Adoption Drivers and Adopter Categories of Improved Rice Seeds in Northern Ghana

Introduction

Besides non-adopters, Rogers (2005) categorized adopters of innovations (technologies) into five main groups, on the basis of their innovativeness. In each adopter category (innovators, early adopters, early majority, late majority and laggards), indi¬viduals are similar in terms of their innovativeness [1, 2]. That is, the extent to which individuals in the society adopt new ideas compared to other members of the same system [3]. People ran¬domly choose to reject, adopt or disadopt innovations. Yet, there is a time period within which rejection, adoption or disadoption is deemed to have occurred. Ideally, it takes about three years for farmers to adopt agricultural innovations and a decade or two to disadopt them, due to abstinence [4, 5]. However, not all agri¬cultural innovations get disadopted due to their obstinate nature [6, 7, 1]. Some are disadopted due to their unsustainability or incompatibility with environmental, technological, institution¬al, sociocultural and socioeconomic conditions of farmers [8, 9, 10]. For example, rice farmers in Ghana and Uganda disadopted Nerica in less than five years after adoption, due to unprofit¬able and unsustainable nature of producing Nerica, relative to farming other rice varieties [11, 12]. Lamptey (2021) penned that a decade is an ample time to decide if an innovation has actually been rejected, adopted, or disadopted, irrespective of the causes. That is because some non-adopters can later become adopters and some adopters can also become disadopters, within the decade. It is therefore very prudent to fill such a research gab in adoption literature by categorizing potential adopters of ISRS in northern Ghana into early adopters, late adopters, con¬tinuous adopters, temporary disadopters, permanent disadopters, and non-adopters. The IRS under study were all technological inventions (innovations) of the Centre for Scientific and Indus¬trial Research (CSIR) of Ghana and her affiliates, promoted for adoption in northern Ghana by 2009. As at 2019, some of these IRS were still being adopted by farmers while others were either rejected or abandoned for diverse reasons [13].

Adoption is the decision of full use of an innovation as the best course of action available while rejection is a decision not to adopt an innovation [9, 10]. Adoption is a binary dependent variable. Many adoption researchers estimate univariate models such as logit, probit or Tobit to analysis adoption [14-17]. This study also estimated a logistic regression model to analyse the drivers of IRS adoption in northern Ghana. The study then cate¬gorized the potential adopters into early adopters, late adopters, continuous adopters, temporary disadopters, permanent disa-dopters, and non-adopters, using frequencies and percentages. That helped to achieve the purpose of this study, based on Rog¬ers (2005) Innovation Diffusion Theory (IDT).

Theoretical and Empirical Reviews

The Innovation Diffusion Theory:- Mateos and Dadzie (2014) noted that there is no unified “theory” of innovation disadoption because studies on adoption of improved technologies abound but little evidence exists on the continued use of improved tech¬nologies. This lack of continued use of technologies leads to dis¬adoption of agricultural innovations or improved technologies. However, Rogers (2005) Innovation Diffusion Theory (IDT) is pivotal in any adoption or disadoption research. Diffusion is the process in which an innovation is communicated through certain channels over time among the members of a social system [3, 9]. Diffusion is a social process that involves interpersonal com¬munication relationships between researchers, communication agents and users of innovations. Diffusion occurs when individ¬uals in the society adopt an innovation. The main purpose of the IDT is to understand the adoption of innovation in terms of four elements, including the innovation, communication channels, time frame and social system.

The innovation is the new idea, skill, process, tool, technique, practice, product, invention or project to be adopted (used) by individuals in the society. It may have been invented a long time ago, but if individuals perceive it as new, then it may still be an innovation for them. For example, organic farming is considered as an innovation in Europe while mechanization and the use of synthetic fertilizers or agrochemicals are European indigenous methods of farming. However, the reverse is the case here in Ghana and Africa [18]. Hence, a prominent hindrance to the adoption of innovations is uncertainty. It means adoption and diffusion will occur very well if uncertainty about the innova-tions and other hindrances are minimized or removed. There is therefore no doubt that the improved rice seeds (IRS) of CSIR and her affiliates fall into the category of new inventions for use by farmers and consumers. The majority of rice farmers in the study area are smallholders who produce rice to feed their fami¬lies and sell the surpluses for income [19, 20].

The communication channels are the media through which the information about innovation passes to its intended beneficia-ries. They include television, radio, telephone, internet, news-paper, magazines, and person to person or group to groups of individuals, among others. The social system is the platform, society, community or the environment in which adoption and diffusion occur. The time frame is the period, usually 10 years, within which adoption and diffusion occur [7, 9]. The time di-mension in any diffusion research indicates the reliability of that research. The IDT as well explains five stages in the innovation decision process.

