Livestock Research for Rural Development 30 (11) 2018 Guide for preparation of papers LRRD Newsletter

Citation of this paper

Effectiveness of using cellphone technology as a dairy management training tool for smallholder dairy farms in Kenya

D N Makau, J A VanLeeuwen, G K Gitau1, J Muraya, S L McKenna, C Walton2 and J J Wichtel3

Department of Health Management, Atlantic Veterinary College, 550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3
dmakau@upei.ca
1 Department of Clinical Studies, Faculty of Veterinary Medicine, University of Nairobi, P O Box 29053-00625, Nairobi, Kenya
2 Department of Applied Human Sciences, University of Prince Edward Island, 550 University Avenue, Charlottetown, Prince Edward Island, Canada C1A 4P3
3Ontario Veterinary College, University of Guelph, 50 Stone Road E., Guelph, Ontario, Canada, N1G 2W1

Abstract

There is increasing need for knowledge on the utility of information and communication technology (ICT) for improved agricultural productivity and enhanced income in smallholder production enterprises. The objective of this study was to determine the effectiveness of using cellphone technology as a training tool on smallholder dairy farms (SDFs) in Kenya.

This field trial was carried out between June and September 2017 on 40 farms randomly selected from members of the Naari Dairy Farmers Cooperative Society in Naari sub-location of Meru County, Kenya. An abridged dairy management handbook, developed by Farmers Helping Farmers and the University of Prince Edward Island, was translated into the local dialect, and disseminated as short message text. After pre-intervention knowledge and attitudes assessments on dairy management, farms were randomly allocated into intervention and comparison groups. Using an online short message service interface (because the study population all had cell phones but only 1.7% had smart phones), short messages on management practices were sent daily, for 3 months, to the phones owned by the farmers in the intervention group. Post-intervention assessment of dairy management knowledge and attitudes related to the messages was done 3 weeks post-intervention. Within and between group comparisons and net changes were determined using t-tests, Chi-squared tests where applicable.

There were no significant demographic or knowledge differences between the two groups pre-intervention. Compared to pre-intervention, trained farmers in the intervention group were significantly more informed on: mastitis prevention, disease (calf diarrhea) prevention, stall management, the role of a balanced nutritious diet on immunity and the resolution of some health conditions post-intervention. Translation of message content to the local language and using easily understandable terminology were reported to be helpful for better understanding and motivation of farmers to implement recommendations.

Cellphone technology with a short message service interface can be an effective training tool for SDFs in remote areas of Kenya located far from where seminars are conducted for dairy farmers.

Keywords: developing country, economic, education, information communication technology, livelihoods, rural farmers


Introduction

Like other developing countries, optimal production of the smallholder dairy industry in Kenya is constrained by various challenges including: inadequate feed quality and quantity, poor storage facilities for feed conservation, high cost of feed inputs and inadequate information on production approaches and technologies (Lukuyu et al 2011). Poor communication of research findings to farmers has been identified as a major stumbling block to uptake of best management practices and existing technologies for better cattle nutrition and production on smallholder dairy farms (SDFs) (Ngwira, 2003; Mwangi & Wambugu, 2003; Hove et al 2003; Franzel et al 2014; World Bank Group, 2017). Participatory education and training of farmers could enhance adoption of improved fodder crop use and establishment and efficient use of pastures in SDFs in Kenya (Mwangi & Wambugu, 2003; Lukuyu et al 2011).

Cellphones have been used in different parts of Africa by farmers and fishermen to support their businesses, with numerous benefits and challenges alike. In Ghana, cocoa farmers were able to save on various transaction costs, such as transportation and operational costs (arranging for inputs and contacting purchasing clerks), through the use of cellphones (Ofosu-asare, 2011). A study on the fishing industry (Ghana) observed that fishermen who had cellphones were able to expand their markets using cellphone communication with clients. In addition, the fishermen were able to make decisions based on current information received through their cellphones (Salia et al 2011). Other benefits highlighted by farmers in northern Ghana included improved communication with farm input suppliers, resulting in increased efficiency in farming (Alhassan & Kwakwa, 2012). However, there has been limited research on the use of cellphones for agricultural education purposes in Africa.

A study to assess the use of cellphones for dissemination of agricultural information in India concluded that farmers mostly used their phones for meeting social needs, and receiving extension messages was incidental (Sahota & Kameswari, 2014). However, a more recent study, in the same area of India, concluded that farmers had used cellphones for communication with universities and veterinary institutions on animal husbandry for more than 3 years prior to the 2014 study (Rathod et al 2016). The authors recommended that adequate measures be undertaken to promote adoption of cellphone technology for effective dissemination and use of livestock-related information (Rathod et al 2016). There has been further innovation in agriculture to increase the impact of human communication and social connections on agricultural productivity and smallholder incomes. These social connections have been achieved through specialized applications that act as conduits of information dissemination in the United Kingdom (The World Bank, 2012). These innovations are needed in Africa as well.

Over the last decade, cellphone technology has become largely accessible in even the remotest parts of Kenya (Karlsen et al 2010). Like the rest of Africa, cellphones in Kenya are used for exchange and dissemination of information such as: disease monitoring, weather monitoring, advertising, marketing, financial transactions, business promotion, credit facility, access to advice, and much more (The World Bank, 2012).

