The Impact of Key Stakeholders and the Computer Skills of Librarians on Research Data Management Support Services (Id so-21-1893.r2)

This study aims to investigate key stakeholders’ impact and to assess its effect on Research Data Management (RDM), thereby setting the agenda for sound local government Research Data Management (RDM) policy for maximum profit. This study relies mainly on the stakeholder’ theory. Our target population is stakeholders in RDM. The survey used the Partial Least Squares and Structural Equation Modeling tool with a sample of 295 received from respondents to find the effect among the stakeholders. The Statistical results confirmed there is a positive effect among stakeholders. The Funders agencies criteria influence researchers’ work. The effect of government support policies on librarians RDM support services was positive but not significant. The impact of computer literacy of librarians on the RDM support service delivered by librarians was insignificant. The finding is mainly due to overreliance on manual approach to work in most libraries in Ghana as a result inadequate fund support. Librarians need to take a leading role in RDM to implement RDM policy fully. This study contributes to the discourse by helping to strengthen the existing relationship and also to understand the role of various stakeholders in RDM. Stakeholders in Ghana and developing countries need to take greater responsibility to achieve success. Government can rely on it for policy formulation. The research can be used to expand the existing stakeholders’ effective support and interactions as well as in teaching. The survey results can also help identify best practices, improve current observation, and compare practices across different countries.


Introduction
Research Data Management (RDM) is a comprehensive set of activities for the organization, storage, access, and Plan (DMP) for research they sponsored as a condition for accessing funds.Its purpose is generally to enable research data to be accessible and shared.The institutions also provide data management resources, infrastructure, training, and enabling environments for others to perform their roles.Researchers, on their part, design the study, specify how data will be collected, analyzed and finally draw conclusions.The government provides policy direction and implement research findings.Academic librarians cannot be left in this task as they have the experience that could be translated into their new role, such as DMP support, archiving, preservation, and other support services.The success of RDM depends on the interaction among key players such as the librarians providing RDM support services, government with support policies, the funders or donors with agency criteria to access funding, institution with support services with the purpose to affect researchers' work (Whitehead & Bourne-Tyson, 2016).Their decisions can affect positively or negatively the research data management agenda.Following the reluctance of some of these stakeholders, especially the researchers and librarians, to keep to best practices in handling research data, there is a clamoring call for educational institutions, funders, and government to get actively involved in the management of the research data (Stall et al., 2019).
Computer literacy is a survival skill in the information age, and it is finding more mutual ground with other literacies.It has been described as literacy with digital texts.Computer literacy is viewed as a general knowledge of both software and hardware.A broad knowledge may not qualify one to be computer literate but rather the ability to run and navigate through the application.It was discovered by Baro et al. (2019) that, to be computer literate or fluent, there should be grounded contemporary skills, and a foundational concept, and intellectual capabilities.There is a need to have basic knowledge of communication relating to applying computer technology to solve problems.According to Binici (2021) technology significantly affects the librarians 'skills and competence.Therefore, there is no doubt that all stakeholders especially librarians need these skills to assist researchers in this technology world for better communication and management.Computer literacy should include among others skills that will enable them to store, share, preserve, and archive research data.
Over the years, many researchers have sought to understand the dynamism of the various stakeholders' support and challenges in RDM to find the best practices that will serve the researchers' purpose (Flores et al., 2015;S ˇuman et al., 2020;Tang & Hu, 2019) and (Tang & Hu, 2019).The authors emphasize that despite the challenges that the data in research is relevant and assert that managing research contributes to quality research and the relevance of RDM is not in doubt.Researchers have been encouraged to embrace the practice to derive the full benefit.For example, RDM policies increase work outputs, make research results verifiable, and promote replication of the research findings.Burgeoning literature shows that the stakeholders benefit more if they participate in policy formulations for sustainable competitive advantage in funded research.Despite this, librarians generally were reluctant to engage in RDM (Wilkinson et al., 2018).Besides that, the discussions on data management in developing countries are still at a very early stage and lagged in RDM efforts (Chawinga & Zinn, 2019).
In the context of Ghana, the practice of research data management is still new and underdeveloped compared to other developed countries.Many research data are not available and, therefore, not reusable as in developed countries such as the USA and UK (Arkorful et al., 2021).There is, therefore, no research data management policy.In a study conducted by Benneh et al. (2016) to better understand data management, it was found that some level of research data management practices exist.However, the universities have no RDM support function.The phenomenon is far from overarching.It is also reported that there is no research data management policy at the University of Ghana.The current RDM at the university is not yet fully developed (Avuglah & Underwood, 2019).The situation of Ghana is not different from many other countries especially in Africa.In South Africa for example, despite the directive of the National research foundation directive to all researchers to make data accessible, there is no compliance, as there are still challenges with access, sharing, retention, preservation, and storage of data, Factors impeding the research data management in developing countries, include lack of time and data misuse, absence of proper RDM training, reward and legal framework, cope with weak regulation.It was also reported lack of data infrastructure and interoperability challenge (Anane-Sarpong et al., 2018).
According to Liu et al. (2020), in the USA, UK, and Australia, data sharing, data management plan, access, and retention are critical requirements from funders as conditions to access research grants.The authors found that there is no difference in these countries regarding RDM.In Addition, there are research data management services to help the researchers to comply with the funder's requirement.Larivie`re and Sugimoto (2018) found that 90% of research funded in US National Institute of Health are freely available.In addition, there are research data management services to help the researchers to comply with the funder's requirement (Tang & Hu, 2019).These services are provided mainly by librarians unlike in Africa as such Ghana.It is a phenomenon practiced worldwide, especially in developed countries which is reflected in their development.The aforementioned requirements from the developed countries can be extrapolated and serve as benchmark in the case of Ghana and other country to derive the maximum benefit.
Extant literature did not test the effect of the stakeholders in RDM and their effect on research work.This paper investigates stakeholders' support effect in Ghana.The absence of RDM may impede the urgency for sound RDM practices, as seen in other countries.