The Innovation Adoption Decision Process

The innovation adoption-decision process involves informa-tion-seeking and information-processing activity, where an in-dividual is motivated to reduce uncertainty about the advantages and disadvantages of that technology. As demonstrated by Rog¬ers (2005), the technology adoption-decision process involves five successive steps, namely knowledge, persuasion, decision, implementation and confirmation. The knowledge stage is where individuals learn about the existence of a new technol¬ogy and seek information about it. The persuasion occurs when an individual has a positive or negative attitude toward the new technology. It does not necessarily lead directly or indirectly to an adoption or rejection of the innovation. The decision stage is where an individual chooses to adopt or reject the new tech¬nology while the implementation stage is where the innovation is put into practice. At the confirmation stage, adopters look for affirmations to help them validate their adoption or rejec¬tion decisions. The decisions can be reversed if the individuals are exposed to contradictory messages about the innovation or maintained if they get validatory messages.

The Innovation Adoption Decisions/Categories

The innovation adoption decisions are basically acceptance and rejection. The acceptance leads to adoption and continuous adoption or discontinuous adoption while the rejection leads to non-adoption and continuous rejection or later adoption. Those who reject an innovation and continue to reject the innovation never accept or use the innovation and are therefore regarded as non-adopters. However, those who initially reject the inno¬vation and accept it later become adopters. Those who initially accept the innovation, use it and continue to do so are continuous adopters or the real adopters. However, adopters who discontin¬ue their adoption decisions become disadopters. The disadoption decisions can be temporary or permanent [21].

The temporary disadoption decisions can lead to re-adoption de¬cisions when the disadopters regret their disadoption decision possibly due to additional information about the innovation to be discovered later. The permanent disadoption decisions have no regrets whatsoever about any vital or additional information to be realized about the innovation after disadopting it. The indi¬viduals therefore live with the consequences of their disadoption decisions, the potential benefits notwithstanding.

It stands to reason that early adopters and later adopters are all considered as adopters of innovations, because they are all aware of the innovation. In this vain, people who are not aware of the innovation cannot be considered as non-adopters.

People who are initially unaware of the innovation but later be¬come aware of it through diffusion and decide to adopt it are considered late/later adopters, not initial adopters, early adopters or laggards. Laggards are potential adopters who intentionally delay their adoption decisions because they are conservatives and risk averse. Therefore, some late adopters can be laggards but not all late adopters are laggards. Hence, initial adopters are innovators and early adopters. Thus, potential adopters can be categorized as early adopters, late/later adopters, continuous adopters, temporary disadopters, permanent disadopters, and non-adopters. These six categorization are synonymous to those of Rogers’ innovators, early adopters, early majority, late major¬ity, laggards, and non-adopters. Hence, it is too elementary to simply categorize potential adopters as adopters and non-adopt-ers, irrespective of whether they maintain their adoption deci¬sions over time or not.

Characteristics of Rice Farmers

Socioeconomic characteristics of farmers such as age of house¬hold head, income, group membership, family size, educational level, farm plot size, and number of contacts with agricultural extension agents affect adoption of modern technologies in rice production [4, 19]. These characteristics of farmers need to be taken into consideration in any adoption or disadoption studies.

Most rice farmers in northern Ghana are males, less educated, older, risk averse, have large farm families, small farm sizes, and are prone to unfavourable environmental conditions due to the harmful effects of climate change. Donkoh and Awuni (2011) opined that when the educational levels of farmers are too low, it would take a lot of efforts to introduce modern technologies to them [22]. That is because educated farmers are more prone to adoption and they have the tendency to co-operate favourably with other farmers [23]. Many rice farmers are unable to attend training sessions on innovation dissemination regularly to get latest information on agricultural innovations. Such farmers rely on their fellow (contact) farmers for information on new farming technologies [24].

Rice farmers’ lack of adequate knowledge of agricultural tech-nologies is normally caused by lack of extension visits to farmers’ farms [25]. Younger rice farmers are also less likely to disadopt IRS compared with aged farmers (Rogers 2005). This is because young people are zealous to adopt innovations and they have many years ahead of them to maintain their adoption decisions. This gives avenues for extension to specifically concentrate on young farmers and use rice to curb food insecurity in the future. Martey (2013), also posits that married household heads have a tendency of adopting IRS better than unmarried male household heads. The reason is that married male household heads are usu¬ally supported by their spouses to produce, process and market rice [26]. Similarly, rice farmers with lager household sizes have the tendency to adopt agricultural innovations because members of their households serve as sources of farm labour.

