarding most banana farming practices by farmers. This study provides further evidence that audio media in the form of radio broadcasts and audio CDs is a vital source for agricultural information to the rural farmers. Moreover, it can be concluded that farmers’ characteristics do not influence knowledge gain for banana farming practices when audio media is used as a medium of information dissemination. And thus, audio media remains 68 http://www.i-jes.org Paper—The Effectiveness of Audio Media in Enhancing Farmers’ Knowledge: The Case of Smallholder... a vital source of information for resource-poor farmers and can greatly enhance their agricultural knowledge when audio media is used as an intervention. Keywords—Audio media, smallholder banana farmers, farmers’ knowledge

other value-added products such as cakes, local brew, juice, chips, and other food products [6]. However, it is sad to note that few smallholder banana farmers have access to information regarding proper banana farming practices, they have no sufficient knowledge such as the use of mineral fertilizers to replace nutrients in the exhausted soils. Similar farmer challenges are pointed out in a study by [7] where it is observed that the state of affairs of the resource-poor farmers is exacerbated by natural hazards such as land erosion, drought, and floods, which can easily be controlled if farmers were equipped with the necessary protective agricultural information.
Farmers also receive reduced net earnings because of limited bargaining power for farm inputs such as fertilizers, herbicides, and other related inputs. These low rate earnings do not only reduce the capacity of farmers but affects their morale to continue engaging in agriculture. This bizarre situation is blamed on inadequate access to agricultural information and knowledge and can be turned around through the transforming of the agricultural sector by putting in place a proper medium of communication to disseminate information to the rural farmers on proper farming practices such as sustainable land and soil improvement, labour productivity and remunerative market structures. One study by [8] observes that insufficient or complete lack of agricultural information has negative consequences on agricultural production, for instance, lack of planting information can result in poor timing whereby farmers either plant early or late which of course leads to low yields and eventually meagre farm incomes.
Accurate information on weather is vital for timing major agricultural activities including but not limited to land preparation, planting and spacing of crops, harvesting, and storage. All these farm activities if not properly planned can affect crop yields, quality of produce and household incomes. Previous studies have pointed out that accessing information on proper farming practices is key for agricultural development in Africa. For example, in the study conducted by [9] observed that close to 75% of the rural-based and resource-poor farmers in Africa derive their living from agricultural production, but astonishingly they still face severe food shortages and poverty. Unfortunately, most African countries have not demonstrated willingness to provide to smallholder farmers agricultural information which could change this situation. But also, most of the farmers are semi-literate or illiterate which limits their access to printed sources of information about farming and cultivation provided by the government through the ministry of agriculture. The situation is worse, especially in rural areas where 70-80% of the African population lives [7].
One study by [8] observes that a major challenge faced by most smallholder farmers in Uganda is lack of timely agricultural information on which farmers can base to make wise economic decisions for proper allocation of scarce resources. This situation has adversely affected agricultural production. Given the existing deficiencies in agricultural information sources and channels, this study sought to use audio media as an intervention that will guarantee sustainable information flow and sharing for improved decision making and planning for smallholder farmers. However, even if audio media (radio broadcasts and CDs) is a major source of information for smallholder farmers, it remains to be established if it enhances the knowledge of farmers regarding proper banana farming practices. There is inadequate empirical work in this area. Therefore, this study attempts to investigate whether audio media increases the knowledge of smallholder banana farmers regarding proper banana farming practices. In addition, the paper investigates the extent to which socio-economic factors (Level of education, source of information gender and age) influence the uptake of knowledge regarding proper banana farming practices. There is not enough research that has assessed the influence of these socio-economic characteristics in regard to the uptake of knowledge by farmers.