Stakeholders and Their Relationship
RDM involves several stakeholders at all levels, such as journal publishers and funding agencies (Chen & Wu, 2017).Others are institutions, librarians, government, researchers.Previous findings in Hall et al. (2013) show that the stakeholders benefit more if they participate in policy formulations for a sustainable competitive advantage of the institutions in funded research.In Darch et al. (2021), it was found that comparing motivation to satisfaction with the project outcome gives a valuable lesson for the success of RDM through partnership.However, opinions are divided among the management of many institutions.While some see the relevance of participating in RDM, others remain adamant that investments into RDM are not competitive.The standpoint group also thinks that not all institutions need DMP (Vilar & Zabukovec, 2019).While cost could be a challenge, it may not be the only reason that handicaps the implementation of RDM (Tenopir et al., 2017).Previous research found conflict and tension among these stakeholders in various institutions (Verbaan & Cox, 2014).In addition, the level of mutual understanding was found to be low among the stakeholders sampled.Despite the attempt to resolve this by providing division of role, it remains doubtful.If this is not resolved, it can impede the success of RDM.In Whyte (2014) librarians and research offices show commitment by coordinating and promoting the RDM policies.There is also some level of mixed stakeholders supervising the implementation and governance.Close collaboration among all the stakeholders is required to initiate, promote, foster, and strengthen collective learning, and overcome new problems, and research activities.The diversity of stakeholders suggests careful management of these relationships along the data management lifecycle (Sturges et al., 2015).There must exist good communication and cooperation among these groups (Jolak et al., 2018).This is deemed very important for a successful RDM.The lack of proper management may be a factor impeding the urgency for sound RM practices, as seen in other countries such as developed countries.

