Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

A two-tiered curriculum to improve data management practices for researchers

  • Kevin B. Read ,

    Contributed equally to this work with: Kevin B. Read, Alisa Surkis

    Roles Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation NYU Health Sciences Library, NYU Langone Health, New York, New York, United States of America

  • Catherine Larson,

    Roles Conceptualization, Project administration, Resources, Writing – review & editing

    Affiliation NYU Health Sciences Library, NYU Langone Health, New York, New York, United States of America

  • Colleen Gillespie,

    Roles Conceptualization, Formal analysis, Investigation, Methodology, Validation, Writing – review & editing

    Affiliation Institute for Innovations in Medical Education, NYU Langone Health, New York, New York, United States of America

  • So Young Oh,

    Roles Conceptualization, Methodology, Resources, Writing – review & editing

    Affiliation Institute for Innovations in Medical Education, NYU Langone Health, New York, New York, United States of America

  • Alisa Surkis

    Contributed equally to this work with: Kevin B. Read, Alisa Surkis

    Roles Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation NYU Health Sciences Library, NYU Langone Health, New York, New York, United States of America

A two-tiered curriculum to improve data management practices for researchers

  • Kevin B. Read, 
  • Catherine Larson, 
  • Colleen Gillespie, 
  • So Young Oh, 
  • Alisa Surkis



Better research data management (RDM) provides the means to analyze data in new ways, effectively build on another researcher’s results, and reproduce the results of an experiment. Librarians are recognized by many as a potential resource for assisting researchers in this area, however this potential has not been fully realized in the biomedical research community. While librarians possess the broad skill set needed to support RDM, they often lack specific knowledge and time to develop an appropriate curriculum for their research community. The goal of this project was to develop and pilot educational modules for librarians to learn RDM and a curriculum for them to subsequently use to train their own research communities.

Materials and methods

We created online modules for librarians that address RDM best practices, resources and regulations, as well as the culture and practice of biomedical research. Data was collected from librarians through questions embedded in the online modules on their self-reported changes in understanding of and comfort level with RDM using a retrospective pre-post design. We also developed a Teaching Toolkit which consists of slides, a script, and an evaluation form for librarians to use to teach an introductory RDM class to researchers at their own institutions. Researchers’ satisfaction with the class and intent to use the material they had learned was collected. Actual changes in RDM practices by researchers who attended was assessed with a follow-up survey administered seven months after the class.

Results and discussion

The online curriculum increased librarians’ self-reported understanding of and comfort level with RDM. The Teaching Toolkit, when employed by librarians to teach researchers in person, resulted in improved RDM practices. This two-tiered curriculum provides concise training and a ready-made curriculum that allows working librarians to quickly gain an understanding of RDM, and translate this knowledge to researchers through training at their own institutions.


Better data management on the part of researchers is recognized as a critical need by researchers, funders, and publishers [13]. Good research data management (RDM) practices provide the means to analyze data in new ways, more effectively build on another researcher’s results, reproduce the results of an experiment, and aggregate like datasets for analysis [46]. While the benefits of RDM are clear, researchers often overlook the importance of RDM throughout the research process. The reasons for this are well-documented [710]: researchers see no benefit to themselves in exercising good RDM practices, they do not believe anyone would want or be able to understand their data, grant and publication pressures leave them no time, and there is no money to support RDM. The goal of this project was to facilitate better RDM on the part of researchers through the development of concise online modules to provide librarians with the knowledge and comfort level to teach RDM, and a ready-made, flexible curriculum for librarians to use for training researchers at their own institutions.