A study done on SDFs in Nakuru county, Kenya, documented significant positive association between increased milk yields and use of cellphones for provision of extension services (Smollo et al 2016). However, although use of cellphones has a huge potential for improving smallholder productivity, various factors influence the gains. These factors include: timeliness, quality and trustworthiness of the information disseminated, type of agricultural practices, skills and knowledge levels of the farmers, institutional policies and regulations (Mittal & Tripathi, 2009 ; Mutunga & Waema, 2016). As a consequence of these factors, under-utilization of animal husbandry information via cellphones has affected milk production in SDFs in Kenya (Smollo et al 2016). There is increasing need for knowledge on the utility of information and communication technology (ICT) for enhanced agricultural productivity, and subsequently improved income in smallholder production enterprises (The World Bank, 2012). However, research on the effectiveness and use of cellphones, as one method of ICT, in training farmers or disseminating extension-related information in the East African region, especially Kenya, is minimal.

The objective of this study was to determine the effectiveness of using cellphone technology as a dairy management training tool on knowledge and attitudes of smallholder dairy farmers in rural parts of Kenya.


Materials and methods

Description of study area

This randomized controlled field trial was carried out in Naari sub-location of Meru County, Kenya (06'0" N and 3735'0" E). Meru County is located on the eastern slopes of Mount Kenya and is 270 kilometers north of Nairobi, the capital city of Kenya. Naari sub-location is in the high agricultural potential region with an altitude of approximately 2,000 m above sea level. The main agricultural activities include: dairying, subsistence crop farming, horticulture and lumbering. Farmers grow food crops such as maize, beans and Irish potatoes and forages for dairy cows. This study area was predetermined since this trial was part of a larger study involving dairy farmers in the area (Makau et al 2018; Muraya et al 2018). A non-governmental organization, Farmers Helping Farmers (FHF), and the University of Prince Edward Island (UPEI) had an existing developmental partnership with the Naari Dairy Farmer Cooperative Society (NDFCS). This rapport provided a strong foundation for the work and the entry point to the community.

Sample size and data collection

The farmers included in the study were from NDFCS, a dairy group with an active membership of 550 farmers (active member is defined as one who regularly sold milk to the NDFCS at the time of the trial). In May 2015, 200 SDFs were randomly selected from the NDFCS registry for a related cross-sectional study using software-based random number generation. One hundred of the 200 SDFs were involved in another related intervention study, and therefore were not eligible for this trial to preserve the integrity of the intervention study. Of the remaining 100 SDFs, participants were selected if they met the eligibility criteria of: active membership with the NDFCS, possession of a cellphone, and subscription to the Safaricom carrier as the cellphone service provider. A total of 95 of the 100 SDFs met the inclusion criteria. Sixty farms were selected for this study through random number generation. Phone interviews were conducted to confirm compliance with the criteria and interest in participating in the study. When a farmer declined to participate in the study, the farm corresponding to the next random number was invited to participate as a replacement. The sample size was determined based on a need to demonstrate differences in knowledge levels between two groups of 30 farmers with respect to the cellphone training intervention, 95% confidence and 80% power.

The 60 farmers were randomly allocated into either a comparison (30) or intervention (30) group. The principal farmers for each group were invited to attend an initial meeting for their group orientation. The meeting for the intervention group was held one day before that of the comparison group. Both meetings were followed by administration of a questionnaire for collection of baseline data (pre-intervention) on knowledge and attitudes on dairy management. The questionnaire was self-administered but facilitated by a local farmer who served as a translator from English to ‘Kimeru’ (local language) where necessary.

Some sections of the questionnaire were borrowed from a questionnaire used in the 2015 study. The questionnaire had 37 questions with sections on farm household demographics and principal farmer’s knowledge and attitudes related to: mastitis prevention and management, teat blockages, nutritional management, stall design, and neonatal calf management practices. After these two initial meetings, held on two consecutive days in June 2017, the 60 selected farmers subsequently began participating in the study (Figure 1).

Figure 1. Flow diagram of participants for a cellphone training intervention trial on dairy management in Kenya in 2017.

The farmers in the intervention group were registered in a database management system using MySQL (Structured Query Language) and content dissemination was managed through an Apache platform. Only 1.7% of the farmers in the study population owned smart phones.

Intervention

An abridged version of a dairy management handbook developed by FHF and UPEI was used to develop the content used for training the intervention group. The abridged handbook was translated into the local language (Kimeru) and compressed into short text messages of 160-200 characters. Using a XAMP server and an online integrated SMS interface, ‘Africa’s Talking’ provided by Safaricom, the short text messages were sent daily to the cellphones owned by the farmers in the intervention group. One message was sent per day, 5 days a week for 3 months between June and September 2017.

Post-intervention data were collected during a follow-up meeting 3 weeks after completion of the intervention. These meetings were held separately for each of the groups (intervention and comparison), at different times on the same day. A local farmer (translator) facilitated the filling of a self-administered questionnaire and subsequent focus group discussions (FGD) for both groups. During these meetings data on knowledge and attitudes of the farmers on dairy management were collected. The FGD for the intervention group was aimed at assessing the overall experience and impact of the cellphone intervention and clarify any issues emanating from the training messages. The FGD for the comparison group served as an avenue to address some challenges the farmers faced on their farms related to feeding and mastitis. The themes for discussion were centered around nutrition and mastitis management questions in the questionnaire. At these meetings, participants in both groups received one-liter of cooking oil and one kilogram of dairy cow mineral supplements as appreciation for their participation. All farmers in the comparison group were subsequently provided with detailed education seminars to address some of the farm management challenges they faced.