Librarians and Research Data Management (RDM)
Globally, librarians over the years have not been involved in RDM.They are basically into collection development (Johnson, 2019), but with time there is a call for librarians to participate in RDM.In making a case for librarians to be involved in RDM, some argued that services in the library in higher institutions are strategically positioned to offer a significant role in RDM (Andrikopoulou Rowley & Walton, 2021).The librarians have practical skills that could be used in RDM.These include advisory and training services.The result of the study in Tenopir et al. (2017) revealed that libraries of 20 best universities demonstrated enough commitment to provide services such as how to write DMP, data preservation, publishing, managing personal, and sensitive data.They suggested that libraries need to take a leading role in managing the research data generated in academic institutions.
Another study conducted by Faniel and Connaway (2018) in which thirty-six (36) academic librarians were sampled about the factors that influence RDM support and experience, the result revealed that 25% of the academic librarians state that there is a negative perception about them by researchers.The libraries initiated the support late, compiling researchers to start their initiative to embrace the current trend.Librarians have to initiate support for researchers to change the negative perception by providing and engaging researchers.
Librarians in RDM are key stakeholders.This is because they have relationships with all these stakeholders' groups, which positions them to be better placed to lead the RDM agenda.A key challenge will be the roles of all the various stakeholders involved in RDM, such as the government, academic institutions, funders, and researchers (Cox & Verbaan, 2016).According to Tenopir et al. (2017), academic librarians are directly collaborating with researchers.There is a need for stakeholders' engagement for librarians in high levels of interaction with researchers while cooperating with other support service providers such as institutions, funders, and government.There must be a reciprocal relationship between stakeholders for data to be managed cheaply (Malthouse et al., 2019).The support from key players such as the library professionals, government, top management in various institutions, and other funding agencies in Ghana for successful RDM.

Conceptual Framework and Hypothesis Development
The leading theory base of this study is that of stakeholders.Although it has developed significantly it remains a focal point.According to Freeman (1984), organizations' work affects other parties that are within and external to the institutions.Identifying such individuals or groups of people is vital to their success since they can negatively or positively affect organizational performance.These stakeholders' requirements must be known and well understood for proper implementation (Wilburn & Swanson, 2016).Key stakeholders are responsible for the adoption of RDMP.With these policies stakeholders have no choice than to comply if they would like to enjoy a good relationship and derive the full benefit.
This theory was used with a proposed model (Cocos & Lepori, 2020).This model is in three level such as policy level (government), program level (funding agency), and project level (performing institution including researchers and librarians).The stakeholders' demands were incorporated to forecast how they will influence organizational response to such multiple stakeholders.The stakeholder's theory as applied here only shows role and responsibility and its effect on research project (Cornel et al., 2021).This is because the various group involved in RDM must share information to enable their roles to be performed most efficiently to achieve all goals.It is important to note that stakeholder theory outline ethics in a pragmatic viewpoint.It says ethics are things that are done daily to solve our problems and make things better to meet individual needs.Example of such ethics in RDM are defining ownership of data, agreement, share data, accessibility, protecting the identity, and information of participant as well as the licensing of data.These ethics are key in DM and are the foundation of RDM as such the application of stakeholder's theory.
In Ghana, most of the funding comes from private individuals, non-profit organizations (NGOs) and international funders instead of the government.The situation is mainly due to economic conditions in most developing countries with scarce resources.As in Rogers ( 2021), scientific research is not integral to national policy.Most of the research are conducted through research grants sponsored by multilateral organizations such as the World Bank, the InterAmerican Development funders, and the European Regional Development Fund.The funding comes with policy requirements that go beyond the funding role and overlap the country's governmental support.Therefore, there is little or no connection between funders and the government concerning research funding.Research institutions receive a little support from the government for administrative and labor costs but they are independent.