2
Literature Review

Effectiveness of audio media in information dissemination among small holder banana farmers
In rural Africa, agricultural information and knowledge are vital factors that can propel economic development. Information dissemination and exchange containing new and proper farming practices can fetch new opportunities to the illiterate and rural small-scale farmers [9]. This is supported by another study that found the use of ICTs especially innovative community radio, mobile phone combined with radio, rural community telecentres and farmers' call centres to be the most affordable and reliable means of enhancing agricultural information and knowledge exchange, not only between extension agents and smallholder farmers, but among the small holder farmers themselves [9]. Furthermore, the benefits associated with radio for broadcasting of agricultural information to smallholder farmers have been increased due to the beginning of many private community radio stations in rural areas. Even if majority of these radio stations are profit-oriented, and therefore, promote programmes that are mostly sponsored by private organisations, they at the same time air out a range of agricultural related programmes.
Public radio stations are inclined to air programmes that satisfy the political agenda of the government, rather than farmers [10]. The same idea is supported by [11] who points out that radio is a medium of choice for resource-poor community people because it does not only create awareness of the best farming practices but has the capacity of increasing adoption of innovative agricultural practices for rural farmers with limited means of farm improvement. This echoes well with one study that pointed out that accessing agricultural information at a time when it is needed can enable rural farmers in Africa compete favourably with their counterparts in the developed world. Reliable and timely information can help them increase production, improve marketing which translates into increased outcomes and welfare for the farmers. On the contrary it is observed that gaps in information circulation to the rural farmers inhibit them across the value chain [8]. Furthermore, the study observes that small scale farmers with medium access to information observe how their counterparts are doing things and draw lessons from them. Their source of information is farmer-tofarmer exchange. They transfer this information and knowledge to their own farms. Unlike NGOs that offer information on service-oriented activities, radio and audio CDs are important sources of farm information and are widely used in most rural communities. Radio broadcasts for farmers provide a wide range of agricultural in-formation including new crop and disease resistant varieties, appropriate chemical use, and weather forecasts.
In the study by [8] it is found that NGOs cooperate with agricultural extension officers for extension delivery to rural farmers. Traders who deal with farm produce such as maize have links with chemical dealers for storage information. The results show that agricultural extension officers do not partner with radio stations to help them share information widely with farmers and this affects rural farmers whose reliable and affordable source of information is the radio. Previous studies have shown that even if face-to-face interaction between extension workers and rural farmers is vital, this cannot satisfy the farmers' agricultural information needs. This is because extension workers are not only few but are ill funded by most African governments. They lack essentials such as transport, training manuals, inputs and other basic training materials. They can hardly address emergency situations like those related to weather such as drought, floods or pest infestation. Consequently, radio broadcasts if well planned and timely communicated to rural peasant farmers could be very effective in addressing rural farmers' information needs and this is because radio reaches almost everyone in the community. This resonates well with one study by [12] where peasant farmers pointed out that the extension agents are not doing a commendable job in solving the issues affecting the rural farmers. Farmers indicated that the extension agents may not be showing equality in the distribution of extension services. In addition, it is noted that unlike other methods of information dissemination, the community radio as an extension tool serves the farmers better because it can take technical information from the agricultural-based research community, including local universities, research institutes, rural development centres and non-governmental organizations (NGOs) and convert this information into the native languages of the peasant farmers.
Findings of the study conducted by [10], indicate that peasant farmers solidly stated that they mostly received agricultural information from fellow farmers and radio daily compared to other sources. This was mainly due to the nature of the information, source and time the programmes were aired on radio and togetherness with each other in the community. Such information was considered immediate and spread widely. Although radio is a preferred medium of communicating agricultural information by most rural farmers, the same study found farmers and extension workers to be effective in the dissemination of agricultural information, for example a respondent in one study observed that other sources used to disseminate agricultural information are less important compared to extension agents. This is because extension agents are easily accessible in urgent and serious situations where rural farmers need information. Farmer to farmer information was also highly evaluated in the same study. However, generally about the effectiveness in information dissemination, audio media was found more effective as an information source than farmer to farmer method of disseminating agricultural information. The reason for this is perhaps the upsurge in access and use of mobile phones, CDs and in addition to many radio stations that promote the call-in services. This echoes well with the study by [13] who observed that the knowledge levels of the peasant farmers who listened to the radio broadcasts on agriculture programme significantly increased compared to the non-listeners. The study further observed that the peasant farmers who listened to the radio programmes had gained knowledge and altered their attitude toward modern farming compared to their counterparts who did not listen. Findings of this study presuppose that the radio was very instrumental not only in the dissemination of agricultural information but also the mastery of knowledge by the peasant farmers. This was also confirmed by earlier studies for example, [14] observed that nearly half of the farmers who listened to the radio and viewed TV broadcasts reported to have gained more knowledge and were able to adopt improved farming practices. This is consistent with study findings of [15] that show that majority of peasant farmers (82.6%) mentioned that they own radio sets and more than half of them 64.3% observed that radio is a vital and reliable source of agricultural information. This indicates thus, radio broadcasts will benefit many farmers and more could benefit if they only care to listen. A related study has similar sentiments where it is observed that consistent listening by small scale farmers to the radio agricultural programmes equips farmers with valuable knowledge about new farming techniques. The information received enables farmers to gradually and progressively change their methods of farming [16].
However, an assessment of the effectiveness of other agricultural information sources indicates that extension workers scored lowest in the dissemination of agricultural information. The reason for this is perhaps because farmer groups are easily accessible by fellow smallholder farmers and the latter can easily and usually at no cost access any agricultural information they need, this makes extension agents encounter just a few smallholder farmers [17]. Findings of another study by [18] shows that radio across the globe has been very instrumental in creating awareness among farmers. Information regarding modern farming practices such as selection of hybrid seeds, manure and fertilizer application, disease and pest control, this and more can conveniently be disseminated to the farmers with less costs. Farmers have different sources of information as can be illustrated in figure 1. However, for the majority poor, illiterate and rural-based farmers the only reliable and accessible source of information is the radio. Most previous studies in developed and developing countries have stressed the importance of radio in information dissemination to the rural farmers. For example, [19] observed that radio is not only vital in information dissemination but cheap and thus can be afforded by majority farmers. It has capacity to reach larger numbers of dispersed rural farmers. It has no limitations, even the remotest and hard to reach environments, the radio is there. It is a key medium for the rapid dissemination of agricultural information to the rural farmers. This echoes well with other previous studies that found radio to be very key in disseminating information to farmers. It was found that two cohorts of farmers differed significantly; those who listened to the radio agricultural programmes and those who never listened at all but relied on other sources of information. The farmers who listened exhibited better farming practices than their counterparts who did not [12], [13]. Consequently, this study sought to investigate the effectiveness of radio in equipping farmers with knowledge on proper banana farming practices. Of all the information sources the radio supplemented by recorded information on CDs was preferred because in the Ugandan context, although there are a multitude of FM radios they are limited studies that have investigated how useful these radios stations can be in equipping rural farm-ers with the necessary knowledge and skills. Although, previous research has shown that of all the agricultural information sources, radio is the most suitable medium for the rural farmers, this has not been fully investigated in the Ugandan context with a convergence of ICTs. A summary of some of the sources of information that farmers use is indicated in figure1.