Data management and descriptive analysis

Data from the questionnaires were keyed into MS Excel 2010 (Microsoft, Sacramento, California, USA) and checked for errors. Data were then transferred to STATA software 13.0 (StataCorp LLC, College station, Texas, USA) for statistical analysis. Descriptive statistical analysis (summarizing distributions, means, and medians) was done for continuous variables. Categorical variables were also summarized using frequencies and percentages.

Knowledge scores were calculated based on responses provided to groups of questions on feeding (3) and mastitis prevention (7). Each right answer given was allocated a value of 1 while each wrong answer was 0. Responses to all questions within a group were then summed up to provide a score for each individual respondent for that group of questions. There were no missing responses to these questions.

For continuous variables (e.g. size of land used for dairy production and knowledge scores), pairwise comparisons were carried out using two sample t-tests for between-group comparisons, and one-sample paired t-tests for before and after comparisons of the same group. For categorical variables, Pearson’s Chi-square and Fisher’s exact tests (if cells had fewer than 5 farmers) were used. The net change was calculated by comparing differences in scores on questions pre- and post-intervention within and between groups. For proportions, confidence intervals were used to identify significant differences between groups and within groups (Barr, 2018). Results were considered significant if p value ≤ 0.05 or confidence interval were not overlapping. Farmers agreed to the use of the data for research purposes as long as confidentiality was maintained.


Results

A total of 40 farmers participated up to the completion of the study, 20 farmers withdrew from the study (Figure 1). Their reasons for withdrawing included: ineligibility because they were no longer selling their milk to NDFCS; getting a job off the farm (making the training irrelevant and not being available for post-intervention assessment); and having a change of farming priorities (resulting in sale of animals, hence no motivation to continue to participate in the project). These reasons were not perceived to be related to the study or its objectives and therefore selection bias was expected to be minimal.

Demographics of and farm characteristics of participating SDFs.

Out of the 40 farmers who fully participated in the study, most were male, with no significant difference in gender between the intervention and comparison groups (p = 0.34) (Table 1). Most of the women (78.6%) had only studied up to primary level education, while most of the men (61.5%) had studied up to secondary school level. The difference in education levels between the two genders in the study population was statistically significant (p = 0.02). However, there was no statistically significant difference between the education levels of the principal farmers between the two study groups (Table 1).

More than two-thirds of farmers reported that a substantial (50-75%) proportion of their total household income was earned through dairy farming (Table 1). On average, farmers had about 3.4 acres (s.d.= 2.4 acres) of land available for dairy and crop production. Most farmers (55.0%) allocated at least 50.0% of their available land to dairy production (Table 1).

Pre-intervention knowledge analysis and comparison between groups

Farmers were keen to increase their knowledge in dairy farming, with 62.5% of them having attended some form of training on dairy farming. The proportion of farmers that had attended some training (through seminars and or educational/experiential trips) on dairy production in the last one year prior to the field trial was not significantly different between the two groups (p > 0.05) (Table 1). However, slightly more farmers in the comparison group reported having attended training than the intervention group. A high proportion of principal farmers in both groups were not able to recall the subject of training sessions they had attended. Although not significantly different between groups, the proportion that could not remember was modestly higher among intervention group members (Table 1).

General knowledge on mastitis prevention was fairly good in the study population pre-intervention. Washing the udder prior to milking was a commonly known practice (82.5% - 33/40), but only a handful knew about using some cleaning agent in the wash water (15.0% - 6/40). Few farmers knew about post-milking teat dip (25.0% - 10/40) and dry cow therapy (30.0% - 12/40). There were no significant differences in mastitis prevention knowledge scores between the two groups pre-intervention (Table 2).

Table 1. Demographic and other characteristics of 40 smallholder dairy farms participating in a cellphone training
trial on dairy management in Kenya in 2017.

Variable Names and Categories

Intervention
Group (n=24)

Comparison
Group (n=16)

p

Gender

0.34

Female

29.2% (7)

43.8% (7)

Male

70.8% (17)

56.3% (9)

Marital status

0.06

Married

87.5% (21)

93.8% (15)

Divorced or widowed

4.2% (1)

6.2% (1)

Single

8.3% (2)

0.0%

Education attained by principal farmer (regardless of gender)

1.00

Primary

50.0% (12)

50.0% (8)

Secondary

45.8% (11)

50.0% (8)

University/college

4.2% (1)

0.0%

Proportion of total income from dairy

0.13

Less than 50%

8.3% (2)

31.3% (5)

50 – 75 %

83.3% (20)

68.8% (11)

More than 75 %

8.3% (2)

0.0%

Proportion of land used for dairy

0.89

25% or less

33.3% (8)

25.0% (4)

50 – 75

50% (12)

62.5% (10)

More than 75%

16.7% (4)

12.5% (2)

Attended any training within the last year

0.06

Yes

50.0% (12)

81.2% (13)

No

50.0% (12)

18.8% (3)

Subject of training if attended training within the last year

0.27

Can’t remember

75.0% (9) *

46.2% (6) *

General husbandry and feeding

16.7% (2) *

15.4% (2) *

Silage making

8.3% (1) *

38.4% (5) *

* Based on n= 12 and 13 in the two groups, respectively (those who attended some training)

Feeding knowledge (and its application) was similar between the two study groups pre-intervention. One-third of farmers (13/40) knew that it was good to supplement calf diets with some concentrate and thought dairy meal would suffice, while 80.0% (32/40) of farmers knew that they needed to supplement the diet of dairy cows with dairy meal for steaming up pre-calving. There were no significant differences in nutrition knowledge score between farmers in the two groups pre-intervention (Table 2).