Library Oriented Model of Institutional RDM
This model by Pinfield et al. (2014) focuses on issues that concern the challenges with RDM by distinguishing the various levels of activities, different stakeholders and drivers, and numerous factors affecting the adoption of any program.This model is adopted because it addressed the relationship' effect among the stakeholders as key actors in RDM and their roles for a successful program.It states that the implementation of RDM is affected by a set of influencing factors that either impact or have been impacted by the program.These factors will either facilitate or constitute a hindrance by influencing the character and the course of action.Therefore, success will be achieved if various stakeholders perform their roles per policies and strategies with various influencing factors in perspective.This study is in line with Shah et al. (2020), which revealed that ''data management, data curation, data mining, data science, data-savvy, and big data purpose is to identify relationships and causal associations, classify and predict events, identify patterns and anomalies, and infer probabilities, interest, and sentiment.'' The model presented here provides a relevant basis for our studies.This model engages various stakeholders.Based on an extant literature review, the research model was designed.Figure 1 shows our research framework.

Development of Research Hypothesis
Funders and Librarians.Funders' role in supporting librarians cannot be overemphasized.Funding research is key for the librarians to be proactive.Funding helps academic librarians to research into new knowledge.It is essential to help them be abreast with the current development in information science, manage other researchers that fall on them for academic direction, and, more importantly, manage research data.All these cannot be done without adequate support through funding.The soul of the library is money.Inadequate funds impede the effectiveness of any library activity, which will affect the work of librarians (Gao et al., 2020).The greater part of the sponsorship for librarians comes from state and local sources; federal funding provides critical support, enabling them to fulfill their communities' obligations.As such, the role of funders is a sine qua non for the success of librarians.Therefore, it is hypothesized that: H1: The funders agency criteria have positive effect on librarians RDM support services.
Institutions and Librarians.Institutions where librarians are working influence positively the work of librarians provided the institutions give the needed support.The quality of service rendered by librarians is dependent on the support provided by the institutions.The Institution is the librarians' employer and provides guidelines that librarians need to follow.Job satisfaction and leadership style improve the relationship between an employee and employer (Albro & McElfresh, 2021).Institution aims and staff development bring positive perception among the staff.It is hypothesized that: H2 The institutions support services affect positively the librarian RDM support services.
Institutions and Researchers.Institutions play a significant role in the work of researchers.The quality of programs and the services offered will affect the researcher's efficiency and increase its output of work (Cowen et al., 2019;Ponomariov & Boardman, 2010).There is growing competition for the fund due to the growing number of institutions.Institutions are now forced to compete for scarce resources and funds.Institutions are therefore very keen on the caliber of researchers that are recruited.The institutions would like to influence researchers' work performance to benefit from funding that only the best researchers can help to access.Therefore, it is hypothesized that: H3 institutions support services positively affect the work of researchers.
Funders and Researchers.In a study conducted by Go¨k et al. (2016), where researchers in all of the four countries have the chance to access the supra-national research funding from the European Union, it was found that citation impact of the researchers' work is positively related to funding variety and negatively related with funding intensity.Funders could influence the researchers' choice of topic (Webster et al., 2021).This burgeoning in terms of researchers' work raises issues for various funders and the quality of work done.Funders expect to see researchers demonstrating the highest possible profit on investment for their grant.It was also argued that funders affect researchers' knowledge production; the hypothesis is that H4 Funders agencies criteria affect the researchers' work.
Government and Librarians.According to a study conducted in Indonesia (Nashihuddin, 2018), it was noted in the conference that the government' program initiated needs to be followed up by a librarian.Government must include the librarians as key stakeholders in program implementation, such as research data management policy, because librarians' work is not generally on the government's top policies.In Osunrinde and Adetunla (2018) the study reveal that the knowledge level of librarians toward the preservation and conservation of information materials was quite high, and the level of training has improved greatly when government sponsored.It is there hypothesized that: H5 Government support policies affect the librarians RDM support services.
Computer Literacy and Librarians.The computer can help things to be done faster and easier.The librarians with computer literacy are likely to be more productive and efficient at work than those who are not.The Librarian's ability to manage research data depends on the computer literacy level with a set of technical skills such languages (Python, Structured Query Language, Java, and eXtensible Markup Language), design and structures of databases, user-centered design, natural language processing tools, Internet of Things, and large data (Semeler