Study Location and Data

Studied Location, Sample Size and Data Collection:- The study was conducted in northern Ghana, where the improved rice seeds (IRS) were adopted and disadopted. Cluster sampling was used to select four main rice producing zones each in Tolon, Kumbungu and Savelugu Districts but two in Nanton District for the study. Simple random sampling was then employed to select households from each of the selected zones to constitute the total sample.

The vegetation in the area comprises grasslands, shrubs and clusters of perennial trees such as mango, neem, shea, baobab, acacia and other drought tolerant trees. There are two main sea¬sons in a year: rainy and dry seasons. The dry season normally starts from October/November to April/May while the rainy sea¬son commences in June and recedes in September/October each year, as a result of variable climatic conditions in recent times. The annual optimal rainfall stands at 1000mm or 35 inches deep [27]. The region alone records about 68,407.25 metric tonnes of paddy rice yearly, and is responsible for about 37% of rice produced in Ghana [28]. The average paddy rice yield in the region stands at 1.32Mt/ha relative to the national average yield of 3.65mt/ha. The map of the study area is as shown in Figure 1.

The study chose a sample size of 385 households from an esti¬mated population of 1,000,000 IRS farmers in northern Ghana, consistent with Krejcie and Morgan [29].

Figure 1: Map of the study area showing the selected districts and communities

Source: Authors’ construction, 2020

Statistical Description of Data Collected (Used)

Table 1 shows the socio-demographic features of the sampled improved rice farmers. Descriptive statistics of the rice farm¬ers showed that the mean age of improved rice adopters was 40 years, with those of males and females being 41 and 39 years respectively. It means the male adopters were relatively older than their female counterparts. It was also observed that 30% of the improved rice farmers were literate/educated, with the ma¬jority (31%) being males than females (29%). This finding is in tandem with who found that male rice farmers in northern Gha¬na were more educated than their female counterparts [16, 30, 31]. The likelihood of literate farmers adopting new agricultural innovations such as IRS is higher than the illiterates. This is be¬cause the literates are better placed to understand the benefits of the innovation being promoted. Since the illiteracy rate is higher among the farmers, it is very probable that the numbers of the IRS adopters in the region would be lower than expected. Aver-agely, there were 11 members in male headed farm households and 7 people in female headed farm families, with a mean score of 9 persons per household in the region, corroborating GLSS7 (2019).

Similarly, male farm households had larger rice fields (5 acres) than those of females (3 acres), giving a mean farm size of 4 acres in the study area. This is in tandem with Donkoh (2020), MoFA (2017) and Ragasa et al. (2013) finding that most rice producers in Ghana (about 80%) are small-scale farmers, whose farmlands measure less than 2 hectares in size. Besides, 90% of the rice farm household heads were males while 10% were fe¬males, corroborating. Most of the household heads (90%) were familiar with government policies for the rice sub sector of the economy of Ghana. Contrary to our expectations, more female (95%) than male (85%) household heads were aware of such policies. In the same way, the results also indicated that equal numbers (90%) of male and female adopters of IRS in the region were able to access both output and input markets respectively. That is unprecedented since females are more market oriented than males [23, 26].

Moreover, an average of 80% farm household heads compris¬ing 79% males and 81% females were able to utilize extension services, and 30% were able to obtain loans to farm rice, cor¬roborating [25]. This is opposed to Doss et al. (2003) who found that female household heads have less access to external inputs, extension services and information due to socio-cultural values. These notwithstanding, an average of 92% farm household heads (91% males and 93% females) were adequately aware of climate change and its devastating effects on rice production in recent times. They were therefore able to employ appropriate climate change adaptation mechanisms like creating fire belts around their farms during the dry seasons and constructing drains/canals in and out of their rice fields during the rainy seasons, corrobo-rating [32, 33]. Then, a mean of 75% adopters (74% males and 76% females) had good transportation systems, equal numbers (50% each) of the male and female adopters were members of farmer based organizations, while 30% of the farm household heads (29% males and 31% females) used mobile phones to get information for their farming and marketing activities. Zakaria (2019) revealed that farmers’ membership in groups/associa-tions facilitates their adoption of agricultural innovations, for maximum output.

