Teaching research data management courses in higher learning institutions in Tanzania

Abstract

Purpose

The study investigated teaching research data management (RDM) courses in higher learning institutions (HLIs) in Tanzania to enable postgraduate students to work with their research data.

Design/methodology/approach

The study triangulated research methods. Postgraduate students were investigated using survey questionnaires to learn about their needs and perceptions of the teaching RDM courses in HLIs. Key informants (academicians, information and communication technologists and library staff) were also investigated using in-depth interviews to explore their experiences and knowledge of teaching RDM courses. SPSS statistical software was used for analysing quantitative data; qualitative data were analysed thematically.

Findings

A total of 70 questionnaires were distributed to postgraduate students with a returning rate of 44 (69%). On the other hand, 12 key informants were interviewed. A low level of RDM literacy was revealed among 38 (86%) respondents. Most respondents 40 (91%) reported the need for HLIs to start teaching RDM courses. A lack of skills and knowledge in teaching RDM courses was revealed among key informants. The competency-based, adaptive and constructive teaching techniques were selected for teaching RDM courses, whereas intensive training and online tutorials were revealed as teaching formats.

Research limitations/implications

This study focused on teaching RDM courses in HLIs. The survey questionnaires were distributed to all 2nd year postgraduate students, however, the findings cannot be generalised to all postgraduate students due to the response rate obtained. The findings obtained from key informants can also not be used as a basis for generalization across HLIs.

Practical implications

This study concluded that postgraduate students need to be well equipped with skills and knowledge on RDM and its related concepts; teaching RDM courses should be regarded as a continuous programme for benefit of students, researchers and the community at large.

Social implications

Appropriate teaching of RDM courses among students not only ensures that students meet the funders’ and publishers’ requirements, but also encourages students to store and share their research among researchers worldwide; thus increasing collaboration and visibility of the datasets and data owners through data citations and acknowledgements.

Originality/value

This is a comprehensive study that provides findings for HLIs to teach RDM courses in HLIs, especially for postgraduate students. The findings revealed the need for teaching RDM courses in HLIs. The study provides the basis for further RDM research in HLIs and research institutions.

Citation

Mosha, N.F. and Ngulube, P. (2023), "Teaching research data management courses in higher learning institutions in Tanzania", Library Management, Vol. 44 No. 1/2, pp. 166-179. https://doi.org/10.1108/LM-04-2022-0033

Introduction

The development of research data as a key source of information is now becoming apparent; thus, to take advantage of this revolution, students and researchers in higher learning institutions (HLIs) need to be trained in research data management (RDM) and related fields. Research data are defined as the output from any systematic investigation that involves a process of investigation, experiment or testing of a hypothesis, as well as from artefacts of cultural heritage (Pryor, 2012; Nielsen and Hjørland, 2014). Lewis (2010) defined RDM as the processing of all types according to raw materials, which is created along every research process. Research data is increasingly recognised as a vital resource whose value needs to be preserved for future research (Perrier et al., 2017); and thus, it places a huge responsibility on HLIs to teach RDM courses. Students in HLIs are always working with research data and data management in different ways (Radecki and Springer, 2020); however, formal education in RDM is very rare at the university level (Whitmire, 2015; Wiljes and Cimiano, 2019).

Studies indicate the need for teaching RDM concepts and practices to enable students to apply new methods of managing their research data very (Tammaro and Casarosa, 2014; Carlson and Bracke, 2015; Whitmire, 2015; Pasek and Mayer, 2019; Perrier et al., 2017; Wiljes and Cimiano, 2019; Yang and Li, 2020). There is a larger gap in understanding graduate students’ RDM needs (Pasek and Mayer, 2019). In this regard, HLIs need to teach RDM courses to ensure students meet their RDM needs in all stages as specified in the research data life cycle (Schneider, 2013). According to Wiljes and Cimiano (2019), the main aim of teaching RDM courses in HLIs is to enable students to apply the contents of the courses in their research work.