Population and Data Collection
The research approach used in this research is the deductive research where the study refers on the previous concept approach and theories in formulating the research hypotheses and objectives.The research approach is also explanatory (Casula et al., 2021).This is to enable us to have better understanding of the phenomenon.The study sample were selected by ensuring that the right target fill the questionnaires in order to collect in-depth data from them.The questionnaire seek to understand the effect of the various stakeholders in RDM namely librarians, funders, researchers, government (Table 1).The researchers were guided by the list released by commonwealth network (Nexus, 2020) to select the institutions or organization in order to recruit the respondents.Through the help of research assistants, the various units were contacted first to check their availability.After acceptance the questionnaire were sent to them.So, the analysis is based on the responses received which are fit for our data analysis.Therefore, the study population are all researchers in academic research institutions in Ghana notably, the University of Ghana's Institute for Medical Research, The Cocoa Research Institute of Ghana (CRIG), African Centre for Conflict Resolution and Reconciliation, Centre for Sustainable Development Initiatives, Council of Scientific & Industrial Research, Geological Survey of Ghana, Ghana Meteorological Services Department, Institute of Local Government.Studies.A purposeful sampling was adopted to ensure the right people fill the questionnaire.For a person to be qualified, the respondent must with the institution for not less than 2 years.In addition, the respondent must be involved in research activities and has individually or as team benefited from research grant.The questionnaires were first piloted with few researchers to check for its feasibility.Upon careful scrutiny some corrections were made and was finally administered which lasted for 2 months (January -February).Each respondent answered all questions.A survey questionnaire was administered to 501 respondents, and a response rate of 59% was successful.So, the analysis was based on 295 respondents.No one was forced to answer any question and confidentiality and anonymity were guaranteed.The questionnaires were coded in IBM-SPSS (version 25).

Instrumentation
The stakeholders in RDM in Sturges et al. (2015) were adapted namely (institutions, funders, researchers, government, and learned society of which librarians are members) with focus on their Support services in RDM from which we derived the variables namely Researchers' works, Librarians RDM support, Institutional support services, Funders' agency criteria, Government support policies, and Computer literacy of librarians.Computer literacy was included to check its effect on of librarians since it is key for their new role.Since RDM policy stipulates roles, duties and responsibilities of each stakeholder, it may be logical to believe that if every stakeholder performs their supporting roles there will be some amount of effect.So, the questions are related to their roles and support services in RDM.Previous finding in Liu et al. (2020), Research data management policies in USA, UK and Australia universities has demonstrated the areas of focus of research data management policies and the stakeholders' roles and its intended benefit.Therefore, the researchers' work was measured with RDM support received, and their roles in RDM such as data sharing and access and DMP data and data analysis.Librarians RDM support service' items were related to the support provided to researchers with regard to 37 itemized questions all related to their duties and responsibilities as shown in Appendix 1.A 5-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) were used.This was adapted to make the questions clear, in turns creates an appropriate data sample.The feedback can come in handy, as you can compare against specific outcomes instead of indescribable goals (Joshi et al., 2015).

Data Analysis
Quantitative data were analyzed with the partial least square-structural equation modeling (PLS-SEM) method.IBM-SPSS (version 25) and the SmartPLS software were preferred because they could explore small data sets in quantitative studies.It also enables path analysis that shows the direction and the effect of the various variables (Sarstedt & Cheah, 2019).The Cronbach's alpha value was used to check data validity in factor analysis.Significant figures were determined at a confidence interval of 0.05%.