2.2
The influence of socio-economic factors on the uptake of knowledge regarding proper farming practices through audio media.
In this study we also sought to establish the influence of socio-economic factors such as level of education, gender, age and source of information on the uptake of knowledge regarding proper farming practices. Socio-economic factors can either negatively or positively influence farmers' knowledge which in turn can affect productivity. According to [10], the use of ICTs in most developing countries for the provision of relevant agricultural information to rural farmers largely depends on the socio-economic background of each individual farmer in the agricultural sector.
In a related research, it is observed that access to agricultural information and knowledge are key socio-economic factors that foster agricultural development in the world. Agricultural information is crucial in enhancing agricultural production of smallholder farmers. The information is obtained from various sources, depending on the type and uses to which the information is intended for. Agricultural information is normally generated from different digital sources and disseminated to smallholder farmers through different channels of communication. However, for the rural and remote areas of Africa, radio is the most affordable communication channel for the rural population. Farmers' reliance on agricultural extension, as the major sources of information and knowledge, has been the greatest hindrance to agricultural production in most rural areas in Africa. It has been found that some of the sources of information to farmers are unclear and counterproductive [10].
Today more than ever before, access to information and acquisition of knowledge are not only key drivers of socio-economic transformation but also the most important factors of development across the globe. Rural farmers have limited access to agricultural information and this affects agricultural productivity. Agricultural information is usually obtained from different sources and circulated to rural farmers through different channels, but out of these, radio has so far proven to be the most affordable and accessible to rural farmers [10]. This view is also supported by [8] who argues that with an ever-increasing population and land fragmentation, the perpetual decline in soil fertility, environmental degradation, increased crop pests and diseases, smallholder farmers need knowledge and information that can help them improve the quality of their production. Therefore, providing appropriate information on proper farming practices can greatly enhance their living standards. In the study by [14], no significant relationship was found between demographic characteristics such as age, land holding, extension services, social participation, and mass media participation. This finding is supported by the study by [20] who observed that there was no significant correlation between education level and knowledge gain by peasant farmers. However, there was a slight difference in knowledge acquisition, knowledge gain was slightly higher among the literate farmers when compared to illiterate ones. The contribution of both women and men to agricultural production is significantly high and yet their access to agricultural information and resources differs. For example, despite their contribution to agricultural production, many rural women do not have as much privilege to access information as men. These gender disparities majorly caused by the socially constructed relationship between women and men negatively affects not only access to agricultural resources but even access to information which key to agricultural production [21]. This gender gap in agricultural production is also confirmed by [22] who observes that in some of the developing countries, there is limited contact between men and women and in cases where we have more men as extension workers than women, the latter are less likely to access agricultural extension services with a presumption that agricultural information received by men will dribble down to women and other household members. However, on the contrary findings of one study by [23] did not find this to be the case.