Librarians, with their knowledge of metadata, preservation, and discovery, are recognized by many as a potential resource for assisting researchers with RDM [1118]. However, this potential has not been fully exploited, particularly in the biomedical research community. While this is in part due to institutional barriers and the failure of researchers to recognize librarians’ expertise in this area, a major barrier to fulfilling this potential lies with the shortcomings of resources available to librarians. Before embarking on this project, the authors disseminated a survey to health sciences librarians through professional listservs asking respondents if they saw a role for their library in teaching RDM, if they currently taught RDM and, if not, what they saw as barriers to doing so [19]. There were 118 survey responses, with 84% of respondents indicating they saw a role in teaching RDM at their institutions, but 75% indicating that they did not currently do so. Barriers identified included a lack of knowledge about RDM (60%), lack of comfort engaging with researchers around the topic (48%), and lack of satisfactory curricula to train researchers (44%) were barriers to supporting their own research communities.

While web-based RDM educational modules for librarians already exist [2022], until recently [23] none have had a biomedical focus or addressed librarians’ familiarity with the research process. Many of these online educational offerings require a more substantial time commitment [22, 23], which can be a significant barrier for working librarians. Another gap has been the lack of a ready-made, biomedically-focused RDM curriculum for use by librarians in training researchers. To fill the gaps that exist in available RDM training, we developed and piloted two curricula: 1) a web-based curriculum that teaches health sciences librarians about RDM, data, researchers and the biomedical research process [24] and 2) a toolkit consisting of slides, script, instructions, and an evaluation form for an introductory RDM class [25].

Materials and methods


The intent of the pilot project was to train health sciences librarians and provide them with the tools to teach RDM to researchers, specifically within the context of the research landscape in the United States (US). The criteria for pilot participation was therefore that learners be a) health sciences librarians, and b) working at an institution in the US. Recruitment of pilot participants was completed through emails to health sciences librarian listservs and newsletters disseminated through the National Network of Libraries of Medicine (NNLM). The platform that hosted the online modules was freely accessible, therefore the modules could be taken by anyone. As a result, learners who had not been directly targeted as part of the pilot recruitment process and who did not fit the pilot criteria found and took the modules. Because we had not anticipated this, none of the assessment questions embedded in the online modules directly assessed learners’ suitability for the study. We therefore filtered out the following from our initial sample (n = 89): 1) non-librarian users at any NYU domain (n = 16), 2) users at any non-US domain (n = 4), and 3) users whose free-text comments indicated they were not US-based health sciences librarians (n = 4). We also removed duplicate users (n = 2), keeping only their first response. The number of librarians completing each module and the number of institutions represented by those librarians is listed in Table 1.

The piloting of the Teaching Toolkit consisted of the project Principal Investigators (PIs) completing site visits to observe librarians using the Teaching Toolkit to teach an in-person class for researchers at their institution. Eligible librarians were those who had completed all seven modules and whose responses in the completion survey indicated that there was a possibility of them teaching within the timeframe of the grant. There were 18 librarians from 15 institutions who met this criteria. Three of these institutions were able schedule an RDM course within the timeframe of the grant, and therefore all three were selected for piloting.

Online modules

We created seven web-based modules for librarians, utilizing content based on our experience teaching RDM to librarians and researchers and designing the modules based on the cognitive science of learning theories to enhance educational effectiveness [2630]. The modules were published online using a platform developed at the NYU School of Medicine that allows for authoring, dissemination, and data collection of web-based learning modules [24]. Initially, we created seven modules, with the order, content, and objectives of those modules mirroring existing classes the project PIs taught to researchers starting in 2012 and health sciences librarians starting in 2014 as seen in Table 2.

Table 2. Online research data management education modules for librarians.

The modules included videos, text, and embedded questions to assess the following: 1) users’ experience of the modules, 2) changes in self-reported understanding and comfort level with the material, and 3) intent to use the knowledge gained. The questions were not drawn from a validated instrument as none exist designed for our purposes. However, the questions were developed in consultation with a collaborator (CCG) with evaluation expertise, adapted from instruments used within our medical school curriculum that have evidence of their reliability (internal consistency) and validity (expected associations among and across evaluation domains) across their use in different contexts (e.g., obtaining feedback on and evaluation of courses, clinical experiences, and online modules), and were pilot-tested and refined with experts and then librarians through several iterative cycles.