Teaching RDM courses is gaining importance worldwide as it ensures data quality, integrity, shareability, discoverability and reuse of data over time (Henderson and Corry, 2021; Perrier et al., 2017; Wiljes and Cimiano, 2019). For example, the Oregon State University (OSU) has established a RDM course that is open to graduate students from all disciplines without any prerequisites and is taught by the libraries’ data management specialists. Curricular materials are drawn from existing resources where possible, but much of the lecture and computer laboratory content was developed from scratch (Whitmire, 2015). Syracuse University developed a program “to train students with knowledge and skills in collecting, processing, managing, evaluating, and using data for scientific inquiry, the project known as Science Data Literacy” (Qin and D’ignazio, 2010). Schneider (2013) adds that forthcoming generations of students of almost all disciplines will have to work with their research data in systematic and scientific ways.

Studies identified gaps based on the lack of teaching RDM courses in HLIs. O’Reilly et al. (2012) explain that researchers do not undergo any training, which leads to unreliability, inaccessibility and short lifespan of their data and data management in general, and they do not receive any help on how to manage their data (Buys and Shaw, 2015). This means that they learn as they go, unsystematically, during their day-to-day research work (Fear, 2011; Peters and Dryden, 2011). On the other hand, research indicates a lack of knowledge and skills in RDM, from data generation to data storage and sharing (Perrier et al., 2017; Wiljes and Cimiano, 2019; Henderson and Corry, 2021).

Lack of competencies, lack of practices and perceptions of data use, range of data sharing and data type, as well as incorrect researchers’ perceptions toward RDM hinder HLIs from teaching RDM courses (Peters and Dryden, 2011; Jahnke et al., 2012; Hickson et al., 2016; Kurata et al., 2017). For example, Whitmire (2015, p. 2) explain the RDM components of open data and data sharing as “…… that many graduate students would probably not be prepared to deposit a dataset into the IR because they would not have received any formal training in data management during their programs, and as such, their data would not be an insufficient condition to be shared (i.e. be well organized and sufficiently documented)”. The quote insisted that competencies in working with data are often not included as a part of a student’s formal education (Carlson and Stowell-Bracke, 2013).

This study investigated the teaching of RDM courses in HLIs in Tanzania. Additionally, the present study focused on postgraduate students who are working on their research. Graduate students do not have a lot of knowledge of RDM; thus, HLIs shall be required to obtain RDM skills for their research progress and job requirements in educational fields (Pasek and Mayer, 2019).

Methods

Study area

This multimethod design study was conducted at the Nelson Mandela African Institution of Science and Technology (NM-AIST), one of the public universities in Tanzania. The institution has four schools, namely the School of Computational and Communication Science Engineering (CoCSE), the School of Materials, Energy, Water and Environmental Sciences (MEWES), the School of Life Sciences and Bio- Engineering (LiSBE), and the School of Business and Humanities (BuSH). However, the School of BuSH does not have any students, since it was established to enable the other three schools to integrate science and engineering subjects with business studies. Therefore, the School of BuSH was not part of this study. The NM-AIST was selected based on the initial preparation for the implementation of the RDM services such as the approved RDM policy and operational procedures, as well as, a well-equipped Information and Communication Technology (ICT) department with equipment like servers and High-Performance Computing (HPC) for storing research data and handling big data.

Study population

The study involved postgraduate students who were in their 2nd year, and who had already started their data collection. The study selected postgraduate students because they have already presented their research proposal, have collected their research data and they were at the stage of analysing data and report writing; thus they have the experience in RDM and its related concepts. Mannheimer and Banta (2017), RDM instruction programs often focus on faculty and graduate students who normally work with research data, thus, improving their data management strategies. Key informants involving academic, library and ICT staff were selected to add more information and experiences on teaching RDM courses and the advantages of RDM in HLIs.

