This study employed a descriptive cross-sectional survey design supported by quantitative analytical approaches to examine the factors influencing the adoption of modern rice technologies among farmers in Kambia District, Sierra Leone. The design was appropriate because it enabled the researchers to collect data from a large number of respondents at a single point in time while analyzing relationships between socio-economic, institutional, and technological factors affecting adoption behavior. Inferential statistical techniques such as logistic regression analysis, chi-square tests, post-hoc analysis, and Kendall’s coefficient of concordance (Kendall’s W) were incorporated to identify significant predictors and associations among variables.

The study was conducted in Kambia District in the Northern Province of Sierra Leone, one of the major rice-producing regions in the country. The district is characterized by inland valley swamps, bolilands, and upland farming systems suitable for rice cultivation. The study was conducted across three chiefdoms in four communities in Kambia District, namely Kasiri, Kychon, in Samu Chiefdom (8.94˚N, −13.11˚W, 8.93˚N, −13.14˚W), respectively, Robana, in Mambolo Chiefdom (8.91˚N, 13.04˚W), and Rokupr in Magbema Chiefdom (9.01˚N, −12.95˚W) were purposively selected due to their high involvement in rice farming and varying levels of access to agricultural technologies and support services. The target population consisted of registered and non-registered rice farmers actively engaged in rice production during the 2025/2026 farming season, including both male and female farmers, smallholder and medium-scale producers, as well as members and non-members of farmer organizations.

A total sample size of 312 rice farmers was used for the study. The respondents were proportionately distributed across the selected chiefdoms as follows: Kasiri (71), Kychon (71), Robana (90), and Rokupr (80). The study adopted a multi-stage sampling procedure involving purposive, stratified, and simple random sampling techniques. First, the chiefdoms were purposively selected based on rice production activities and accessibility. Second, farmers were stratified according to chiefdom, gender, membership in farmer organizations, and type of rice cultivated to ensure balanced representation. Finally, simple random sampling was used to select respondents from farmer lists obtained from cooperatives and extension officers. The sample size for the study was determined using the Yamane [22] formula for sample size determination. The formula is commonly used in quantitative research to determine a representative sample from a known population (Table 1).

n = N/(1 + N(e²))

where:

n = required sample size;

N = total population size;

e = margin of error (0.05).

Both primary and secondary data sources were utilized in the study. Primary data were collected through structured questionnaires, field interviews, and farmer consultations, while secondary data were obtained from Ministry of Agriculture reports, Food and Agriculture Organization publications, agricultural policy documents, journals, extension service records, and rice development project reports.

Table 1.Proportionate distribution of sample size.

The questionnaire contained both closed-ended and multiple-choice questions designed to capture information on demographic characteristics, farming experience, farm size, and access to agricultural inputs, irrigation, credit, extension services, training opportunities, and adoption of modern rice technologies. The instrument was reviewed by agricultural experts and supervisors to ensure content and face validity, while pre-testing was conducted in a nearby farming community outside the study area to improve clarity and reliability. A pilot survey was also carried out to test internal consistency, and ambiguous items were revised accordingly.

Data collection was conducted through face-to-face interviews administered by trained research assistants familiar with local languages and farming practices. Permission was obtained from local authorities and farmer leaders before data collection commenced. Respondents were informed about the purpose of the study, and their participation was voluntary. Interviews were preferred because many respondents had low levels of formal education, making self-administered questionnaires unsuitable. The dependent variable in the study was the adoption of modern rice technologies, coded as 1 for adopters and 0 for non-adopters. Independent variables included age, gender, education level, farming experience, farm size, land ownership, access to fertilizer, access to machinery, irrigation availability, access to credit, extension services, training opportunities, labour availability, farmer organization membership, chiefdom location, input availability, and main source of household income.

Qualitative data were collected through Key Informant Interviews (KIIs) with purposively selected stakeholders who had extensive knowledge of rice production and technology dissemination in the study area. The key informants included agricultural extension officers, rice farmer association leaders, community leaders, agricultural input dealers, representatives of non-governmental organizations, and officials from relevant agricultural institutions. A semi-structured interview guide was used to explore the availability of resources and factors influencing the adoption of modern rice technologies, including access to improved seeds, credit facilities, extension services, farm inputs, training opportunities, market access, and institutional support. Interviews were conducted face-to-face, recorded with participants’ consent, and supplemented with field notes. The qualitative data provided detailed insights into the opportunities and constraints affecting farmers’ decisions to adopt modern rice technologies and were used to complement and validate the quantitative survey findings.