An optimum percentage (80%) of the respondents, which were made up of 85% males and 75% females, did both ploughing and harrowing before planting/sowing rice, corroborating Lamptey et al. (2022) that modernized rice varieties have high input and labour requirements to give optimum yields. Finally, only 10% of the farm household heads (11% males and 9% females) em¬ployed the services of combined harvesters to harvest their rice, chiefly due to constraints with affordability and availability of mechanization services in the region. It means an average of 90% the adopters harvested rice manually using sucks, sickles and tarpaulins, corroborating MoFA (2017) and APS (2015) that majority of rice farmers in Ghana produce rice on small scales using simple farm tools and equipment.

Table 1: Socio-Demographic Features of Improved Rice Seed Adopters

                                                                  Source: Field data, 2020.

Conceptual and Analytical Frameworks

Conceptual framework: Adoption:-Several factors affect each of these adoption decisions of farmers, to adopt or not to adopt, to continue to adopt or to disadopt, to temporarily disadopt or permanently disadopt [34]. The farmers are also concerned about whether to re-adopt innovations they disadopt or maintain their disadoption decisions for such innovations. Farmers’ tem¬porary disadoption decisions lead to re-adoption, when they lat¬er regret their actions [21]. The theories of perceived usefulness and perceived ease of use of technology determine farmers’ mo¬tive to use agricultural innovations whereby the motive to use serves as an intermediary of real innovation adoption [35]. That is because farmers also consume the rice they produce. Farmers’ decision to continue to adopt or disadopt an innovation is deter¬mined by their need to maximize profit or derive the maximum utility from it. Farmers tend to continue the adoption if the ex¬pected utility or gains from persistence use is more than that of abandoning it. The expected satisfaction of a farmer is a matter of both household and farm specific factors [36]. Several other factors affect farmers’ early adoption, late adoption, non-adop-tion, continuous adoption, temporary disadoption (re-adoption),and permanent disadoption decisions. These factors include so-cioeconomic, demographic, institutional, technological and en¬vironmental factors. The complex nature of farmers’ adoption decisions can best be explained descriptively, since no econo¬metric model can adequately analyze the entire phenomenon. Hence, this study estimated a univariate binary response model (logit) to analyze the drivers of adoption, and used percentages to categorize the various groups of adopters.

Analytical Framework: The Logistic Regression Model

Logistic (logit) regression model was estimated to analyze adop¬tion drivers of IRS in northern Ghana. The binary response in the model was adoption or non-adoption, where Y denoted a random binary variable that took the value of 0 or 1. The 0 then denoted non-adoption and 1 denoted adoption. 1 represented the socio-economic, demographic, farm level, technological, in¬stitutional and environmental characteristics of the farmers rel¬ative to adoption of IRS [37, 38]. The logistic model specified the likelihood that Y=1, leading to the values of X1, expressed econometrically as follows:

The binary response model, as a regression model, was written as: yi = 1 - f(Xi-β) + εi with yi as the dependent variable denoting adoption of IRS; Xi as adoption drivers of the IRS and εi as the residue representing the deviation of the binary from its condi- tional mean. In conformity to Cameron and Trivedi (2010) and Bayard et al. (2006), the empirical model specified to analyze farmers’ adoption decisions of IRS was:

Where Pi represents the likelihood of farmers adopting improved rice varieties and {Pi / (1-Pi)} is the uneven ratio in favour of adoption and X_i denotes the independent variables influencing adoption. εi Is the error term and β stands for the logistic coeffi¬cient of the independent variables.

Definitions of variables and their a-priori expectations

The Table 2 illustrates the definitions of variables and their a-pri-ori expectations relative to adoption of IRS. Here, adoption is considered as a binary variable and denoted by 1, if a farmer was an adopter and 0, if otherwise. Any farmer who did not cultivate IRS for a minimum of three years, from 2009 to 2019 cropping seasons, was considered a non-adopter [39].

Table 2: Adoption Variables and their a-priori Expectations of Improved Rice Farmers

                                                   Source: Author’s construction, 2022

Discussions and Conclusions

Adoption and non-adoption levels of improved rice seeds:- In-formation on Table 3 shows that the mean adoption level of the improved rice seeds (IRS) in northern Ghana is 51.95%. It means about 48.05% of the farmers in the study area are non-adopters of IRS. That necessitates the use of a binary regression model (logit) in analyzing the drivers of adoption, consistent with Don¬koh and Awuni (2011), Bruce et al. (2014), Abdallah (2016), and Lamptey (2022), as shown on Table 4.