Theoretical Framework for the Study
This research uses the Diffusion of Innovations Theory authored and propagated by [24] in his Diffusion of Innovations and expounded on in Communication of Innovations: A Cross Cultural Approach by [25]. This work explains the process by which new ideas, concepts, thoughts and innovations move through channels among people of a given social system. For elaboration, check the SMCRE Model (Source-Message-Channel-Receiver-Effect) in Figure 2. In this model (S) stands for Source and this could be some agricultural agency or any NGO poised to disseminate agricultural information to the farmers, (M) stands for message, and this could be any advice regarding some agricultural information to the farmers such as site selection, planting and spacing, fertilizer application or disease and pest control, The message reaches the farmers through a conduit or channel, in this model abbreviated as (C). This channel could be a radio, fellow farmers, farmer groups, government or extension agents. In this study the channel was MMU community radio "Amagezi Murro-Knowledge is power" 105.2 FM. In the model (R) stands for receiver and in this study, these were the small holder banana farmers organised under radio listening clubs and (E) indicates the effect. This means what farmers do with the agricultural information. For example, do farmers change their long held traditional farming practices to proper and modern farming practices? In this study, the knowledge levels of the farmers were assessed before the intervention of radio broadcasts. We did broadcast agricultural information for one month to the farmers who were found lacking knowledge regarding proper banana farming practices. The (E) is what happens after farmers have received the agricultural information. In this model, it is not only enough to know the source of information through which farmers get information but is imperative to know whether the source of information is effective or not. This knowledge is important because it would assist policy makers in making decisions whether to uphold or change the channel of communication. However, it should be noted that this study did not focus on other scales of the model, it focused on the effect of agricultural information (i.e. What farmers do with the agricultural information they receive). These other aspects will either be handled by other researchers or by us in further studies. The main interest of this study has been to focus on the 'E' or the effect of agricultural information on smallholder banana farmers' farming practices.

Methods
This study used a mixed design approach to assess the effectiveness of audio media in enhancing smallholder banana farmers' knowledge regarding proper banana farming practices. The underlined research objectives were investigated in this study: 1. To assess the effectiveness of audio media in enhancing smallholder banana farmers' knowledge regarding proper banana farming practices. 2. To evaluate the extent to which socio-economic characteristics (age, gender, level of education and source of information) influence the uptake of knowledge regarding proper banana farming practices.

Study design and procedure
A total of 1000 participants was included in this study. Multi-stage and convenient sampling techniques were used. In the first stage, 4 districts out of 7 districts in the Rwenzori region were selected. The second and third stages involved the selection of sub-counties and zones respectively. The fourth and final stage involved the selection and generation of lists of households from each zone. In each zone, 20 participants were sampled from each of the 50 selected zones in the Rwenzori region. Zones selected were only those where MMU radio signals were very clear. Information regarding radio coverage and the specific zones where MMU community radio signals were very clear was obtained from the radio station at MMU. Before selecting participants for this study in June 2017, we made pre-survey visits to each of the selected zones in a district. The visits were intended to achieve the following: 1. To confirm that the radio signals are clear in that zone, 2. To request the Chairperson Local Council 1 (C/P LC1) to randomly select and make a list with phone contacts of 20 smallholder banana farmers from his/her village and 3. To request the C/P LC1 to accept to work with the principal investigator in the mobilization of the identified participants. Two months later, in August 2017, we conducted a needs analysis study involving all the identified households with the intention of establishing their knowledge levels regarding proper banana farming practices. Data from this survey was analysed and two cohorts were identified. One cohort consisted of farmers who were found doing well regarding proper banana farming practices and the other cohort contained farmers who were found lacking almost in all aspects of the farming practices. The focus of this research was on the latter cohort. This group was organised into radio listening clubs (RLCs) and a well-structured leadership was put in place consisting the chairperson, secretary, and timekeeper. Researchers have pointed out that radio listening groups or forums are on the increase as listenership models. One study by [26] defined a radio forum or listening club as a "small listening and discussion group that regularly meets according to the broadcast schedule to listen and discuss the content of a given special radio programme". This was followed by identifying and hiring two agricultural specialists from the School of Agriculture and Environmental Sciences who were tasked to develop and broadcast content to the cohort that was found lacking knowledge regarding proper banana farming practices. Each RLC was given a small radio set with batteries that were estimated to last for a month. They were also given a broadcast schedule and asked to follow the schedule and meet at an identified meeting place to listen to the radio broadcasts. The radio programmes lasted for a month and they were very interactive whereby the members from different RLCs were given a chance to call in and give their opinions on the programme or ask questions. To supplement the radio broadcasts, we developed CDs containing all the con-tent on proper banana farming practices and distributed them to all the RLCs in the 50 zones. This audio information was intended to assist the farmers either fully grasp the content or comprehend some aspects of the content that were missed or not understood during the radio broadcasts. After a period of 6 months, a period hoped to be long enough for the smallholder banana farmers to have put in practice what they were taught, we conducted an end line survey. As can be seen in the results section, baseline survey data was compared with end line survey data to establish if there are any significant changes in knowledge before and after an intervention of radio broadcasts on proper banana farming practices.