Results and Discussion
The data assessment was done to check for internal consistency, reliability, convergent and discriminant validity to check the strength of the measuring items and how they are holding together.The minimum criteria have all been met for various loadings, the average variance of various extracted (AVE) ranged from 0.504 to 0.698.The AVE value of ø 0.5 satisfies the condition for convergent validity.Apart from funders and computer literacy, Cronbach a was all above .7.The composite reliability ranges from 0.774 to 0.898.As shown in Table 2, the result was consistent with the study (Hair et al., 2016).

Evaluation of Measurement Model
Table 3 shows the Heterotrait-Monotrait Ratio of correlation (HTMT).It was also used to assess the discriminant validity.The HTMT is below.

Evaluation of Structural Model
Discriminant validity is how a construct is different from other constructs.It infers that a construct is unique.In order to evaluate the discriminant validity of the model, the Fornell & Larcker criterion, Heterotrait-Monotrait Ratio of correlation (HTMT) and Cross Loadings were used.Results of the discriminant validity of all the constructs computer literacy, funders, government, Institution, librarians, and researchers obtained using the square root of the AVE and cross-loading matrix, respectively, are shown in Table 3.According to Hair et al. (2016), the square root of the AVE of a construct should be greater than its correlation with other constructs for satisfactory discriminant validity, and the diagonal elements must be larger than the entries in corresponding columns and rows to satisfy discriminant validity (Henseler et al., 2009).Additionally, to use Fornell and Larcker's criterion, the squared root of the AVE values was compared with the latent variables (Fornell & Larcker, 1981).

Path Analysis and Hypothesis Testing
The Structural model (Figure 2) shows relationships between the constructs in the research model.The collinearity calculation, path coefficient (b) using t-statistics, coefficients of determination (R 2 values), effect size, blindfolding and predictive relevance (Q 2 values), and effect size (q 2 values) were used to test the strength of the relationship between the dependent and independent variables.Furthermore, the collinearity among latent variables is measured through the variance inflated factor (VIF).There was no collinearity as they were less than 5.According to Hair et al. (2011), the maximum value of VIF greater or equal to 5 (VIF ø 5) is an indication that the latent collinearity is problematic.The value of R 2 in Table 4 is substantial.A range of R 2 from 0 to 1 is acceptable.Any value higher than this shows some levels of predictive accuracy.In Hair et al. (2011Hair et al. ( , 2013) ) and Henseler et al. (2009), R 2 values of .75 are substantial, .50 is moderate, and .25 is weak predictive abilities of the variable.Chin (1998) also articulates the values of 0.67, 0.33 and 0.19 as substantial, moderate and weak, respectively.The Q 2 is an indicator of the model's predictive power or relevance.The result is obtained by using blindfolding procedures for a specified omission distance D with values between 5 and 10.Q 2 values have predictive relevance as they are above zero (0), and q 2 values of 0.35 indicating an exogenous construct, meaning that there is large predictive relevance for the endogenous constructs.
These effects were tested based on the criteria proposed by Hair et al. (2017) and Zhao et al. (2010) .The indirect and direct effects are also significant.