Analysis of adoption and disadoption as complex phenomena are discussed in other studies by Lamptey et al. (2021), Azu-mah et al. (2022), Donkoh (2020), Donkoh et al. (2019) and Azumah (2019); using Generalized Multivariate Regression, descriptive statistics, Endogenous Switching Poison Regression, and Multivariate Probit approaches respectively. Lamptey et al. (2022) also used the Propensity Score Matching (PSM) model to analyze the effect of adoption of improved varieties on rice productivity in the Northern Region of Ghana. They found that adoption of improved rice varieties had a high and positive effect on rice productivity of farm households in the region.

Five different approaches were involved in the PSM model.First, Logistic Regression (logit) was used to analyze factors that affected adoption of improved rice varieties. Secondly, a histo¬gram was employed to check for overlaps and common supports in the propensity score distribution. Thirdly, a test was carried out to analyze the propensity score of the variables in the mod¬el, by categorizing adopters and non-adopters respectively as treated and control groups. Then, an overall quality test of fac¬tors affecting adoption was conducted both before and after the matching, by treating adopters and non-adopters as unmatched and matched samples in the study. Finally, an Average Treat¬ment Effect (ATE) model was employed to estimate the effect of improved rice varieties adoption on rice output among the farm households. All these tests and models used indicate the robust and complex nature of the phenomenon of adoption (and disa¬doption). Hence, analyzing adoption as binary requires the use of logit as a univariate model. However, to treat the same subject as multifaceted concept would demand more robust models like Generalized Multivariate Regression, Propensity Score Match¬ing, Endogenous Switching Poison Regression, Multivariate Probit, and Heckman Two-Stage Models, among others, which is not the focus of the present study.

                                Table 3: Adoption Levels of Improved Rice Seeds in Northern Ghana

                               Source: Field data, 2020. N = 385 Mean adoption level = 51.95

Adoption drivers of improved rice seeds

Improved rice seeds (IRS) modelled in this study are Jasmine-85, Nabogu, Nerica-1, GR-18, Digang and Mandii. Jasmine-85 and Nerica-1 happen to be among the latest to be promoted in the study area while the rest were selected from older varieties under cultivation, corroborating Ragasa et al (2013), MoFA (2017) and Lamptey et al. (2022).

Outcome of the logistic regression analysis in Table 4 showed that age of household head, level of education, household size, access to extension services, access to production credit, and harvesting method did not affect adoption of IRS in northern Ghana. These findings contradict those of Donkoh et al. (2019) and Martey et al. (2013) but are in tandem with Abdallah, (2016) and Bruce et al. (2014). Donkoh et al. (2019) also found an in-significant association between extension service delivery and adoption of agricultural technologies, due to the low extension to farmer ratio in Ghana. However, ownership of mobile phones, membership in farmers based organizations, and access to input markets positively affected adoption of IRS at 1%. Donkoh et al. (2019) also found a positive relationship between farmers’ membership in FBOs and adoption of agricultural technologies. Farmers’ awareness of climate change (rainfall perception), gov¬ernment policies and total farm plot size also positively affect¬ed adoption of the IRS modelled, corroborating Donkoh et al. (2019).

However, gender and mechanization negatively affected adop¬tion at 5%. This implies that female farm household heads have the tendency of adopting agricultural innovations more than their male counterparts, contrary to the findings of Zakaria et al (2019) and Martey et al. (2013). It also means that farmers who do not practice mechanization are more likely to adopt the innovations better than other farmers who practiced mechani¬zation. This is contrary to the a-priori expectations of this study because rice cultivation essentially requires mechanization (till¬age). Besides, the nature of soils in the study area is such that no meaningful arable crop farming can take place in the absence of mechanization, except by the means of simple farm tools such as hoes and cutlasses for tillage. This finding confirms the fact that about 30% of rice farmers in northern Ghana cultivate essential¬ly for subsistence purposes (Azumah 2019, APS 2013). It shows that adopters of agricultural innovations in northern Ghana are not necessarily large scale or commercial farmers who have ac¬cess to adequate mechanization services.

Access to input market positively affected adoption at 10% while that of road network negatively influenced adoption at 10%. These imply that farmers who have access to input market are able to buy IRS, fertilizers and agrochemicals to enhance their adoption decisions. On the other hand, lack of good roads in the region hindered adoption of IRS because access to good roads is an incentive to adoption (MoFA 2017, APS 2015). Fi¬nally, the adoption constant was statistically significant at 1%, but negative, meaning the tendency for the farm household heads to reject the IRS was prevalent in the region, due to other extraneous factors.