Participants
Participants in this study were 1000 smallholder banana farmers drawn from 50 villages in Kabarole, Kyenjojo, Bunyangabu and Kamwenge districts in Western Region in Uganda. Out of 1000 respondents, 456 respondents were found with the necessary knowledge regarding proper banana farming practices and these were left out at the baseline survey level. The target group of this study was the 544 respondents who were found lacking as regards the recommended farming practices. Out of 544 respondents, 2 died before the end line survey and 2 declined to participate in the study. Therefore, 540 respondents filled the questionnaire completely and participated in observations. Out of these respondents, 420 (77.7%) were male while 120 (22.2%) were female. These results indicate that the majority of smallholder farmers are males. This supports the findings of [15]; [17] who in their research found that there were more male farmers than females. Respondents were also asked to indicate their age and most respondents 185 (34.2%) were aged 50 years and above, between 30 to 39 were also slightly many 152 (28.1%), between 40 to 49 were 134 (24.8%) and the least in number 69 (12.7%) were aged between 20 to 29 years. Participants were also required to indicate their level of education and most of the 317 (58.6%) indicated to have completed Primary level, while those who indicated to have completed secondary were 130 (24.0%), a small number 31 (5.7%) indicated to have finished tertiary and 62 (11.5%) indicated to have not attained any level of education.

Instruments and measurements
We used a questionnaire and observation checklists in the collection of quantitative and qualitative data respectively. Some of the items used in the questionnaire were drawn from validated studies such as [29]; [30]. The rest of the items were selfdeveloped. In addition, the instrument was checked to verify proper wording, completeness, and appropriateness for content validity by professors at the Faculty of Agriculture and Environmental Sciences. Finally, the instrument was checked and approved by the project supervisors. Training of research assistants and translation of the instrument into Rutooro language was done ahead of data collection by a Rutooronative professor with bias in agricultural extension and instructional strategy. The instrument was administered by the first author assisted by 9 research assistants. The instrument includes the following scales of measurement: site selection, land prepara-tion, planting and spacing, fertilizer, manure and pesticide use, weed management, soil and water management, pruning and de-suckering, pests and disease management, staking, harvesting, storage, and marketing. Qualitative data was collected through the use of structured observation. Research assistants were asked to reach the 110 randomly selected households and after filling the questionnaire to request the household head to observe the plantations, while at the plantation as can be seen on the results section, they were to indicate on the observation checklist whether they have observed recommended or not recommended practices. To analyze the results, a two-step analytic process was employed. We first arranged responses per each category or theme based on the farming practices. Second identifying the key observed items per each farming practice based on the frequency of occurrence so as to determine the percentages of observed recommended or observed not recommended farming practices. A summary of results from the observation checklist is shown in the results section.