The modules could only be accessed in the order listed in Table 2 for the purposes of the pilot. Questions embedded before the first module assessed each librarian’s overall background and interest level in RDM. Questions embedded at the conclusion of each module assessed librarian satisfaction with the module and their self-reported change in understanding. Questions embedded after completion of all modules assessed librarians’ self-reported change in comfort level with the material, plans to teach RDM, and intent to use the material learned in other aspects of their work. The self-reported changes in understanding and comfort level were assessed using a retrospective pre-post design to correct for participants’ tendency to overestimate understanding and confidence at baseline and then re-calibrate more accurately after training [3134]. A four point scale was used to assess both self-reported understanding (no, minimal, moderate, strong) and self-reported comfort level (not, somewhat, mostly, very). Embedded questions and responses are included in S1 Data and S2 Data.

It was decided that, while not part of the original protocol, a more complete assessment of the strengths and weaknesses of the online modules could be gained through semi-structured telephone interviews with participants who had completed all seven modules. Because of the time required by this additional phase, we selected 14 interviewees, 2 based on their RDM expertise and 12 based on their suitability to be pilot participants for using the Teaching Toolkit at their own institution. Suitability criteria was based on librarians’ self-reported interest in, and intent to use the material in the online modules.

Teaching Toolkit

We developed a Teaching Toolkit which consisted of slides, a script, and an evaluation form for class attendees [25]. The material was designed to be used to teach a 60 to 90 minute introductory RDM class with the content drawn from the curriculum of the first seven online modules (Table 2). We piloted the Teaching Toolkit with librarians from three institutions. The two project PIs traveled to the institutions to observe the classes being taught and conduct semi-structured interviews with the librarian(s) who had taught the class.

The semi-structured interview (see S1 File) asked the librarians to reflect on the use of the Teaching Toolkit, describe further plans for use of the Teaching Toolkit, describe their professional background, and reflect on the suitability of the Teaching Toolkit for their particular audience. Interviews were designed to elicit feedback from the librarians that would elucidate the strengths and limitations of the Teaching Toolkit’s content, approach, and suitability for a range of instructors and audiences. The interviews were transcribed and a simple content analysis was used to code the major themes that emerged from the interviews regarding the strengths and limitations. Co-authors (KR and AS) discussed codes to ensure agreement and quotes representing identified strength and limitation codes are reported.

Class attendees at the three institutions were asked to evaluate the class (see S2 File), answering questions about their satisfaction with the material and their intent to use what they had learned. Seven months after the class, a follow-up up survey (see S3 File) was sent to researchers to assess whether they had actually used what they learned. Researcher data is included in S3 Data.

The collection of evaluation and interview data was approved by the NYU Langone Health Institutional Review Board.

Data analysis

We used a Wilcoxon signed-rank test to test the difference between self-reported understanding before and after each module, and between self-reported comfort level before and after the series of modules. In both cases, we used the one-sided test because, having done retrospective pre-post assessments, we did not expect decreases in self-reported understanding or comfort level. We used a normal approximation because of ties in the data. We used a chi-squared test of the effective ratings across the modules to determine if there was any significant differences in the learners’ perceptions of the effectiveness of the different modules. We otherwise used descriptive statistics to characterize responses.


Online modules


Responses to the embedded questions at the completion of each module, indicated that the majority of librarians found the modules to be mostly or highly effective (Fig 1). A chi-squared test of the ratings across the modules indicates that there is no significant difference in effectiveness across the modules (p = 0.18). Across all modules, 91% indicated that they found the level of the material to be “just right” and 91% indicated that the length of the module was “just right”.

Fig 1. Librarian rating of effectiveness of each of the seven online modules.

Using the Wilcoxon signed-rank test, we found the differences between self-reported understanding from pre- to post-module to be significant, with p < 0.0005 for each module. The mean difference in level of self-reported understanding varied by level of self-reported pre-understanding, with larger increases in self-reported understanding seen for those who reported having had no or minimal understanding of the content before the module (Fig 2).

Fig 2. Change in self-reported understanding, categorized by initial level of self-reported understanding, aggregated across modules.