Binary logistic regression analysis was employed to determine the factors influencing adoption of modern rice technologies. The model estimated the probability of adoption using socio-economic and institutional predictors, while model fitness was assessed using deviance statistics, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), McFadden R², Nagelkerke R², Cox and Snell R², and Tjur R². Chi-square tests were used to examine relationships between categorical variables such as rice variety cultivated and household income source, as well as access to improved seed varieties and income source. Where significant associations were identified, post-hoc pairwise comparisons with Bonferroni correction were conducted to determine specific group differences. Kendall’s coefficient of concordance (Kendall’s W) was also used to measure the level of agreement among farmers regarding the ranking of factors influencing adoption of modern rice technologies.

Key Informant Interview (KII) data were analyzed using thematic content analysis. Interview recordings were transcribed verbatim and carefully reviewed to identify recurring patterns and emerging issues. Responses were coded and grouped into themes based on similarities in meaning and relevance to the study objectives. Major themes identified included observed weather changes, climate-related agricultural challenges, impacts on farming systems, and adaptation constraints. The qualitative findings were subsequently triangulated with quantitative survey results to enhance the validity and depth of interpretation. Representative quotations from key informants were used to illustrate key themes and provide contextual explanations for the quantitative findings. This approach enabled a comprehensive understanding of farmers’ experiences and perceptions regarding climate change and its effects on agricultural livelihoods.

Ethical considerations were strictly observed throughout the study. Respondents participated voluntarily, informed consent was obtained, and confidentiality and anonymity of responses were guaranteed. Participants were assured that the information provided would be used strictly for academic purposes. Despite challenges such as limited financial resources, poor road accessibility, low literacy levels among respondents, incomplete farmer records, and time constraints, adequate measures were taken to ensure the reliability and validity of the study findings.

The results from Table 2 reveal that rice farmers in Kambia District are predominantly male (65.1%), although female participation is notably high in Rokupr (62.5%), indicating some gender variation across chiefdoms. The age distribution shows that most farmers fall within the economically active age groups (33 - 57 years), suggesting a relatively mature and experienced farming population, which is further supported by the finding that a large proportion have 11 - 20 years of farming experience, suggesting that rice farming is largely undertaken by mature farmers with substantial practical knowledge [23]-[25]. Educationally, the majority of respondents have low levels of formal education, with 35.6% having no formal schooling and only a small proportion attaining tertiary education, which may limit access to agricultural information, extension services, and the effective adoption of modern rice technologies [26]-[28]. Farm sizes are generally small to medium-scale (2 - 5 acres), reflecting the dominance of smallholder farming, although a few farmers manage larger holdings. Rice farming is the primary source of household income (93.6%), highlighting its central role in livelihoods and food security in the district [29][30]. In addition, more than half (57.7%) of respondents belong to farmer organizations, particularly in Robana and Rokupr, which may improve access to information, inputs, and collective support systems [31][32]. The findings indicate that rice production in Kambia is largely driven by experienced smallholder farmers with limited education but strong dependence on agriculture, with variations in gender roles, organizational membership, and resource access across chiefdoms.

Table 2. Demographic characteristics of respondents.

The findings indicate that access to key resources for adopting modern rice technologies in Kambia District is uneven and highly dependent on location, with better access generally observed in Robana and Rokupr compared to Kasiri and Kychon (Table 3). A majority of farmers cultivate improved varieties such as ROK 10 (55.1%), though a significant proportion still rely on traditional varieties (27.9%), especially in less advantaged chiefdoms. Access to improved seeds and fertilizers is relatively strong in Robana and Rokupr but remains limited in Kasiri and Kychon, where many farmers report no access [33][34]. Mechanization is generally low, with only 1.6% owning machinery, while most farmers either rent/borrow (43.3%) or have no access at all (45.2%) [35][36]. Similarly, irrigation is inadequate, as 41.4% depend solely on rainfall, highlighting vulnerability to climate variability and drought risks [37][38]. Financially, farmers rely mainly on personal savings (38.8%) and cooperatives (34.0%), with minimal access to formal banking services, and half of respondents reported that credit is never available when needed [39]. Training on modern technologies is also limited, with 42% receiving no training, particularly in Kasiri and Kychon [40]. Labour availability varies, being more sufficient in Robana and Rokupr, while land access is dominated by rented (43.9%) and leased systems,indicating tenure insecurity. Additionally, timely availability of inputs remains a challenge, with delays or shortages reported by many farmers. The results suggest that while some progress has been made in improving access to inputs and support services, significant constraints in finance, training, irrigation, mechanization, and input distribution continue to limit the widespread adoption of modern rice technologies across the district [41][42].