Table 4: Logistic regression of factors affecting adoption/adopters of improved rice seeds

                  *Significant at 10%       ** Significant at 5%      ***Significant at 1%

                                                                      Source: Field data, 2020.

Six Adopter Categories of the Improved Rice Seeds

Below was an attempt to categorize potential adopters of IRS in northern Ghana into six groups. Namely; early adopters, late adopters, continuous adopters, temporary disadopters, perma-nent disadopters, and non-adopters. The reason being that, the general categorization of potential adopters into adopters and non-adopters does not elucidate the diverse shades of real adopt¬ers (Lamptey 2021). Farmers are by nature very selective. They pick and choose aspects of innovations that suit them most and they continue to do so as long as they derive satisfaction from those they have chosen to adopt (Mateos and Dadzie 2014, Doss 2006). They disadopt the innovations they are not comfortable with and they may re-adopt them if they later find them useful. They may entirely abandon them for other innovations, if they no longer derive maximum utility from their continuous adop¬tion or usage (Donald and Parker 2012). Hence, this study clas¬sified potential adopters into six distinct categories: early adopt¬ers, late adopters, continuous adopters, temporary disadopters (re-adopters), permanent disadopters (deserters) and non-adopt-ers, as presented in Table 5.

Early adopters are farmers who adopted the IRS as soon as they were released/promoted in the region, particularly from 2010 to 2014. Late adopters are those who accepted and used the seeds later on, especially from 2015 to 2019, through diffusion. Con¬tinuous adopters refer to only farmers who utilized the seeds continuously from 2010 to 2019 while non-adopters are farmers who never accepted or used the seeds in that decade. Disadopt¬ers refer to farmers who ever adopted the innovations (either from 2010 to 2014 or from 2015 to 2019) but abandoned them along the line, with or without any intentions of adopting them again. Disadoption is therefore abandonment of an innovation after adopting it. Temporary disadopters (re-adopters) on the other hand refer to only farmers who ever adopted the innova¬tions, stopped adopting them for some time, went back to adopt them again and intend doing so until further notice. Permanent disadopters on the other hand are disadopters who have no in¬tentions of ever re-adopting the IRS (Rogers 2003). These clas¬sifications are in tandem with Donald and Parker (2012) who found that disadoption can be a permanent or temporary change but the aftermath reaction to it may be positive and reinforcing or negative and regretful.

They opined that temporal disadoption can result in re-adoption when individuals regret their disadoption decisions or when they are able to overcome the challenges associated with continuous adoption. It means there could be temporary disadopters or per¬manent disadopters. The outcomes of disadoption can therefore be cumulative and involving rather than dichotomous (Mateos and Dadzie 2014, Donald and Parker 2012, Doss 2006).

It is therefore very imperative to re-group potential adopters into early adopters, late adopters, continuous adopters, tem-porary disadopters (re-adopters), permanent disadopters, and non-adopters. The results as shown in Table 3 reveal that the highest continuous adoption level was about 15%, meaning only about 15% of farmers in the region are continuous adopters of IRS. This is consistent with Lamptey (2022) and Kijima et al. (2011) that adoption rates of agricultural innovations in Ghana and across Africa are generally low. The adoption rates in this sense refer to the numbers (percentages) of farmers who adopted the IRS in the study area during the period under review [40].

Table 5: Six Categories of Improved Rice Seeds Adopters

N = 385; ** = 100%;    T* = Temporary; P* = Permanent; F = Frequency; % = Percentage

Summary and Conclusions

This study was carried out to analyze drivers of adoption and categorize potential adopters of improved rice seeds in northern Ghana into six specific groups. Using a survey data collected from 385 rice farm households, the study employed logistic re¬gression to account of the factors that affected adoption of im¬proved rice seeds in northern Ghana and accordingly used fre¬quencies and percentages to categorize the potential adopters into early adopters, late adopters, continuous adopters, temporal dis¬adopters (re-adopters), permanent disadopters, and non-adopt-ers. That categorization helped to specifically group potential adopters into six, rather than the two fundamental groups of adopters and non-adopters, irrespective of whether they main¬tained their adoption decisions over time or not. The mean level of adoption was 51.95%. Meaning, about 52% of rice farmers in northern Ghana adopted IRS. That had policy implications to ensuring that adoption levels of IRS in northern Ghana rose drastically above average. Outcome of the logistic regression also showed that positive drivers of IRS in the region included farmers’ awareness of government policies for rice, knowledge of climate change, ownership of mobile phones, membership in farmer based organizations, total farm plot size, and access to input markets. However, gender, road network, and mechaniza¬tion negatively affected rice farmers’ adoption decisions of IRS northern Ghana. The study recommends that the government collaborates with Non-Governmental and Farmer-Based Orga¬nizations to encourage rice farmers to continuously adopt IRS in northern Ghana, via interventions like the Planting for Food and Jobs program. Also, female rice farmers in northern Ghana, should be encouraged to embark on mechanization and commer¬cialization of their farming enterprises. Besides, the government should construct motorable roads to the farming communities. Finally, adoption researchers should use this study as a guide when categorizing adopters of agricultural innovations in Ghana and elsewhere.