Data analysis
This study collected both quantitative and qualitative data. For the former, descriptive statistics were used to analyse demographic data of the respondents such as age, gender, level of education, electronic gadgets owned, etc. Second, the assumption of homogeneity of variances was tested and satisfied via Levene's F test with p > 0.05. This was done as a basis of comparison for the two groups before and after an intervention. The comparison in terms of means and standard deviation to assess the change in means before and after the intervention was performed and results showed a change after the intervention of audio media. We ran independent t-test to assess if they are significant changes between the two groups after the intervention of audio media. We also used Person correlation to establish the association between farmers' demographic characteristics and knowledge gain regarding proper banana farming practices. The quantitative findings of this study are substantiated by the observations made by enumerators during the end line survey. A summary of observations is made from various respondents with whom we interacted during our study. Given in the results section are some extracts from the observations made.

Key Research Findings
In this section, we present our study results starting with quantitative findings which are followed by a summary of extracts from the observation checklists. Observation results come from the randomly sampled households from the 50 villages where the research was conducted.

Effectiveness of audio media in enhancing smallholder banana farmers' knowledge regarding proper banana farming practices.
Results regarding the effectiveness of audio media in enhancing knowledge of banana farmers on proper farming practices indicate that there was a change in the knowledge levels of smallholder banana farmers after an intervention of live radio broadcasts supplemented by audio information on CDs. The mean scores of the 50 out of 61 items in table 3 increase after the intervention was done via Radio broadcast and audio CDs. Forty-five items through all practices have shown a big change after the intervention with an increase of more than 0.5 scores, showing significance in t-tests as described in the next paragraphs. Examples of such items are the method used to clear the land, time of land preparation, size of holes dug, number of suckers per plant, time of de-suckering, etc. Before running the t-tests the assumption of homogeneities of variances was tested and satisfied via Levene's F test. This is consistent with studies done by [27]; [28] where they observe that homogeneity of variance and normality are vital and should precede tests like t-tests and ANOVA. Homogeneity of variance results indicate F (6) After, the researcher performed the t-test to assess if there are significant changes after the Radio broadcast. It was found out that the independent samples t-tests were associated with a statistically significant effect before and after the radio intervention and specifically in the following practices:

Findings from observation
To substantiate findings of the quantitative data "see table 2", this study also made use of observation checklists, where we examined recommended farming practices based on the works of [29]; [30]. Only practices that could easily be observed from the banana plantations were included. On the overall, findings show that a big number of respondents to have adopted the proper banana farming practices as can be seen in table 2. However, in some cases flaws were observed for example (40%) of the farmers were found not to have implemented proper de-suckering as many suckers were seen per plant, also almost half (45%) of the observed smallholder banana farmers had a variety of different types of bananas in the same plots of land. This is bad because mixing many varieties in the same garden, encourages pests and diseases. Regarding staking no enough windbreaks were seen in and around the plantation. Farmers were not doing well either as some banana plantations looked to be infested with diseases such as sigatoka as could be seen from ripening young bananas and bananas with yellow leaves

Pests and disease control
Banana plantation appeared to be free from weevils and nematodes 63(69.3%) Plantations looked to be infested with diseases such as sigatoka as could be seen from ripening young bananas and bananas with yellow leaves 47(51.7%)

The extent to which socio-economic characteristics influence the uptake of knowledge regarding proper banana farming practices
The second objective sought to investigate the extent to which socio-economic characteristics such as level of education, gender, age and source of information in-fluence the gain in knowledge regarding proper banana farming practices by smallholder banana farmers that participated in this study. To achieve this aim, we ran Pearson correlation "see table 3". The research wanted to find out to what extent does age contribute to the best farming practices however, results from Pearson Correlation tests between age and farming practices reveal a negligible association (Pearson correlation r lie between 0.01 and 0.09 or below 0). Furthermore, results also indicate that there is a low association between soil and water management with the age (r =.139, p=.001).
Similarly, regarding respondents' gender, the Pearson correlation have shown that the gender and farming practices have a negligible association with gender (r =.029, p=.507). These results indicate that the effect if any is small and insignificant. Additionally, the Pearson correlation showed that the level of education and harvesting & storage management have a low association [ r (541) =.144, p=.053], and on the other side the relationship with the level of Education and other farming practices are not related at all. The researchers affirm that this was caused by the fact that all farmers were subjected to follow the radio broadcast practices and the practices were easy to implement without further technicalities either in terms of content or context. There is also a belief that the farmers share their experiences as well.
However, the Pearson correlation showed that there is substantial association between the Radio information broadcast & CDs and knowledge about site selection r(541) = 0.615 , p> 0.05 , pruning & de-suckering r (541) = 0.664 , p> 0.05 . It was observed at the same time there is a strong relationship between the radio broadcast and land preparation and also staking management with r (541) = 0.745, p> 0.05, r (541) = 0.867, p < 0.05. On the other hand, radio information sharing has demonstrated a moderate relationship with other farming practices" see table 3".