Sample size and sampling procedures

The whole population (2nd postgraduate students) were involved in this study because the population size was very small (70 postgraduate students) then surveying all 2nd year students was ideal and more rigorous. Key informants were conveniently selected. The interview data were studied and collected until it was clear that perspectives were being repeated and data saturation was reached. In this study, the point of saturation was reached when the researcher conducted interviews with 12 key informants. The interview venue was selected based on the convenience and preferences of the respondents, taking into consideration their privacy and comfort. The library staff were assisted to set up appointments for the interviews and distributing the questionnaires to the respondents. All respondents were requested to fill out the informed consent before the interviews.

Data collection

In line with the multimethod design, qualitative data were collected from key informants to supplement quantitative data collected from postgraduate students. Survey questionnaires were distributed to 70 postgraduate students, with a return rate of (n = 44, 69%), whereas 12 key informants were interviewed using a developed interview guide.

Data analysis

In this study, both quantitative and qualitative data were analysed separately, and then qualitative data were merged with the quantitative data set to provide support to the results of the quantitative data set (Creswell and Plano-Clark, 2007). The study performed data cleaning and analysis using SPSS version 20 (Statistical Package for Social Science Software (SPSS), IBM SPSS Statistics). The thematic analysis approach was used to analyse qualitative data from key informants. The analysis was carried out in three stages: first, the line-by-line coding of field notes and transcripts (unpacking of text into discrete elements to expose underlying thoughts and meanings); second, the in-depth examination and interpretation of the resultant codes into descriptive themes and; third, interpretation of the descriptive themes into more abstract analytical themes. The study further employed between-method triangulation, whereby qualitative data were used to corroborate the findings obtained from quantitative data.

Results

A total of 44 postgraduate students aged 21–50 years were enrolled in this study of which 33 (75%) were males. The demographic information for postgraduate students was provided (Table 1). Further qualitative data was collected from 12 key informants (6 Academicians, 4 Librarians and 2 IRC staff).

RDM literacy level

Overall, postgraduate students were not aware of the RDM and its related concepts. Respondents 38 (86%) reported a low level of RDM literacy, with four respondents reporting a high level and two respondents saying they had a moderate level (Table 2).

Key informants described RDM literacy level as the skills obtained by students to enable them to work with their research from data collection/generation to data curation. They saw RDM courses as being essential for students to work with their research. They believed that students gain elementary knowledge in RDM during research methodology courses offered at NM-AIST as follows: research methods and communication (Master and PhD), applied research methods (Master and PhD), research methods and statistics (Master), and biostatistics (Master). Indeed, they identified a need for RDM for students to learn how to work with their research. As one respondent reported:

RDM courses shall be taken as among the core courses to enhance RDM literacy level for students to work with their research” [Academician #1]. Capacity building among librarians, IRC and academicians was considered important in teaching RDM courses. Another respondent reported, “We think that there is a need for the institution to teach RDM courses among students; In fact, the institution should empower its staff especially librarians to be competent in RDM [Librarian #1].

The need for teaching RDM courses in HLIs

Respondents were further asked to indicate if there is a need for HLIs to start teaching RDM courses in HLIs. Respondents 40 (91%) reported the need for teaching RDM courses in HLIs. Key informants considered RDM as among the topics in HLIs and they supported the need for teaching RDM courses in HLIs. However, capacity building for facilitators and instructors for teaching RDM courses is highly recommended.

One respondent reported that:

Even though RDM skills to teach our students are needed, we cannot wait until we get training since most of us, we have research methodology skills to start with [Academician #2].

Teaching techniques for RDM courses

Generally, teaching techniques for teaching RDM courses in HLIs were required. In this case, competency-based teaching and adaptive teaching techniques were selected by most respondents 39 (23%) (Table 3).

Key informants selected the constructive teaching technique since it involves both lecturers and students and it encourages teamwork. One respondent added:

I think constructive teaching technique enables learners to contribute and share their knowledge [Librarian #2].

Another respondent insisted:

Since RDM is among the new courses in HLIs, constructive teaching technique could enable both students and lecturers to participate well in teaching and learning activities [Academician #5].