Table 2. Mean knowledge scores on mastitis prevention and feeding for 40 smallholder dairy
farms participating in a cellphone training trial on dairy management in Kenya in 2017.

Mean knowledge scores

Intervention
(n=24)

Comparison
(n=16)

p

Pre-intervention

Mastitis prevention

3.8 (s.d. 1.9)

4.7 (s.d. 1.1)

0.07

Feeding

2.2 (s.d. 0.5)

2.0 (s.d. 1.0)

0.48

Post-intervention

Mastitis prevention

4.3 (s.d. 1.4)

1.8 (s.d. 0.8)

<0.001

Feeding

2.3 (s.d. 0.7)

2.4 (s.d. 0.8)

0.69

Net change

Mastitis prevention

0.5 (s.e. 0.4)

- 2.9 (s.e. 0.3)

<0.001

Feeding

-0.1(s.e. 0.2)

0.4 (s.e. 0.3)

<0.001

Intervention summary and feedback

All farmers in the intervention group received cellphone training messages during the 3-month intervention period. Most (70.8% - 17/24) of these farmers did not keep track of the number of messages sent to them and mentioned that they received many messages. Although a message was sent out daily for 5 days a week, the mean number of messages reported to be received by farmers was 4.4 messages per week, with a s.d. = 2.0 messages per week. Some farmers (29.2% - 7/24) estimated they had received between 4-7 messages during the entire training period, lowering the average. Most farmers reported that they always read 100% of the message (the entire message) (Figure 2-1).

From the post-intervention meeting with the intervention group, farmers generally found the content of the message understandable, except for one farmer who had some difficulty understanding the messages (Figure 2-2). On a scale of 1 (very easy to understand), 2 (easy to understand), 3 (somewhat easy to understand), 4 (difficult to understand), and 5 (very difficult to understand), the mean, s.d. and median scores for content understandability were 2.3, 0.9, and 3.0, respectively.

On a scale of 1 (very informative), 2 (informative), 3 (somewhat informative), 4 (not very informative), and 5 (not informative at all), the mean, s.d. and median scores regarding how informative the messages were comprised of 2.3, 1.0, and 3.0, respectively. More than a third of farmers reported that the messages were very informative (Figure 2-3). Over half of the farmers felt extremely or very motivated (Figure 2- 4) to practically implement the dairy cow management practices from messages such as those covering mastitis prevention and Napier grass feeding and other cow nutrition practices. Additionally, most farmers felt that the messages received (such as management of cases of retained placenta) were very effective for their dairy management systems (Figure 2-5).

Since the messaging was one-way (farmers could not ask questions for clarification), the extent of the challenge faced by the farmers regarding not being able to call back to inquire about the messages was assessed on a scale of 1 (not challenging at all), 2 (slightly challenging), 3 (challenging), 4 (very challenging), and 5 (extremely challenging). Eleven of the 24 intervention farmers were indifferent and so didn’t respond to the question. The challenge of not knowing who to call back about the messages was not largely experienced among the farmers except for those who found this a big challenge (Figure 2-6), with mean, s.d. and median scores of 2.0, 1.5 and 1.0, respectively.

A third of the farmers (8/24) had some questions and concerns about some messages received in the 3-month intervention period. A few of these farmers with concerns (37.5% - 3/8) thought that the messages were not very clear and orderly for thematic continuity in each message, while 25.0% (2/8) of these farmers had concerns that some of the translations from English to the local (Kimeru) language were difficult to contextualize on their farms. However, 37.5% (3/8) of farmers with concerns chose not to articulate their concerns altogether. Some of the 8 farmers (25.0% - 2/8) that had concerns chose to ask for help from veterinarians, veterinary technicians or their neighbors to read and better understand the knowledge, while the rest chose to ignore the concerns and understand the messages as they had read them.

From the FGD, some farmers expressed a challenge not previously envisioned. Since the screen of the feature phones was small, scrolling through to read a full 160-character message took some time.

Post-intervention comparison between and within groups

Knowledge on the different practices taught as beneficial methods of mastitis control (using a different wash cloth for each milking cow, drying udder before milking with a clean cloth or paper towel, using a different drying cloth for each milking cow, using a teat dip post-milking, giving fresh feed soon after milking, using dry-cow treatment when drying cows off prior to calving, and not leaving milk in the udder to allow calves to suckle) was again assessed for the two groups post-intervention.

The mean mastitis prevention knowledge score on comparison farms decreased, but in the intervention group, there was an increase in mean score on knowledge of mastitis prevention practices, producing a net change in knowledge on mastitis prevention of 3.4 between the 2 groups, which was significant (p < 0.01) (Table 2). From FGD, it was evident that, although farmers in the intervention group were more knowledgeable about some of these practices post-intervention, the rationale was not always clear to them. Clarification was provided on how each of the practices was relevant in reducing mastitis occurrence on farms.