Discussions
The structural model in Figure 2 shows that funders agency criteria have a significant positive effect on librarians RDM Support service with a path coefficient of 0.092.The amount of funding that a library receives directly influences the quality of its services (Arnott et al., 2020).Their findings buttress our results in that a Note.The factor loading are all higher than 0.70.This means that causal relationships under study are truly distinct from each other.In other words, they are different and not measuring the same thing.There is no multicollinearity.
grant funding a librarian to join a research team to provide direction regarding research data management has significantly impacted the team's data gathering, storage, and curation methods.It is noted, however, that the path coefficient is below 0.1.This is because funders agency criteria do not directly affect librarians, since the funding  is still relatively low.The librarians in Ghana and Africa still need more support.This study also observed that the institutions support services positively affect the librarians RDM support service in RDM with a path coefficient value of 0.102.This result is clear evidence that the institutions' support can increase librarians' output in research data.Regular training to upgrade librarians' skills and knowledge will go a long way to produce much more positive results to boost the confidence level (Lockhart & Majal, 2012).Many institutions though independent in Ghana and other African countries still need government support and policy direction as a result are not able to provide as much as librarians work demand.The institutions depend on government for infrastructure development.Even though the institution may have the willingness and prepare the librarians until government policy support them, this vision will never be fulfilled (Darch et al., 2021).
The institutions RDM support service has a positive effect on the work of researchers with a path coefficient of 0. 515.The Institution where the researchers work is responsible for policy implementation.It is also responsible for the provision of some infrastructure and training despite the autonomy to researchers to complement that of government.All these contribute to the level of significance reported in this study.This assertion is also in line with Moher et al. (2020) that institutional support positively affect researchers' work in fulfilling better the RDM.The existing guideline is impactful.The institutions perform a leadership role as well; the leadership style exhibited, therefore, positively influences the researcher's job satisfaction, which in turn translates to greater efficiency and output of researchers.
The funders agency criteria influence researchers' work with a path coefficient of 0.363.The impact of research alludes to how research influences a wide variety of phenomena and trends in society, which emerged from the combined effect of research findings (Hove, 2020) that generally manifest over the long term.Funders (research councils, charities, foundations) required researchers to make research data underpinning their work freely available, reusable, accessible, and stored.These funders' policies go a long way to positively affect researchers' work by causing them to produce quality research.Our finding is consistent with Benner and Sandstro¨m (2000) who argued that funding positively affects the researchers' output and behavior.Funders agencies criteria have a positive impact on the quality of the university industry of the researcher (Hottenrott & Thorwarth, 2011).In Go¨k et al. (2016), a citation impact of the researchers' work positively impacts fundingrelated variety and negatively impacts funding intensity.This could account for the level of significance reported in this work.
Government support policies influence on Librarians RDM support services has no significant influence (p = .03).This means that hypothesis 5 (''government support policies affect librarian RDM support services ") was rejected as the study results did not confirm it.However, this does not mean that government does not have any influence on librarians.It could mean that its effect is an indirect one, given that government interventions to librarians are not direct, which was found to be the main hindrance to RDMP.This may also be due to the lack of policies since the government relies on such policies as RDMP to release funds.Our result is line with the finding of Huang et al. (2021) that government effect on librarians' work in China is insignificant.The authors argued that the librarians in China lack professionalization.Furthermore, there is a ''wait and see'' attitudes because they were not involved in RDM policy formulation and implementation.It is a wake-up call to all governments yet to formulate RDM to fully involve librarians at early stage in order not to have any implementation drawback the case in China.Librarians need to update their knowledge to remain relevant.The positive effect could be demonstrated by getting a national award, successful grant applications, and project work, especially in efficiently and effectively managing research data (DeLuca, 2020).
Computer literacy of librarians and Librarian RDM support service recorded a path coefficient value of 0.004.The result shows an insignificant effect of computer literacy on the work of librarians.According to Julien et al. (2018) and Zhou (1996) the effect of computer-related skills is very substantial, if not critical, for librarians in academic institutions.However, this is not so significant in our findings.Our study found that most of the Librarian sampled have either no computer skills or have only a basic level of computer literacy, so instead of a computer helping them to be faster, it rather slows them.As a result, they rely on manual ways of working instead.
Furthermore, again this may be because the automation of Ghanaian libraries for research data management would not automatically lead to greater efficiency and productivity unless librarians acquire essential skills to use this technology effectively.The most common problems cited were the system's frequent breakdown, followed by electric power failure, and inadequate computers in the libraries.
Similarly, in Emojorho and Adomi (2006), unreliable telecommunication, infrastructure, and electrical power outages are barriers that negatively influence the use of IT facilities by the majority of staff.These challenges are common in developing countries such as Ghana and many other African countries.Librarians and researchers do not also possess analytical skills and lack the abilities needed for working confidently in this new field of the changing information environment (Kirkwood & Price, 2016).In Baro et al. (2019), many librarians rank themselves low in skills related to RDM.As a result of these challenges, more time is spent on the work than expected, leading to the level of insignificant reported in this study, thereby rejecting hypothesis 6.In order to compensate for this weakness, librarians and researchers need to be prepared to adapt to the new circumstances and comply with the emerging requirements of research data management.Unless these issues are resolved, the impact may not be so significant as expected.