Discussion
This study investigated the potential of audio media in enhancing farmers' knowledge regarding proper banana practices. The study was guided by two objectives; to assess the effectiveness of audio media in enhancing smallholder banana farmers' knowledge regarding proper banana farming practices and to evaluate the extent to which socio-economic factors influence farmers' uptake of knowledge regarding proper banana farming practices. In the context of Uganda with abundant FM radio stations and limited agricultural extension services to smallholder farmers, it is imperative to establish if audio media (radio broadcasts and audio CDs) could be helpful in equipping farmers with knowledge regarding the proper farming practices.

The effectiveness of audio media in enhancing smallholder banana farmers' knowledge regarding proper banana farming practices
Before the intervention of audio media (radio broadcasts and audio CDs) smallholder farmers had poor initial knowledge regarding proper banana farming practices on most banana farming practices especially the time of land preparation, tools used for de-suckering, planting, method of weeding, etc. However, based on our findings there was evidence of improvement in the knowledge levels of majority smallholder banana farmers on almost all farming practices after making an intervention. Our findings are consistent with the study by [31] where the effectiveness of community radio programme was investigated through knowledge test whereby knowledge levels of rural women were assessed before and after they listened to a community radio programme. Just like our study, this study found significant changes in the pre-test and post-test scores of overall knowledge of the small-scale farmers who participated in the study. Nevertheless, quantitative findings revealed that some farming practices, such site selection, soil and water management, pruning & de-suckering, and pests & diseases management did not show a statistically significant change before and after the intervention. This is perhaps due to the nature of the practices themselves, as they cannot easily be changed and also some of them may be caused by external factors on which the farmers do not have any control.
To substantiate quantitative findings, this study also used observation method and results reveal increase in knowledge levels after the intervention as indicated by improved farming practices for example 79% of the farmers were able to plant as a recommended practice perennial crops such as legumes in their banana plantations and 75% of the farmers had banana plants planted in blocks as recommended. The same improvement is noticed in almost all other banana farming practices. This indicates that the radio broadcasts supplemented by audio CDs were of great benefit to the farmers who before the intervention showed low levels of knowledge. In this study, when asked about their source of agricultural information, more than half of the farmers 51.3% mentioned the radio, while 13% mentioned extension workers, farmer groups 27.4%, TV and newspapers 2.2% and 6.6% respectively. This indicates that radio can be very instrumental in extending information to the rural farmers despite the fact that extension agents are taken to be the official medium for agricultural information to the farmers. This resonates well with a study by [10], who found that peasant farmers mostly received agricultural information from fellow farmers and radio daily compared to other sources. Our study findings are congruent with [17] whose study on the assessment of the effectiveness of other agricultural information sources found that extension workers scored lowest in the dissemination of agricultural information. The reason for this is perhaps because farmer groups can easily be reached by fellow farmers and normally at no cost. Although in their study [32] observed that traditionally, the primary sources of access to innovative farming information for both rural and resource-poor farmers are agricultural extension agents, our study provides additional evidence that radio is key in enhancing knowledge levels of rural farmers.
However, our findings indicate that even if on the overall smallholder farmers demonstrated to have gained knowledge regarding proper banana farming practices, some farmers were found to be lacking in some aspects of agriculture, for example, 45% of the farmers contrary to what is recommended many varieties were observed in the same garden or plots. Mixing many varieties in the same garden encourages pests and diseases. Based on our observation this could be prompted by the small pieces of land where people are living as most land is highly fragmented due to increased population. More still 46.2% of the farmers never applied organic fertilizer in their plantations, this was mostly observed in the villages of Kabarole and Bunyangabu districts and the reason for this could be perhaps the soils are fertile and self-sustaining and thus no need for manure or fertilizer application. One absurd finding is that even if farmers have tried to follow the recommended farming practices, in some areas diseases such as Sigatoka and banana bacterial weevil (BBW) have persisted as could be seen from the ripening of young bananas and bananas with yellow leaves. More than half of the respondents 51.7% were affected. This situation could be explained by the lack of agricultural information or advice to rural farmers. This calls for urgent government intervention otherwise farmers' plantations will be wiped out, leaving them without any means of survival. However, this notwithstanding, overall findings showed audio media (radio broadcasts and audio CDs) to be effective in terms of knowledge gain by the farmers and this is not only showed by increased mean scores after implementation the intervention but even from the observation results. Likewise, one study found radio programmes related to agriculture to be vital in equipping farmers with farming methods, so as the farmers listen to the programmes they steadily change their old-fashioned farming methods to not only new but recommended methods [33].