There was also a difference between groups in knowledge level on diarrhea prevention post-intervention (p < 0.01). Most of the intervention group (87.5% - 21/24) and (25.0% - 4/16) of the comparison group - knew that housing the calf in a clean and dry place would help reduce occurrence of calf diarrhea cases. Similarly, post-intervention, more farmers in the intervention group (66.7% - 16/24) knew that always providing dry bedding and removing manure from the stall daily was helpful in preventing diarrhea in calves compared to the comparison group at 0% (p < 0.01). There were no differences in diarrhea prevention knowledge between groups pre-intervention. From the FGD, it was apparent that although farmers had calves on their farms, most of them did not have conventional stalls/pens for their calves. Because of this farming practice, bedding in calf pens was not a major consideration for them.

Figure 2. Descriptive analysis of attitudes and experience of 24 farmers in the intervention group participating in a cellphone training trial on dairy management in Kenya in 2017.

On causes of teat blockage, there was a significant difference (p = 0.02) in the understanding that udder infection was a risk factor for teat blockage between the comparison (62.5%, 10/16) and intervention (91.7%, 22/24) groups post-intervention. Similarly, more farmers in the intervention group (58.3%, 14/24) than the comparison group (18.8%, 3/16) knew (post-intervention) that improper milking techniques (pulling hard on the teats during daily routine milking) was associated with teat blockage (p = 0.01). Compared to the comparison group, post-intervention, the intervention group was also more aware that teat blockage problems could be an inherited problem (0% - 0/16 vs 25.0% - 6/24, respectively) (p = 0.03). There were no differences in teat blockage knowledge between groups pre-intervention.

From the FGD, farmers indicated that pulling the teat during milking was necessary for some cows because they had small teats due to cross breeding of Bos taurus breeds with Bos indicus breeds, with the latter mostly having small teats. With this cross-breeding being common in the area, most farmers had habituated to this pulling technique of milking, even when the cows had standard size teats that could be milked easily using the squeezing technique.

The mean knowledge score on feeding practices was assessed based on the understanding of ideal height for harvesting Napier, need for dairy meal for steaming up cows pre-calving, and colostrum feeding times for newborn calves. Although there was a 0.4 increase and a 0.1 decrease in the mean knowledge score in the comparison and intervention groups respectively, these changes were not significant. There was no significant difference between the two groups post intervention. However, the net change of 0.5 in scores on knowledge about recommended feeding practices was significant (Table 2). When asked about the benefits of good nutrition post-intervention, 58.3% (14/24) of the intervention farmers were more knowledgeable (p = 0.04) on the role of a balanced nutritious diet in supporting the resolution of rain scald compared to 25.0% (4/16) in the comparison group. During the FGD, farmers from both groups mentioned that feeding cows on short Napier grass and steaming up were not very novel concepts to them since they had been trained about them in other seminars as well. However, the physiological rationale behind these practices were not clear to them.


Discussion

Analyses in this trial ultimately involved 40 farms randomly selected and allocated into intervention and comparison groups. Training the intervention group through SMS on smallholder dairy management best practices for 3 months resulted in significant increases in dairy management knowledge scores on various husbandry aspects in the intervention group. Farmers in the intervention group were more knowledgeable on mastitis prevention practices, associations between stall hygiene and calf disease, as well as some beneficial nutritional management practices post-intervention; indicative of the improvement of knowledge for better production, irrespective of previous training and formal education levels. This improvement in knowledge could be attributed to the fact that by using cellphones as a training tool, farmers could keep the information with them at their fingertips for potentially long periods of time (Martin & Hall, 2011). Moreover, the cellphone messaging as a training tool was well-received by the farmers, who read the messages and were largely motivated to implement the recommendations. In addition, the farmers considered a frequency of one message a day as a suitable and effective way of delivering training content to them.

These trial findings had some semblance to findings in other SDFs in Kenya; a study by (Staal et al 2003) highlighted a positive effect on milk production (not evaluated in this study) when farmers in Njoro sub-county used husbandry information received through mobile phones. However, a trial in Machakos County found that: 1) use of cellphones had no significant effect or influence on agricultural practices by farmers; 2) use of SMS among farmers was more likely when farmers had formal education of high school level and above; and 3) more farmers preferred use of voice to SMS since SMS was tedious (Mutunga & Waema, 2016). Nevertheless, neither of these studies was expressly designed to evaluate effectiveness of using cellphones as a training tool towards improved knowledge, which was the focus of this study.

This study population was generally representative of other SDFs in Kenya. Most of the principal farmers were female, which has been observed in other studies (Gallina, 2016). However, in the current training and research sessions, more men (65.0%) than women participated, which is likely a result of women being busy with chores at home and the patriarchal culture. Men are more frequently involved in off-farm activities, such as attending training and research sessions, than women. Some of the men attending the sessions indicated that they were representing their wives. Comparative pre-intervention analysis between the two groups showed that the groups were generally alike.

Similarly, on dairy management practices, such as feeding and mastitis prevention, there was no significant difference between the groups prior to intervention. The random allocation assisted in mitigating possible selection bias (Kahan et al 2015).

Analogous to other findings in other areas in Kenya (Richards et al 2015), dairy production was the main source of income for most (55.0%) farmers. The land acreages in this study population were small, with an average of 3.4 acres available for dairy and crop farming. The average land size of these SDFs was slightly higher than the average (2 -2.8 acres) documented in the region (Mugambi et al 2015) but within the range documented by other studies in Kenya (Omiti et al 2006; Vanleeuwen et al 2012).