The qualitative findings revealed that “rice farmers in the study areas have access to limited forms of resources that support the adoption of modern rice technologies. Respondents identified both physical and human resources as the main resources currently available. Physical resources included basic farming tools, land, and labor, while human resources mainly referred to agricultural extension workers and experienced farmers who provide technical guidance and farming advice. However, participants explained that these resources are inadequate to fully support large-scale adoption of modern rice technologies”.

A significant chiefdom effect was observed in the adoption of modern rice technologies, indicating that adoption rates varied across locations due to differences in local resource availability and institutional support. Chiefdoms with higher adoption levels generally had better access to improved seed varieties, farmer training programs, extension services, and agricultural inputs, which enhanced farmers’ knowledge and confidence in using modern technologies. Differences in irrigation facilities and reliable water sources also influenced adoption, as farmers with access to irrigation were more willing to invest in improved practices than those relying solely on rainfall. Similarly, greater availability of mechanization services, such as tractors and threshers, reduced labour constraints and facilitated technology uptake. Qualitative findings further revealed that some chiefdoms benefited from stronger support from government agencies, NGOs, and agricultural development projects, leading to increased training, input distribution, and technical assistance. Hence, the chiefdom effect reflects spatial inequalities in access to resources, extension services, irrigation infrastructure, mechanization, and institutional support, highlighting the need for targeted interventions to ensure equitable dissemination of modern rice technologies across all chiefdoms in Kambia District.

Table 3. Access to available resources for adoption of modern rice technologies.

The logistic regression analysis assessed the factors of modern rice technology adoption among farmers, incorporating a range of socio-demographic, economic, and resource-related variables. The logistic regression model significantly improved the explanation of adoption behavior compared to the null model, as indicated by the reduction in deviance from 397.2 to 246.7 and the highly significant chi-square value (Δχ² = 150.49, p < 0.001) (Table 4). This confirms that the selected socio-economic and institutional variables collectively explain variations in farmers’ adoption of modern rice technologies [43][44]. The explanatory power of the model was relatively strong, with McFadden R² = 0.379, Cox and Snell R² = 0.383, Nagelkerke R² = 0.531, and Tjur R² = 0.427. These statistics suggest that the included predictors account for approximately 38% - 53% of the variation in adoption behavior [45][46]. Overall, the model demonstrates that factors such as access to resources, institutional support, and farm characteristics are important in influencing farmers’ technology adoption decisions [47][48].

The logistic regression analysis examined the factors associated with modern rice technology adoption among farmers using socio-demographic, economic, and resource-related variables. The model significantly improved the explanation of adoption behavior compared to the null model, as evidenced by the reduction in deviance from 397.2 to 246.7 and the highly significant likelihood ratio chi-square statistic (Δχ² = 150.49, p < 0.001) (Table 4), indicating that the included variables collectively contribute to explaining adoption decisions (Assaye et al., 2023; Melaku Baye et al., 2025) [43][44]. The model also demonstrated good fit, with McFadden R² = 0.379, Cox and Snell R² = 0.383, Nagelkerke R² = 0.531, and Tjur R² = 0.427. However, these pseudo-R² measures should be interpreted cautiously, as they are not directly comparable to the R² statistic used in ordinary least squares regression. Consistent with Train (1977) [49], who suggested that values between 0.20 and 0.40 indicate excellent model fit in logistic regression, the results suggest that the model provides substantial explanatory capacity and effectively distinguishes adopters from non-adopters rather than explaining a fixed percentage of variation in adoption behavior.

Table 4. Logistic regression model summary for the adoption of modern rice technology.