In addition, constructive teaching technique provides both practical and theory to enable students to learn in deep and be competent in the subject area. One of the respondents contributed that:

I also think that any teaching technique that involves both hand-on and theory will enable students to be able to practice well, thus, HLIs should use techniques which are focusing more on practical than theory [IRC Staff #1].

Another respondent added that:

I encourage RDM modules which involve “hands-on activities” to enable researchers to learn about RDM software and tools, therefore, instead of just demonstrating the tools to allow students use them [Academician #6].

Teaching format for RDM courses

Respondents were also required to select a teaching format that will enable them to catch lessons very easy. Many respondents 34(27%) preferred intensive training (online) (Table 4).

The findings from key informants mentioned intensive training (online) and online tutorials (video clips, information etc.) as teaching formats for teaching RDM in HLIs. One respondent added that:

Online teaching and learning formats need to be considered in teaching RDM courses in HLIs. These formats are very important since they can enable students and other learners to study in their own time and to have learning resources to assist them in their studies [Librarian #3].

Another respondent insisted on the sharing of learning materials to students using online mechanisms, and he, therefore, reported that:

We like facilitators to produce and circulate teaching and learning materials and share them online to enable students to use them at their own time [Librarian #4].

Another respondent provided the advantages of online formats:

I think online teaching formats will enable tutors to reach a big number of students wherever they are and these are good formats especially, in this time when the world experiencing covid-19 pandemic in which social distancing is among the preventive measures [Academician # 8].

RDM topics to be covered

RDM topics were also identified, and the majority of respondents 34(20%) selected research data policy and guidelines (Figure 1).

Key informants’ respondents selected research data life cycle and Data Management Plan (DMP). One respondent reported that:

I suggest practical implementation and follow-up of the DMP. DMP is very important during early stages of data collection to data storage [Academician #7].

Benefits of teaching RDM courses in HLIs

There are several benefits for teaching RDM courses in HLIs. Respondents 41(24%) indicated meeting publishers’ requirements, whereas key informants mentioned funders’ requirements. One of the respondents felt that meeting funder requirements were of primary importance for students, HLIs and the nation in general:

Most of the funders’ requirements need to see their data to be open, hence to enable other researchers to use the data for more research findings and to use the fund for the development of the country and its citizens [Academician #3].

Another respondent added:

Data sharing and open data are among the components in RDM that enable funders to invest more in research [Academician #4].

Other benefits mentioned included saving money and time for recollecting research data on the same area and strengthening scholarly communication.

Discussion

The findings revealed that RDM literacy was low among respondents. A similar finding was provided by Mushi et al. (2020), which revealed that the University of Dodoma (UDOM) lecturers and stakeholders (students) have a low level of RDM and related areas. This finding was inconsistent with the study of Adika and Kwanya (2020) which indicated RDM literacy levels among 50 students interviewed on research data literacy skills as 2(4%) poor, 38(76%) good, and 10(20%) very good. Also, the study conducted at the University of Oregon by Carlson et al. (2015) revealed that “… students may have picked up their skills on the job training because a lot of them had a former life in a professional field … or it’s something they got as an undergraduate”. Again, the study by Carlson and Stowell-Bracke (2013) noted that graduate students at Purdue were involved in the collecting, processing and analysing of research data. It is thus concluded that postgraduate students should have a kind of RDM skills to work with their research as compared to undergraduate students.

Generally, teaching RDM courses is mandatory in HLIs. The study findings revealed that RDM courses will equip researchers and students in HLIs with more skills and knowledge. This finding was consistent with preceding studies by Carlson et al. (2011), Buys and Shaw (2015), Mandinach et al. (2015) and Cowie and Cooper (2017). The study by Wiljes and Cimiano (2019) also added that The Cluster of Excellence Cognitive Interaction Technology (CITEC) [1] released the CITEC Open Science Manifesto [2] stated that “CITEC recognizes the need to extend the educational curriculum for young scientists towards topics of research data management and offers training and personal consulting for advanced researchers, thus contributing to awareness among young researchers of good practice in scientific research”. Teaching RDM courses in HLIs enables researchers to work with their research data as well as ensure risks regarding data security and privacy, which many researchers are unaware of (Jahnke et al., 2012). Buys and Shaw (2015) found that researchers need help, guidance and training regarding long-term access to data, long-term preservation of data, becoming acquainted with services that support these processes, as well as providing backup copies of data during the project lifetime.