Most farmers in this study were keen on dairy production and thus had attended some form of training on dairy management. This is a common happening in dairy cooperatives in Kenya where the dairy cooperative organizes seminar/extension sessions for farmers to increase knowledge and improve production (Wambugu et al 2011; Ettema, 2012). However, a short-coming of this form of farmer training has been that the knowledge retention can be relatively low among session attendants. Less than a half of the farmers in this study were able to recall details of the trainings they had attended within the last year pre-intervention.

Use of cellphone messaging for information dissemination in Kenya has increased in the last decade; in the agricultural sector, this dissemination has played a great role in enhancing information transfer between farmers, researchers and industry representatives (Kiptum, 2016). In recent years, cellphones technology has been adopted and is now used for some agricultural purposes in Kenya. Most of the innovations being prioritized include using SMS on cellphones for information access to farmers (The World Bank, 2012). Using one-way messaging in this study was done through MySQL on an Apache-based SMS gateway server which allows for transmission of uniform messages from a server to many individuals as a promotional item (Hussain, 2016). MySQL is essentially a common language for accessing databases (Oracle Corporation, 2013). Apache is one of the most widely used software for database interaction, visualization and management (Balkhi, 2009).

Overall, the farmers reported that the message reception was good. On assessing how the farmers felt about the information in the messages, the general feeling was that the messages were informative, and hence more than half of the farmers were extremely or very motivated to implement dairy management advice in the messages. This messaging encouraged discussions between farmers, as well as consultations with veterinary service providers, especially when some components were unclear. The use of cellphone messaging as a form of information dissemination has been shown to increase farmer-to-farmer training and uptake of various technologies (The World Bank, 2012). Unfortunately, most feature phones have small storage capacities and thus farmers sometimes need to delete older messages when the phone memory is full. However, farmers mentioned that they read most of the messages sent to their phones at any one moment and could retain the messages that provided new information to them and they preferred not to delete them. For this reason, receiving the messages made them happier compared to one-day farmer seminar trainings. Similar findings in relation to content retention were observed in another study (Farm Africa, 2015).

The main message-related challenge highlighted by the farmers in the FGD was that parts of the message were not easily readable on the small screen of the feature phone. For example, some farmers reported that some messages were longer than the phone screen display could handle at one time and took a long time to scroll through it at the time of receiving the message. Farmers in the FGD said they sometimes took a break in reading one long message, and then they sometimes forgot to read the rest of the message later. However, the farmers indicated that translation of messages into the local language was a welcome idea. Although some farmers had a challenge in understanding some translated words, the messages were still considered by most farmers to be easy to understand. Cellphone training has been considered a sustainable approach to support the use of media technologies for training purposes (World Bank Group, 2017).

The mastitis prevention knowledge scores in the comparison group appeared to decrease significantly, while the intervention group scores increased slightly. The relatively high pre-intervention knowledge scores for the comparison group (compared to post-intervention) may have been a result of the pre-intervention meeting for the intervention group being held a day before that of the comparison group. Farmers in the intervention and comparison groups may have discussed the contents of the questionnaires and the meeting since they were all within the same relatively small community, resulting in higher comparison group scores. This unintended dissemination was mitigated post-intervention where both groups were interviewed at different times of the same day. It is therefore recommended that reducing the interval between assessments of the study groups would reduce probability of information transfer between the comparison and intervention groups.

Knowledge scores on feeding practices, such as the amount and time of colostrum given to calves, and ideal height of Napier grass harvesting for optimum milk production, also had significant net changes between the two groups. The comparison group appeared to have increased their knowledge scores while the intervention group didn’t change much. This unexpected net change in knowledge score could be attributed to the fact that such information was also communicated to the farmers by another NGO in the area and in other training forums. However, knowledge on other nutritional information (such as for some skin conditions (rain scald) which could quickly resolve when cows are fed a well-balanced diet) (Roberson et al 2012)was significantly higher in the intervention group than comparison group. Similarly, more farmers in the intervention group were knowledgeable on the benefits of hygiene in calf diarrhea management post-intervention.

A shortcoming of the current study was due to technological constraints. The basic format of content deliverable to the feature phones meant that non-text information, such as pictures and diagrams, could not be used to augment the messages for better understanding by the farmers. While smart phones are currently uncommon in rural parts of Kenya (only 1.7% of the study population had a smart phone in the family), as the price of smart phones decreases, leading to wider utilization, auxiliary material can be provided to farmers for training purposes.

Another limitation was loss to follow-up in both the intervention and comparison groups, reducing the final sample size and power of this study. The reasons for farmers not completing the trial were unlikely to be related to the study objectives, minimizing any bias from this attenuated sample size. However, a smaller sample size leads to reduced power to detect significant differences between groups. Fortunately, we were still able to find significant differences in knowledge between the groups, even with the smaller sample size.

A final limitation of the study was the unintended dissemination of knowledge during the trial, leading to farmers in the comparison group improving in their knowledge scores. A change in study design to have different populations farther apart and a shorter period between evaluation of study groups would help reduce unintended information diffusion between the groups and ‘contamination’ of the comparison group.