Note: M₁ includes What is the level of access to farm machinery (tractor, power tiller, thresher)?, Do you own the land used for 0?, Age, Education level, Farm size, Main source of household income, Access to fertilizer, Access to credit, Access to extension services, Training on modern rice technologies, Is labour readily available during peak farming periods?, Are farm inputs available on time in your locality?, Name of Chiefdom (IN CAP).

The findings indicate that the adoption of modern rice technologies in Kambia District is primarily driven by economic and productivity-related factors, alongside institutional and socio-cultural influences, with noticeable variation across chiefdoms (Table 5). Key factors such as the cost of technologies (46.5%), expected increase in yield (48.7%), and availability of improved inputs (43.3%) emerged as the most influential, showing that farmers are largely motivated by profitability but constrained by affordability and access [50]. Factors like access to credit (34.3%) and availability of labour (34.9%) also play important but less consistent roles, reflecting financial and labour challenges [51]. Institutional and social influences, including extension services (37.5%) and peer learning from fellow farmers (32.4%), significantly enhance adoption, especially in Robana and Rokupr, emphasizing the importance of information sharing and advisory support [52][53]. In contrast, chiefdoms such as Kasiri and Kychon recorded lower influence across several factors, suggesting disparities in access to resources and services. Meanwhile, education level and farming experience showed moderate influence, acting as supportive rather than primary drivers. Market availability presented mixed effects across locations. Overall, adoption is shaped by a combination of affordability, expected benefits, input access, and institutional support, and can be improved through targeted policies such as subsidies, better input distribution, strengthened extension services, rural financing, and farmer training to address local inequalities [54][55]. The qualitative findings revealed that “limited access to modern agricultural tools and inadequate technical knowledge are major factors influencing farmers’ decisions regarding the adoption of modern rice technologies. Respondents explained that many farmers are unable to adopt improved technologies because they lack the necessary support, equipment, and practical training required for effective utilization. Farmers further stated that inadequate knowledge on the operation and management of modern technologies, particularly in large-scale rice farming, discourages adoption and reduces confidence in using improved farming methods”.

Table 5. Major factors influencing adoption of modern rice technologies.

The logistic regression analysis identified key factors of modern rice technology adoption among farmers. Ownership of land significantly increased the likelihood of adoption, with farmers who owned their land being approximately 2.85 times more likely to adopt modern technologies compared to those who did not (OR = 2.854, p = 0.005) (Table 6). Farm size also emerged as a strong positive predictor, with larger farms associated with a markedly higher likelihood of adoption (OR = 7.329, p < 0.001) [54][56][57]. Training on modern rice technologies substantially enhanced adoption, with trained farmers nearly seven times more likely to implement these technologies than untrained farmers (OR = 6.659, p < 0.001) [56][58]. Geographic location, represented by the chiefdom of residence, was another significant factor, with farmers in certain towns exhibiting a 2.76-fold higher probability of adoption (OR = 2.757, p = 0.023). In contrast, limited access to credit (OR = 0.110, p = 0.047) and extension services (OR = 0.308, p = 0.021) were negatively associated with adoption, indicating that inadequate financial and technical support can constrain uptake [58][59]. Other variables, including age, education level, household income source, labor availability, and timely access to inputs, were not statistically significant predictors.

Table 6. Logistic regression analysis of factors of modern rice technology adoption.

Note: Adoption of modern rice technology level “1” coded as class 1.

Thresholds can vary, Landis and Koch (1977) [60] came up with the widely accepted benchmarks for interpreting the strength of agreement. These are:

0.00≤w < 0.20—Slight agreement;0.20≤w < 0.40—Fair agreement;0.40≤w < 0.60—Moderate agreement;0.60≤w < 0.80—Substantial agreement;w≥0.80—Almost perfect agreement.