The self-reported comfort level of learners (n = 27) increased from a median of 2 scale points (somewhat comfortable) to a median of 3 scale points (mostly comfortable), and the change in comfort level was seen to be significant using the Wilcoxon signed-rank test (p = 0.0002). Librarians change in self-reported comfort differed by initial comfort level: not comfortable (n = 6) mean increase of 1.33 scale points, somewhat comfortable (n = 15) mean increase of 0.87 scale points, mostly comfortable (n = 4) mean increase of 0.25 scale points, very comfortable (n = 2) mean increase of 0. Fig 3 shows the final self-reported comfort level for each librarian grouped by their initial comfort levels, and indicates which of the librarians felt they had sufficient knowledge to teach at the conclusion of the modules. Finally, 82% of librarians indicated that they would otherwise use what they had learned in their work.

Fig 3. Learner counts of final comfort level with RDM grouped by initial comfort level.

For each initial comfort level, final comfort levels are shown, and are grouped by whether or not the learner felt they had sufficient knowledge to teach RDM.

Post-module telephone interviews.

Telephone interview responses indicated that a common barrier to teaching RDM was a lack of knowledge about strategies for finding institutional avenues for teaching a class. We therefore created an eighth module to provide strategies for libraries to initiate RDM services locally. We discussed likely partners and strategies for implementation and included both those that had been successful in our own institution [10, 18] as well as other institutions [35]. Of those librarians interviewed, three were able to teach within the timeframe of the pilot, and so formed the pilot cohort for the Teaching Toolkit.

Teaching Toolkit

The Teaching Toolkit was piloted at three health sciences libraries. There were 16 total attendees across the three classes, with 10 of those consenting to the use of their evaluation data. Attendees self-identified as having the following roles: staff (4), faculty (3), postdoc (1), fellow (1), and student (1).

The attendees’ ratings of the level, length, and effectiveness of the class (Fig 4) indicated a high degree of satisfaction with the class. All ten attendees indicated that they had had no previous exposure to RDM educational materials, and all ten indicated they would either definitely or probably use what they had learned (Fig 4). The follow-up survey (see S3 File) administered to the attendees seven months after the class received five responses, and in all cases attendees reported that they used what they had learned. Four of the five responded to a free text question asking for a description of how they had used what they learned in the course, and all four stated that the course had helped them with the organization of their data and three specifically pointed to using best practices for improving file naming conventions.

Fig 4. Learner satisfaction reports.

Self-reporting of satisfaction with in-person RDM class for the following: level of class, length of class, effectiveness of presentation, whether the learner will use what they learned.

Our observations of the classes at each institution documented a range of presentation styles, with customizations made to the script and the slides. One instructor added more interactive elements to the class to encourage participation. Each pilot institution made visual modifications to the slides provided, added institutional resources for data storage and sharing, promoted library services around data management, and added examples of data management best practices to make the slides more engaging and better suited to their personal style.

Observing the in-person classes provided an opportunity to identify topics of particular interest to the audience, as well as elements of the presentation that could be improved upon. For example, all of the instructors had some level of discomfort with the material related to teaching the NIH rigor and reproducibility requirements. Based on this observation, we modified the content of the slides and script to streamline and clarify this component of the class. In addition, we made a number of other minor modifications to improve the clarity and flow of the class. We observed that the class attendees were the most engaged during the sections of the curriculum that discussed data organization, preservation, and standards.

Semi-structured interviews (see S1 File) conducted with the instructors (n = 4) from each pilot institution yielded a number of common themes. All instructors discussed the challenge of using material that they had not created. Instructors mentioned difficulties connecting with the material, specifically citing discomfort in the use of a script they had not written themselves. Despite the lack of comfort with the script, instructors felt strongly that the classes were successful, which was confirmed by the evaluation data (Fig 4). Pilot instructors all mentioned a desire to add interactive elements to the class. Instructors also stated that they saw value in all of the material included in the Teaching Toolkit, and did not plan to remove any of the topics discussed.