We hypothesize that, the improvement in knowledge of the farmers in this study would most likely translate into better dairy management, production and improved incomes to the farmers. This effect of training was observed by (Richards et al 2016) where good use of high protein forages coupled with continuous on-farm education and training on best management practices, significantly increased daily milk production in SDFs in Kenya. While a significant difference was observed on knowledge levels of trainees, further investigation on effectiveness of this form of training on actual milk production and practices would be helpful.

A cost-benefit analysis would likely show that use of cellphones for training is a cost-effective approach for knowledge transfer from the farmer’s perspective, given that in Kenya, most farmers already have a cell phone and do not pay anything to receive messages. Therefore, a subscription fee would likely be the only real cost to the farmer, along with a slight increase in charging costs if the phone was used more. Benefits to the farmer could be substantial, depending on the improvements made on the farm. A comparative investigation of effectiveness of seminar training vs cellphone training would be informative on the impact and sustainability for such alternative farmer training methods.

With this study population, interventions implemented needed to be based on a feature phone interface. Evaluation of a smartphone application that allows a more interactive interface between the farmer and messages on smartphones could be explored in the future when smartphones become more common among rural farmers. Furthermore, a trial to compare the differences in cost and impact of training using the feature phone and smartphone would be more informative on the best cellphone interventions for SDFs in Kenya.


Conclusion and Recommendations


Acknowledgements

We are grateful to the primary funding program for this research, the Canadian Queen Elizabeth II Diamond Jubilee Scholarships (QES) which are managed through a unique partnership of Universities Canada, the Rideau Hall Foundation (RHF), Community Foundations of Canada (CFC) and Canadian universities. This program is made possible with financial support from the Government of Canada, provincial governments and the private sector. We also acknowledge the large contribution made by volunteers and staff of Farmers Helping Farmers. As well, the support of the NDFCS, the cooperation of Upendo Women’s Group and the primarily women dairy farmers who made it all possible.


References

Alhassan H and Kwakwa P A 2012 The Use of Mobile Phones by Small Scale Farmers in Northern Ghana: Benefits and Challenges. Journal of Entrepreneurship and Management, 1(3), 40–45. Retrieved from http://www.publishingindia.com/GetBrochure.aspx?query=UERGQnJvY2h1cmVzfC8xMzg1LnBkZnwvMTM4NS5wZGY=

Balkhi S 2009 What is Apache? - What is a Web Server? Retrieved June 16, 2018, from http://www.wpbeginner.com/glossary/apache/

Barr D R 2018 Using Confidence Intervals to Test Hypotheses. Journal of Quality Technology, 1(4), 256–258. https://doi.org/10.1080/00224065.1969.11980385

Ettema B F 2012 Dairy Development in Kenya. Retrieved May 11, 2018, from http://www.dairyfarmer.net/fileadmin/user_upload/40_downloads/kenya-dairying-ETTEMA.pdf

Farm Africa 2015 Testing the Use of Mobile Technology to Improve Smallholder Sesame Cultivation. Retrieved May 14, 2018, from https://www.farmafrica.org/downloads/resources/ict244lowres.pdf

Gallina A 2016 Gender dynamics in dairy production in Kenya: A literature review (No. 182). Copenhagen, Denmark. Retrieved from https://ccafs.cgiar.org/publications/gender-dynamics-dairy-production-kenya-literature-review

Hove L, Franzel S and Moyo P S 2003 Farmer experiences in the production and utilisation of fodder trees in Zimbabwe: Constraints and opportunities for increased adoption. Tropical Grasslands, 37, 279–283.

Hussain S N 2016 SMS Requirements and Highlevel Design. Retrieved June 16, 2018, from https://cwiki.apache.org/confluence/pages/viewpage.action?pageId=65865302

Kahan B C, Rehal S and Cro S 2015 Risk of selection bias in randomised trials. Trials, 16(1), 1–7. https://doi.org/10.1186/s13063-015-0920-x

Karlsen C, David A and Hopmann N 2010 Kenya Connected: Mobile Technology is linking Journalists to Local Sources. Journalism. University of Southern Denmark.

Kiptum J C 2016 ICT Framework for Adoption in the Dissemination of Agricultural Information in Kenya :( Case Study of Agricultural Development Corporation). University of Nairobi. Retrieved from http://erepository.uonbi.ac.ke/handle/11295/97618

Lukuyu B, Franzel S, Ongadi P M and Duncan A J 2011 Livestock feed resources: Current production and management practices in central and northern rift valley provinces of Kenya. Livestock Research for Rural Development., 23(5), 112. http://www.lrrd.org/lrrd23/5/luku23112.htm

Makau D N, VanLeeuwen J A, Gitau G K, Muraya J, McKenna S L, Walton C, and Wichtel J J 2018 Animal and management factors associated with weight gain. Preventive Veterinary Medicine, (In press).

Martin B L, and Hall H 2011 Mobile Phones and Rural Livelihoods: Diffusion, Uses, and Perceived Impacts Among Farmers in Rural Uganda (7 No. 4) (Vol. 7). Annenberg.

Mittal S, and Tripathi G 2009 Role of Mobile Phone Technology in Improving. In Agricultural Economics Research Review (Vol. 22, pp. 451–459). Indian Council for Research on International Economic Relations (ICRIER).