The study found competency-based and adaptive teaching techniques in teaching RDM courses in HLIs. A similar finding emerged from the study done by Wiljes and Cimiano (2019), namely that the competency teaching technique that relies on practical exercise is useful for teaching RDM courses in HLIs. The study is inconsistent with the study done by Biernacka et al. (2021), which proposed ways of teaching and learning data mining and learning analytical skills by using adaptive learning (adaptive teaching technique) to orchestrate the interaction with learners and deliver customised resources and learning activities to address the unique need of each learner. Whitmire (2015) added that the active learning exercises were successful and generally enjoyed by students. In addition, the active approach is driven by practical reality and thus enables students to engage with the content and techniques of RDM rather than lecture alone (Whitmire, 2015). The study also found that a constructive teaching technique should be used in teaching RDM in HLIs, which is similar to the study of Mannheimer and Banta (2017), which found that to make RDM relevant among students in HLIs, instructors need to use constructive learning for strategies that create meaningful learning experiences. This study found that any mode which involves “hands-on activities” enables students to work with their research data practically. Wiljes and Cimiano (2019) add that hands-on demonstrations, group discussions and individual presentations allow students to link the acquired knowledge to their research works. HLIs should use diverse teaching and activation methods to increase engagement and active participation among students (Biernacka et al., 2021). In addition, application, knowledge transfer, and replication methods need to be integrated to facilitate the learning process (Biernacka et al., 2021).

The study findings revealed that an intensive training (online) format should be considered in teaching RDM courses in HLIs. Online learning systems learning management systems, learning platforms, and learning software can capture streams of fine-grained learner behaviours, and the tools and techniques described above can operate on the data to provide a variety of stakeholders with feedback to improve teaching, learning, and educational decision making. A study by Mushi et al. (2020) recommended several online training modules available freely for teaching RDM courses.

The study finding also revealed that DMP should be considered among the topics to be covered in teaching RDM in HLIs. The finding is consistent with studies done by EDINA and Data Library (2011), and Henkel et al. (2012) on DMP, metadata and documentation, organisation, storage, backup and preservation practices, and data publication and sharing are among the topics to be considered in teaching RDM and their instructional programs. In addition, Baker et al. (2015) recommended topics that include data collection techniques, development of a data management plan, keeping data collection activity logs and providing training in summarising the results. DataONE (2016) proposed RDM topics to be covered including data management, data sharing, DMP, data manipulation, data quality control assurance, data protection, metadata, data citation, analysis and workflow, as well as legal and policy issues. NLM (2012) developed seven modules for teaching RDM courses, namely Module 1: an overview of RDM; Module 2: data types, strategies and format; Module 3: metadata; Module 4: data storage, backup and security; Module 5: legal and ethical considerations; Module 6: data sharing and reuse policies; and Module 7: archiving and preservations.

Parallel to previous studies done by Kennan and Markauskaite (2015), Patterton et al. (2018), Bangani and Moyo (2019), Chigwada et al. (2019), Adika and Kwanya (2020), and Mushi et al. (2020), the present study found that publishers’ and funders’ requirements are among the benefits for teaching RDM courses in HLIs. Kennan and Markauskaite (2015) added that “There is increasing pressure from funders, publishers, the public, universities and other research organisations for researchers to improve their data management and sharing practices”. Funders’ pressure to add value to expensive research and to solve cross-disciplinary grand challenges as well as pressure from publishers to ensure transparency and reproducibility of the scientific record, made researchers acquire RDM knowledge (Kennan and Markauskaite, 2015; Patterton et al., 2018; Bangani and Moyo, 2019). López-Meneses et al. (2020) explored the benefit of RDM, namely to enable students for their professional development. Additionally, Yang and Li (2020) added that teaching RDM courses enable students to conduct their research activities effectively and HLIs to improve the quality of their research services. EPSRC (2018) mentioned the following benefits of RDM in HLIs, especially on data storage and sharing:

1. Increases the visibility of research data for citation and reuse by other researchers.

2. Reinforces scientific enquiry by ensuring that enthusiasts and sceptics alike can test, validate and replicate research results.