The curricula we developed provided significant innovation over existing data management education modules [20, 21, 36] in several areas. Our online modules were aimed at librarians and focused on biomedical research. These online modules included training on the processes, data, culture, and language of biomedical research to provide critical context that would allow librarians to overcome the barriers between librarians and researchers. The online modules were concise and directly tied to the Teaching Toolkit, a curriculum specifically created for use by the librarians to teach RDM locally, thus addressing the time constraints of working professionals seeking to enter this area.

One limitation of this study was the sample size, both for librarians taking the online modules and for researchers attending the Teaching Toolkit class. While the sample size of librarians would be low for online educational modules aimed at the general population, our modules were intended for the very specific population of U.S. based health sciences librarians. A rough estimate of the number of potential participants in this pilot is 300 (157 academic health sciences libraries, typically with no more than two librarians engaged in research data management services). Given that, our sample size of 63 librarians for the initial module and 27 librarians completing all modules constitutes approximately 20% and 10% respectively of the total population on which we were drawing. The small sample size of researchers attending the Teaching Toolkit classes is more problematic, and limits any generalizations that can be made based on that data. However, the unanimity of responses is encouraging, and results from a second pilot project by the authors, discussed later, support the conclusion that the Teaching Toolkit is effective and generalizable.

Online modules have been seen to have a high dropout rate; reported in multiple studies to be 90% or higher [3739]. We saw a 57% dropout rate between online modules 1 and 7. This dropout rate is low relative to what has been reported in the literature, but we acknowledge that it may have introduced bias into our results. We have some evidence that argues against this; during the pilot we conducted a survey of librarians who had expressed a strong interest in RDM but dropped out of the online modules. The two question survey (see S4 File) asked respondents to report the reason(s) for their non-completion, and elaborate on those reasons. The survey was distributed to 25 librarians and received 13 responses, 92% of whom indicated lack of time as one reason for non-completion, and for 62% of respondents, lack of time was the only reason indicated. The only other reason provided by more than one respondent was that it was not applicable to the respondents’ work (23%). These results were not unexpected since our target population was working librarians, who often must fit in professional development on their own time.

Another limitation is that we have no direct measurement of knowledge gain, since during the piloting process, a number of issues were uncovered in the modules’ knowledge gain questions, rendering that data unusable. We are therefore forced to rely on self-report of change in understanding of subject matter, however, our primary outcome was the librarians’ ability to teach an RDM class, not their overall knowledge of RDM.

A strength of this study is that the online modules proved to be effective in increasing librarians’ self-reported understanding of and comfort level with RDM. The increase in self-reported understanding for those with little or no understanding of the topic was greater than for those with previous experience. This outcome is expected as the modules were created to provide a concise introduction to RDM, rather than a deep dive. Future RDM education for librarians could modularize topics based on discipline, to provide training on targeted issues faced by specific research communities. The expressed need from the 14 phone interviews for additional guidance on initiating RDM services indicated that providing the knowledge and tools needed to teach a class only partially addresses the needs of librarians looking to provide services in this area.

Another strength of the study was that the Teaching Toolkit showed promise as an effective educational intervention; all respondents from the in-person class evaluation indicated that they would probably or definitely use what they had learned in their work, though the small sample size (n = 10) limits the conclusions we can draw. The positive reviews of the in-person class are particularly striking in light of the discomfort expressed by the instructors at teaching material that was not prepared in their own voice. Despite the awkwardness reported with use of the script, all instructors indicated that it was useful to have as a starting point. It is likely that the effectiveness will increase as librarians continue to customize the material to reflect their own teaching styles.

While the low attendance was a limitation, we do not believe this reflects an issue with the curriculum. In our own institution we have seen attendance at RDM classes grow from a low of three attendees in 2014 to over 100 signups for an RDM class in 2018. As guidelines and mandates from publishers and funders further increase the pressure on researchers to employ good RDM practices, and as the role of the library in supporting RDM is increasingly recognized within institutions, we would expect that class attendance would increase at other institutions as it has at ours. To this point, our pilot institutions have since reported additional trainings with much higher attendance.