Mugambi D K, Mwangi M, Kairu S and Gitunu A M 2015 Assessment of performance of smallholder dairy farms in Kenya: an econometric approach. Journal of Applied Biosciences, 82, 7891–7899. https://doi.org/10.4314/jab.v85i1.13

Muraya J, VanLeeuwen J A, Gitau G K, Wichtel J J, Makau D N, Crane M B, McKenna S L B, Tsuma V T 2018 Cross-sectional study of productive and reproductive traits of dairy cattle in smallholder farms in Meru, Kenya. Livestock Research for Rural Development. Volume 30, Article #171. http://www.lrrd.org/lrrd30/10/jmura30171.html.

Mutunga I M and Waema T M 2016 Context of mobile phone use and its effects on smallholder farmer’ s livelihood outcomes in Kenya. International Journal of Scientific Research and Innovative Technology , 3(4), 121–136.

Mwangi D M and Wambugu C 2003 Adoption of forage legumes: the case of Desmodium intortum and Calliandra calothyrsus in central Kenya. Tropical Grasslands, 37, 227–238.

Ngwira A 2003 Forage demand and constraints to adoption of forage technologies by livestock keepers in Malawi. Tropical Grasslands, 37, 274–278.

Ofosu-asare K 2011 Trade Liberalisation, Globalisation and the Cocoa Industry in Ghana: The Case of the Smallholder Cocoa Farmers Kwaku Ofosu-Asare. University of Westminster. Retrieved from http://westminsterresearch.wmin.ac.uk/10484/1/Kwaku_OFOSU-ASARE.pdf

Omiti J, Wanyoike F, Staal S and Delgado C 2006 Will Small-Scale Dairy Producers in Kenya Disappear Due to Economies of Scale in Production? Retrieved July 22, 2018, from https://ageconsearch.umn.edu/bitstream/25674/1/cp060516.pdf

Oracle Corporation 2013 What is MySQL? Retrieved June 12, 2018, from https://dev.mysql.com/doc/refman/8.0/en/what-is-mysql.html

Rathod P, Chander M and Bardhan D 2016 Adoption Status and Influencing Factors of Mobile Telephony in Dairy Sector: A Study in Four States of India . Agricultural Economics Research Review, 29(1), 15–26. https://doi.org/10.5958/0974-0279.2016.00015.X

Richards S, VanLeeuwen J, Shepelo G, Gitau G K, Kamunde C, Uehlinger F and Wichtel J 2015 Associations of farm management practices with annual milk sales on smallholder dairy farms in Kenya. Veterinary World, 8(1), 88–96. https://doi.org/10.14202/vetworld.2015.88-96

Richards S, VanLeeuwen J, Shepelo G, Gitau G K, Wichtel J, Kamunde C and Uehlinger F 2016 Randomized controlled trial on impacts of dairy meal feeding interventions on early lactation milk production in smallholder dairy farms of Central Kenya. Preventive Veterinary Medicine, 3954, 1–8. https://doi.org/doi:10.1016/j.prevetmed.2016.01.006

Roberson J R, Baird A N and Pugh D G 2012 Diseases of the Integumentary System. In D. G. Pugh and A. N. Baird (Eds.), Sheep and Goat Medicine (Second Edi, pp. 256–290). W.B. Saunders.

Sahota C and Kameswari V L V 2014 Mobile Phones Agricultural Extension. Journal Pengajian Media Malaysia, 16(1), 1–11.

Salia M, Nsowah-Nuamah N N N and Steel W F 2011 Effects of Mobile Phone Use on Artisanal Fishing Market Efficiency and Livelihoods in Ghana. The Electronic Journal of Information Systems in Developing Countries , 47(1), 1–26. https://doi.org/10.1002/j.1681-4835.2011.tb00334.x

Smollo J W, Ali-Olubandwa A M and Ng’endo C M 2016 Influence of Utilizing Animal Husbandry Information from Mobile Phones on Milk Yield Among Smallholder Dairy Farmers in Njoro Sub- County, Kenya. International Journal of Agricultural Extension, 04(01), 41–47.

Staal S J, Waithaka M, Njoroge L, Mwangi D M, Njubi D and Wokabi A 2003 Costs of milk production in Kenya: Estimates from Kiambu, Nakuru and Nyandarua Districts. Nairobi, Kenya. https://doi.org/10.13140/2.1.2945.9206

The World Bank 2012 Using ICT to enable Agricultural Innovation Systems for Smallholders ICT innovations. Retrieved May 14, 2018, from http://www.fao.org/docrep/018/ar130e/ar130e.pdf

Vanleeuwen J A, Mellish T, Walton C, Kaniaru A and Gitau R 2012 Management, productivity and livelihood effects on Kenyan smallholder dairy farms from interventions addressing animal health and nutrition and milk quality. Tropical Animal Health and Production, 44, 231–238. https://doi.org/10.1007/s11250-011-0003-2

Wambugu S, Kirimi L and Opiyo J 2011 Productivity trends and performance of dairy farming in Kenya (WPS 43/2011). Tegemeo Institute of Agricultural Policy and Development (Vol. WPS 43/201). Nairobi, Kenya.

World Bank Group 2017 Agriculture ICT Extension Services. Retrieved May 15, 2018, from https://www.innovationpolicyplatform.org/system/files/4_Agri ICT Extension_Agri_Nov20.pdf


Received 2 October 2018; Accepted 23 October 2018; Published 1 November 2018

Go to top