3. Promotes new research and different ways of testing and analysing research data.

4. Discourages unnecessary duplication of effort by researchers.

5. Saves financial and other resources that are wasted when similar data sets are created by different researchers.

6. Enables discoveries from old research data sets.

The main strength of this study is its composition of respondents (students, academicians, librarians, ICT Personnel) involved in teaching RDM courses in HLIs. Awareness of these issues is essential for HLIs’ management together with their units within the HLIs, to be able to design RDM courses and develop RDM services that suit the needs of the research community. The findings can also be used to develop RDM curriculum based on different RDM topics and shall be recognized as a standard guideline.

Conclusion

This study concludes that there is a need for HLIs to start teaching RDM courses to their students and thus, RDM courses shall be considered among the topics to be included in curriculums. Despite the benefits mentioned in this study, RDM shall be required to enable students in HLIs to share their data and other researchers worldwide and store their data in data repositories for future use. Research data has a long lifespan than the project, thus, students need to plan for how long they will manage their data after the project has finished, as well as before and during research. Research data literacy skills are very important for students even after completing their studies. Therefore, teaching RDM in HLIs shall be considered an essential part of academic education ranging from online guides, classroom presentations, and workshops. Teaching techniques that includes both theoretical and practical sessions shall be accommodated for teaching RDM courses in HLIs. Additionally, the blended learning format is appropriate for teaching RDM courses since it combines both traditional classroom (face-to-face) and online courses. This study will be the basis for other HLIs to set up similar courses for their students, researchers and staff.

Recommendations

1. HLIs’ should think of teaching RDM courses to their students especially for those engaging in research activities.

2. HLIs’ worldwide should implement the teaching technique which will involve both theory and practical sessions to enable students to be well informed about the intended courses.

3. HLIs should ensure that capacity building is provided for all HLIs’ staff who will be involved in teaching RDM and related courses. This can be through long and short course training, workshop programs, advocacy and seminars.

4. HLIs shall ensure that the RDM infrastructure is well implemented to enable data analysis, storage and sharing.

Figures

Figure 1

RDM topics

Table 1

Demographic information of postgraduate students (N = 44)

Item(s)		Frequency (n)	Percentage (%)
Gender	Male	33	75
	Female	11	25
Age	21–30	10	23
	31–40	25	57
	41–50	9	20
Education level	Bachelor Degree	27	61
	Master Degree	17	39
School	School of Computational and Communication Science Engineering (CoCSE)	15	34
	School of Materials, Energy, Water and Environmental Sciences (MEWES)	12	27
	School of Life Sciences and Bio- Engineering (LiSBE)	17	39

Table 2

RDM literacy level

RDM literacy level	Frequency	Percentage
High	4	10
Moderate	2	4
Low	38	86
Total	44	100

Table 3

RDM teaching techniques

Teaching techniques	Frequency	Percentage
Distributed teaching techniques	20	13
Collaborative teaching techniques	32	19
Adaptive teaching techniques	39	23
Competency-based teaching techniques	39	23
Constructive teaching techniques	37	22
Total	167	100

Note(s): NB: Multiple responses were allowed

Table 4

RDM teaching formats

RDM topics	Frequency	Percentage
Presentation or seminars (online)	17	14
Presentation or seminars (live)	30	24
Online tutorials (video clips, information etc.)	12	9
Interactive workshops	6	5
Intensive training (online)	34	27
Intensive training (live)	26	21
Total	125	100

Note(s): NB: Multiple responses were allowed

Notes

1. http://cit-ec.de.

2. https://www.cit-ec.de/en/open-science/manifesto