Further validation of the value of our materials to the library community has come through two avenues. First, six of the eight online modules were used as a core component of a training program developed by the National Network of Libraries of Medicine (NNLM) titled Biomedical and Health Research Data Management for Librarians, which recruited 40 librarians from across the U.S. to take a comprehensive data management training over an eight-week period [40]. Second, both the online modules and the Teaching Toolkit formed the core of a subsequent NNLM-funded pilot project led by the PIs to facilitate the development of RDM services in libraries.

The NNLM-funded pilot project enrolled 26 librarians from six libraries [41]. This pilot resulted in every institution using the Teaching Toolkit to teach between one and five classes with a total of 294 attendees (average attendance = 21), where 99% of those who submitted an evaluation (n = 111) indicated that they would use what they had learned, and 95% (n = 112) indicated that they would recommend the class. Between the NNLM-funded pilot project and the project described in this manuscript, seventeen classes were taught across eight institutions (five health sciences libraries, three university libraries), to over 300 attendees from a wide range of backgrounds (e.g., health sciences, engineering, social sciences). These results indicate that the curriculum is broadly generalizable across a range of academic contexts and audiences.

The online modules and Teaching Toolkit provide effective, approachable RDM training and a ready-made, flexible, proven curriculum. The goal of our minimalist approach is to provide just enough background knowledge to equip librarians to provide effective trainings for researchers in the essentials of RDM, as well as sufficient context about researchers and the research process to increase the librarians’ comfort in engaging with researchers. This approach meets a critical need for practicing librarians with limited time who are seeking to initiate RDM services. While this does not provide the depth of RDM knowledge of more extensive curricula, we believe that researcher engagement resulting from teaching RDM classes is the most effective driver for working librarians to develop deeper RDM expertise. With the help of our curricula, institutions in search of a resource for supporting data management in their research community can look to their library for effective support.

Supporting information

S1 Data. Librarian self-reported satisfaction with and knowledge gained from each of the seven online modules.


S2 Data. Librarian self-reported satisfaction, change in comfort level and intent to use at the conclusion of all online modules.


S3 Data. Researcher self-reported satisfaction and intent to use material at the conclusion of the class taught by librarians using the Teaching Toolkit, and researcher report of actual use of material seven months post class.


S1 File. Semi-structured in-person interview form conducted after the completion of the in-person RDM class taught by pilot participants.


S2 File. Evaluation form administered to attendees/researchers at the completion of the in-person RDM class to gauge satisfaction with the material and their intent to use what they had learned.


S3 File. Follow up survey administered to attendees/researchers seven months after the in-person RDM class to ask how they used what they learned in their work.


S4 File. Survey administered to librarians who did not complete all seven online modules to ask reasons why they were not able to complete the entire curriculum.



This effort was made possible by the NIH Big Data to Knowledge initiative training grant R25-LM012283-02. We would like to thank Karen Yacobucci for her efforts in building video content for the online educational modules, Richard McGowan for his editorial input, and Fred Willie Zametkin LaPolla for his assistance with data visualization. We would also like to thank Dr. Jeremy Paul, Dr. Gyorgy Busaki, Dr. Samuel Solomon, Margaret Henderson, and Aileen McCrillis for their valuable feedback as advisory committee members.