Conclusions and Recommendation
This paper sought to address a set of research questions to discover the causal effect among key variables in research data management: researchers' work, librarians RDM support services, funders agencies criteria, government support policies, and Institutions RDM support service with a focus on librarians' roles toward research data management.The stakeholder's theory has been extended through its application to research data management instead of the organization.Six hypotheses were formulated (Table 4) and four were accepted, and two were rejected.Funders 'agencies criteria and Institutional support service positively affect the librarians' RDM support service work.Besides that, funders agencies' criteria also significantly affect the outcome of researchers 'work in research data management.The study found that the Librarian's contribution to research data management is impactful and positively affects researchers' work.Hench, it is concluded that there is a positive relationship on the research variables and as such, stakeholder relationships affect the research data management (Kim et al., 2020).This means that when various stakeholders play their roles and responsibility, it motivates one another and cause the other party to be more efficient and effective in managing research data management resulting in quality research.Concerning the librarians, the more the support, the more the output.Librarians are well-positioned to lead the RDM agenda and therefore should take a leading role.This study is consistent with Andrikopoulou et al. (2021), who found a positive collaboration among RDM stakeholders.It was also found that corporate stakeholders such as institutions can influence the performance directly through other internal staff or indirectly through other external stakeholders such as the government and funders (Li et al., 2020).The study recommends computer adaptation in research data search and management.The researchers should be encouraged to use FAIR data to harness socio-economic development benefits (Stall et al., 2019).In Africa, as elsewhere globally, RDM can advance research and accelerate development through data sharing and reuse.However, this demands that governments, funders, researchers and institutions work as a team to realize its full potential through policies and incentives that promote data sharing.Moreover, because the effect of computer literacy was not significant, we recommend that librarians should be trained in computer literacy to achieve a significant impact.Other related tools to help them operate without interruption must be in place.This includes an uninterrupted power supply, provision of computers and good network infrastructure in order not to fall back on manual ways of performing these tasks.Technical/technological skills, with a focus on metadata, knowledge of data resources is needed; other skills are communication and project management and soft skills, oral communication skills and the ability to cultivate relationships with patrons.It is not easy to practice FAIR data principles without research data management policies.The institutions through the government should also provide an enabling environment and needed infrastructure to do the work smoothly with RDM policies.
There will be a need for data-intensive services such as data sharing, access, preservation, storage, and services to assist researchers in writing DMP.Government and funders should continue to provide the necessary support, and more funds should be released.The higher the funds which funders and government provide, the more significant the impact on Librarian's work in managing the research data and other services they engage.The researchers should take full advantage of RDM.This will go a long way to increase productivity and help manage better research data, thereby increasing the impact factor for the full realization of the RDM benefit.Such collaborations with librarians can significantly enhance the scientific research process.The movement of librarians into new informationist roles has significant implications for the information professionals and researchers to perform better.This study is useful to all developing countries for education and teaching as it effectively bridges the gap between theory and practices.It could serve as the basis for policy formulation.

Limitation and Future Work
This research is limited to only key stakeholders' contribution in data management, which is because their role is critical to the success of RDM.However, future research can broaden the scope of stakeholders and other aspects of RDM, such as infrastructure development and training to verify and complement this research's findings.

Figure 2 .
Figure 2. PLS-SEM path model showing relationships between study variables.

Table 1 .
variables and Descriptions.

Table 2 .
Data Reliability and Convergence Test.

Table 3 .
Discriminant Validity of Factors.