References

1. Doss, C. R. (2006). Analyzing technology adoption using microstudies: limitations, challenges, and opportunities for improvement. Agricultural economics, 34(3), 207-219.
2. Van Braak, J. (2001). Individual characteristics influenc­ing teachers' class use of computers. Journal of educational computing research, 25(2), 141-157.
3. Rogers, E. M. (2010). Diffusion of innovations. Simon and Schuster.
4. Lamptey, C. Y., Donkoh, S. A., Sulemana, N., Zakaria, A., & Azumah, S. B. (2022). Drivers of initial adoption, con­tinuous adoption and dis-adoption of selected rice varieties among rice farmers in Northern Ghana. Ghana Journal of Agricultural Science, 57(2), 76-95.
5. Kasirye, I. (2013). Constraints to agricultural technology adoption in Uganda: evidence from the 2005/06-2009/10 Uganda national panel survey.
6. Lamptey, C. Y., Sulemana, N., Donkoh, S. A., Zakaria, A., & Azumah, S. B. (2022). The effect of adoption of improved varieties on rice productivity in the Northern Region of Ghana.
7. Lamptey, C. Y. (2021). ADOPTION AND DISADOPTION OF IMPROVED RICE VARIETIES AMONG FARMERS IN THE NORTHERN REGION, GHANA (Doctoral disser­tation).
8. Guizar-Mateos, I., & Dadzie, N. (2014). Financial Ser­vices and Divisible Technology Dis-adoption among Farm Households: Theory and Empirical Application Using Data from Ethiopia (No. 329-2016-13086).
9. Wonglimpiyarat, J., & Yuberk, N. (2005). In support of innovation management and Roger's Innovation Diffusion theory. Government Information Quarterly, 22(3), 411-422.
10. However, Rogers (2005) failed to give the time frame for such categorizations, giving room for speculations (Mateos and Dadzie 2014, Doss et al. 2003).
11. Lamptey, C. Y. (2022). Adoption Rates of NERICA Inno­vation among Rice Farmers in Northern Ghana. Journal of Experimental Agriculture International, 1-12.
12. Kijima, Y., Otsuka, K., & Sserunkuuma, D. (2008). Assess­ing the impact of NERICA on income and poverty in central and western Uganda. Agricultural Economics, 38(3), 327-337.
13. Lamptey, C. Y. (2021). ADOPTION AND DISADOPTION OF IMPROVED RICE VARIETIES AMONG FARMERS IN THE NORTHERN REGION, GHANA (Doctoral disser­tation).
14. Donkoh, S. (2020). Rice commercialisation and improved agricultural technology adoption in Northern Ghana: En­dogenous switching Poisson approach. Alanya Akademik BakıÅ?, 4(1), 105-121.
15. Donkoh, S. A., Azumah, S. B., & Awuni, J. A. (2019). Adoption of improved agricultural technologies among rice farmers in Ghana: A multivariate probit approach. Ghana Journal of Development Studies, 16(1), 46-67.
16. Bruce, A. K. K. (2015). IMPROVED RICE VARIETY ADOPTION AND ITS EFFECTS ON FARMERS’OUT-PUT IN GHANA (Doctoral dissertation).
17. Asante, E. G., Appiah, M. R., Ofori-Frimpong, K., & Afrifa,A. A. (2004). The economics of fertilizer use on some peas­ant cocoa farms in Ghana. Ghana J Agric Sci, 33, 183-90.
18. Donkoh, S. A. (2022). AFRICAN AGRICULTURE AT A CROSSROAD: CAN FARMERS SOCIOECONOMIC IN­DICATORS GIVE US DIRECTION?.
19. Amanor-Boadu, V., Zereyesus, Y., Ross, K., Ofori-Bah, A., Adams, S., Asiedu-Dartey, J., ... & Salin, M. (2015). Ag­ricultural production survey for the Northern Regions of Ghana: 2013-2014 Results. Final Report. METSS/USAID Ghana.
20. Ragasa, C., Dankyi, A., Acheampong, P., Wiredu, A. N., Chapoto, A., Asamoah, M., & Tripp, R. (2013). Patterns of adoption of improved rice technologies in Ghana. Interna­tional Food Policy Research Institute Working Paper, 35(2), 6-8.