The Kendall’s W ranking results for Kambia District show a moderate level of agreement among respondents (W = 0.42) regarding the key factors influencing the adoption of modern rice technologies, and the relationship is statistically significant (p = 0.002; χ² = 16.79) (Table 7). Among the factors, availability of improved inputs (mean = 3.80) is ranked as the most important factor, followed by the cost ofmodern rice technologies (mean = 4.70), confirming that access and affordability are the primary drivers of adoption [54][61]. Social and market-related factors such as influence of fellow farmers and market availability (mean = 5.20 each) also rank highly, highlighting the role of peer learning and market access in shaping farmers’ decisions [62]. Institutional support through extension services (mean = 5.50) and financial access (credit facilities, mean = 5.60) are moderately influential, while farming experience (mean = 5.30) plays a supportive role [63]. In contrast, expected increase in yield (mean = 6.10), availability of labour (mean = 6.20), and especially level of education (mean = 7.40) are ranked lower, suggesting they are less decisive factors in adoption decisions [64]. Overall, the findings indicate that practical access to inputs, affordability, and social influence are more critical than personal characteristics in determining the uptake of modern rice technologies in the district [65][66].

Table 7. Kendall’s W ranking for major factors influencing adoption of modern rice technologies.

The chi-square test of independence result was conducted to examine the association between the type of rice cultivated and the main source of household income. More than half of the expected counts were below 5, and because there is an assumption that no more than 20% of expected counts should be below 5, fisher’s exact test for p-value was done to validate chi-square p-value for Kambia District indicate that there is no statistically significant association between the type of rice cultivated and the main source of household income, χ² (12, N = 312) = 12.12, p = 0.436, and fisher’s p = 0.253 (Table 8). This implies that farmers’ choice of rice variety, whether ROK 10, NERICA, traditional, or deep water rice, does not significantly depend on whether their primary income comes from rice farming, livestock, non-farm activities, or other crops. Although the majority of respondents whose main income is rice farming predominantly cultivate ROK 10 and traditional varieties, this pattern is consistent with overall production trends rather than income source differences. Similarly, farmers engaged in non-farm activities or livestock also cultivate a mix of rice types, though in smaller numbers. Overall, the findings suggest that rice variety selection is influenced more by factors such as access to inputs, agro-ecological conditions, and technology availability rather than household income source, highlighting the need to focus on improving input access and extension support rather than targeting farmers based on income categories [67]-[69].

Table 8. Chi-square test on the type of rice cultivated associated with main source of household income.

The chi-square test of independence results was conducted to examine the association between farmers’ main source of household income and their access to improved rice seed varieties. More than half of the expected counts were below 5, and because there is an assumption that no more than 20% of expected counts should be below 5, fisher’s exact test for p-value was done to validate chi-square p-value for Kambia District reveal a strong and statistically significant association between farmers’ main source of household income and their access to improved rice seed varieties χ² (9, N = 312) = 36.49, p < 0.0001, and fisher’s p ≤ 0.0001 (Table 9). This indicates that access to improved seeds varies considerably depending on income sources. Farmers whose primary income is rice farming constitute the largest group with regular access (147 respondents), although a substantial number (109) still report no access, reflecting persistent inequalities even within this group. In contrast, those engaged in non-farm activities show more varied but generally limited access, while farmers relying on livestock or other crops have very small sample sizes but tend to have either regular access or none at all. The significant relationship suggests that income source influences farmers’ ability to obtain improved seeds, likely due to differences in financial capacity, market engagement, and institutional support [70][71]. The findings highlight the need for particularly for farmers with limited or diversified income targeted interventions to improve equitable seed distribution, sources.

Table 9. Chi-square test on the access to improved rice seed varieties associated with main source of household income.

The post-hoc test results provide deeper insight into the differences in access to improved rice seed varieties across household income groups. After applying the Bonferroni correction (α = 0.003125), only a few pairwise comparisons remain statistically significant (Table 10). Notably, there are significant differences between “Rarely” and non-farm activities (p < 0.001), as well as “Rarely” and rice farming (p < 0.001), indicating that farmers relying on non-farm income or rice farming differ meaningfully in how infrequently they access improved seeds [72][73]. Similarly, “Yes, occasionally” versus non-farm activities (p < 0.001) shows a significant difference, suggesting that farmers engaged in non-farm activities have distinct patterns of occasional access compared to other groups. However, most other comparisons, including those involving regular access are not statistically significant after correction, indicating broadly similar patterns across income groups. The findings suggest that while some disparities exist, particularly for limited or occasional access to improved rice seeds is generally uneven but not consistently differentiated across all income sources, pointing to systemic constraints in seed distribution rather than differences driven solely by livelihood type.

Table 10. Post-hoc test on the access to improved rice seed varieties associated with main source of household income.