6.2
The extent to which socio-economic factors influence the uptake of knowledge regarding proper banana farming practices The second objective of this study sought to evaluate the extent to which socioeconomic characteristics such as age, gender, level of education and source of infor-mation influence the uptake of knowledge regarding proper banana farming practices. Studying socio-economic characteristics of farmers is vital if they are to be helped improve farming, for example, one study found that the use of ICTs in most developing countries for the provision of relevant agricultural information to rural farmers largely depends on the socio-economic background of each individual farmer in the agricultural sector [10]. Nevertheless, this study found no significant relationship between age and all other banana farming practices. However, there was a low association between soil and water management.
Additionally, regarding respondents' gender, the Pearson correlation revealed that gender and farming practices have a negligible association. Moreover, the Pearson correlation also showed no statistically significant relationship between the level of education and farming practices but there was a low association between level of education and harvesting & storage management. Our findings resonate well with the findings by [14] who found no significant relationship between demographic characteristics such as age, land holding, extension services, social participation, and mass media participation. Furthermore, the findings of this study are supported by a study of [20] who observed that there was no significant correlation between education level and knowledge gain by peasant farmers. Nonetheless, this study contrary to ours found a slight difference in knowledge acquisition, knowledge gain was slightly higher among the literate farmers when compared to illiterate ones. The researchers affirm that this could have been caused by the fact that all smallholder farmers that participated in this study were subjected to participate in the radio broadcasts and revise the content taught by listening to the audio CDs. The information was easy to implement without further technicalities either in terms of content or context. In addition, the radio broadcasts were delivered in the local language of the farmers which they understand very well. The radio programme presenters were natives of the region and endeavoured to do anything possible to simplify the content. There is also a belief that the farmers share amongst themselves their experiences as well. All these considerations could have been enabling factors when it came to content mastering and implementation.
However, study findings showed a positive correlation between the source of information and knowledge gain by smallholder banana farmers regarding site selection and pruning & de-suckering. More so, results indicated a strong relationship between the Radio broadcast and land preparation and staking management. On the other hand, audio media as a source of information sharing has demonstrated a moderate relationship with other farming practice.

Conclusions, Implication & limitations
The first outcome of this study is that we found audio media in the form of radio broadcasts and audio CDs to be the main source of agricultural information for the category of farmers included in this study. Audio media was also found to have played a significant role in assisting farmers gain knowledge on proper banana farming practices. Although smallholder farmers had low knowledge about most banana farming practices before the intervention, there was a significant change in knowledge levels after implementing the interposition of audio media. This change therefore is attributed to this intervention. The second outcome of this research is that socioeconomic characteristics mainly age, gender and level of education do not have significant influence on knowledge gain regarding proper banana farming practices by smallholder farmers. However, source of information (radio broadcast and audio CDs) was observed to positively influence knowledge gain of farmers. This could be because of the advantages associated with the medium such as easy accessibility, information aired in the local language, flexibility as one can listen as they carry out other chores, being economical etc.
However, what this study did not do, which could be undertaken in future research, was to assess the effectiveness of audio media compared to other sources of agricultural information such extension workers, farmer groups, farmers' experience etc. Nevertheless, the findings from this research can contribute in numerous ways because no study of this kind has been conducted in the Ugandan agricultural sector to evaluate the influence of audio media (radio broadcast & audio CDs) on knowledge uptake regarding proper banana farming practices by smallholder farmers. This study will not only act as a guide but also a point of reference for agricultural managers at different levels of government and private sector in selecting the kinds of media technologies to use in disseminating agricultural information to rural farmers. A clear understanding of how effective audio media (Radio broadcasts and CDs) is regarding enhancing farmers knowledge on proper banana farming practices is vital for policy makers especially in the Ministry of Agriculture, yet this aspect has not been sufficiently addressed by previous research. Therefore, findings from this research if taken seriously will be a great relief to the already strained agricultural sector in terms of extension services. Radio could be subsumed as a suitable alternative means of agricultural extension to the resource-poor farmers in the region.