21. Donald, L. and J. Parker 2012. ‘Disadoption, in NA - Ad­vances in Consumer Research’. Eds. Zeynep Gürhan-Canli, Cele Otnes, and Rui (Juliet) Zhu, Duluth, MN: Association for Consumer Research, (40), 313-317.
22. Donkoh, S. A., & Awuni, J. A. (2011). Adoption of farm management practices in lowland rice production in North­ern Ghana.
23. Wiredu, A. N., Asante, B. O., Annin, K., Dogbe, W., Attoh, C., & Al-Hassan, R. M. (2013). Factors Influencing Partic­ipation in Rice Development Projects The Case of Small­holder Rice Farmers in Northern Ghana.
24. Etwire, P. M., Martey, E., & Goldsmith, P. (2021). Factors that drive peer dissemination of agricultural information: evidence from northern Ghana. Development in Practice, 31(5), 606-618.
25. Zakaria, A., Azumah, S. B., & Donkoh, S. A. (2019). Wel­fare effects of livelihood diversification of farm households in northern ghana: a quantitative approach. UDS Interna­tional Journal of Development, 6(3), 214-226.
26. Modzakah, D., & Angelucci, F. (2019). Analysis of price incentives for rice in Ghana 2005-2013. Gates Open Res, 3(301), 301.
27. MoFA (Ministry of Food and Agriculture) 2020. 2019 An­nual Report on Rice Farmers in Tolon, Kumbungu, Savelu­gu and Nanton Districts, Northern Region, Ghana. January, 2020.
28. Ministry of Food and Agriculture (MoFA). (2017). Plant­ing for Food and Jobs: Strategic Plan for Implementation (2017–2020)..
29. Smith, S.M. 2019. ‘Determining Sample Size, How to En­sure You Get the Correct Sample Size’. Available at www. qualdrics.com. (Accessed November 19, 2019).
30. GLSS7 (Ghana Living Standard Survey Round 7) 2019.Main Report. Ghana Statistical Service, June 2019, Ghana.
31. Azumah, S. B. (2019). Agricultural technology transfer, adoption and technical efficiency of rice farmers in North­ern Ghana (Doctoral dissertation).
32. Azumah, S. B., Donkoh, S. A., & Ansah, I. G. K. (2017). Contract farming and the adoption of climate change coping and adaptation strategies in the northern region of Ghana. Environment, Development and Sustainability, 19(6), 2275-2295.
33. Lolig, V., Donkoh, S. A., Obeng, F. K., Ansah, I. G. K., Jasaw, G. S., Kusakari, Y., ... & Kranjac-Berisavljevic, G. (2014). Households’ coping strategies in drought-and flood-prone communities in Northern Ghana. Journal of Disaster Research, 9(4), 542-553.
34. Soto, I., Achten, W. M., Muys, B., & Mathijs, E. (2015). Who benefits from energy policy incentives? The case of jatropha adoption by smallholders in Mexico. Energy Poli­cy, 79, 37-47.
35. Samaradiwakara, G. D. M. N., & Gunawardena, C. G. (2014). Comparison of existing technology acceptance the­ories and models to suggest a well improved theory/model. International technical sciences journal, 1(1), 21-36.
36. Soto, I., Ellison, C., Kenis, M., Diaz, B., Muys, B., & Mathi­js, E. (2018). Why do farmers abandon jatropha cultivation? The case of Chiapas, Mexico. Energy for Sustainable De­velopment, 42, 77-86.
37. Cameron, A. C., & Trivedi, P. K. (2009). Microeconomet-rics using stata stata press. College Station, Texas.
38. Bayard, B., Jolly, C. M., & Shannon, D. A. (2006). The adoption and management of soil conservation practices in Haiti: the case of rock walls. Agricultural Economics Re­view, 7(389-2016-23356), 28-39.
39. Doss, C. R., Mwangi, W. M., Verkuijl, H., & De Groote, H. (2003). Adoption of maize and wheat technologies in east­ern Africa: a synthesis of the findings of 22 case studies. CIMMYT.
40. Sahin, I. (2006). Detailed review of Rogers' diffusion of in­novations theory and educational technology-related stud­ies based on Rogers' theory. Turkish Online Journal of Edu­cational Technology-TOJET, 5(2), 14-23.