  1. 1. Everyone needs a data-management plan. Nature. 2018;555(7696):286. Epub 2018/03/16. pmid:29542698.
  2. 2. Schiermeier Q. Data management made simple. Nature. 2018;555(7696):403–5. Epub 2018/03/16. pmid:29542709.
  3. 3. Barone L, Williams J, Micklos D. Unmet needs for analyzing biological big data: A survey of 704 NSF principal investigators. PLoS computational biology. 2017;13(10):e1005755. Epub 2017/10/20. pmid:29049281; PubMed Central PMCID: PMCPMC5654259.
  4. 4. Chan AW, Song F, Vickers A, Jefferson T, Dickersin K, Gotzsche PC, et al. Increasing value and reducing waste: addressing inaccessible research. Lancet (London, England). 2014;383(9913):257–66. Epub 2014/01/15. pmid:24411650; PubMed Central PMCID: PMCPMC4533904.
  5. 5. Gardner D, Goldberg DH, Grafstein B, Robert A, Gardner EP. Terminology for neuroscience data discovery: multi-tree syntax and investigator-derived semantics. Neuroinformatics. 2008;6(3):161–74. Epub 2008/10/30. pmid:18958630; PubMed Central PMCID: PMCPMC2663521.
  6. 6. Merelli I, Perez-Sanchez H, Gesing S, D'Agostino D. Managing, analysing, and integrating big data in medical bioinformatics: open problems and future perspectives. BioMed research international. 2014;2014:134023. Epub 2014/09/26. pmid:25254202; PubMed Central PMCID: PMCPMC4165507.
  7. 7. Bardyn TP, Resnick T, Camina SK. Translational Researchers’ Perceptions of Data Management Practices and Data Curation Needs: Findings from a Focus Group in an Academic Health Sciences Library. Journal of Web Librarianship. 2012;6(4):274–87.
  8. 8. Tenopir C, Allard S, Douglass K, Aydinoglu AU, Wu L, Read E, et al. Data sharing by scientists: practices and perceptions. PloS one. 2011;6(6):e21101. Epub 2011/07/09. pmid:21738610; PubMed Central PMCID: PMCPMC3126798.
  9. 9. Tenopir C, Dalton ED, Allard S, Frame M, Pjesivac I, Birch B, et al. Changes in Data Sharing and Data Reuse Practices and Perceptions among Scientists Worldwide. PloS one. 2015;10(8):e0134826. Epub 2015/08/27. pmid:26308551; PubMed Central PMCID: PMCPMC4550246.
  10. 10. Read KB, Surkis A, Larson C, McCrillis A, Graff A, Nicholson J, et al. Starting the data conversation: informing data services at an academic health sciences library. Journal of the Medical Library Association: JMLA. 2015;103(3):131–5. Epub 2015/07/28. pmid:26213504; PubMed Central PMCID: PMCPMC4511052.
  11. 11. Deardorff A, Florance V, VanBiervliet A. Assessing the National Library of Medicine's Informationist Awards. Journal of escience librarianship. 2016;5(1). Epub 2016/11/08. pmid:27818846; PubMed Central PMCID: PMCPMC5096841.
  12. 12. Kerby EE. Research data services in veterinary medicine libraries. Journal of the Medical Library Association: JMLA. 2016;104(4):305–8. Epub 2016/11/09. pmid:27822153; PubMed Central PMCID: PMCPMC5079493.
  13. 13. McEwen L, Li Y. Academic librarians at play in the field of cheminformatics: building the case for chemistry research data management. Journal of computer-aided molecular design. 2014;28(10):975–88. Epub 2014/07/21. pmid:25038898.
  14. 14. Medina-Smith A, Tryka KA, Silcox BP, Hanisch RJ. Librarians and Scientists Partner to Address Data Management: Taking Collaboration to the Next Level. Digital library perspectives. 2016;32(3):142–52. Epub 2016/11/29. pmid:27891247; PubMed Central PMCID: PMCPMC5120403.
  15. 15. O'Malley D, Delwiche FA. Aligning library instruction with the needs of basic sciences graduate students: a case study. Journal of the Medical Library Association: JMLA. 2012;100(4):284–90. Epub 2012/11/08. pmid:23133328; PubMed Central PMCID: PMCPMC3484943.
  16. 16. Pollock L. Data management: Librarians or science informationists? Nature. 2012;490(7420):343. Epub 2012/10/19. pmid:23075976.
  17. 17. Read KB, LaPolla FWZ. A new hat for librarians: providing REDCap support to establish the library as a central data hub. Journal of the Medical Library Association: JMLA. 2018;106(1):120–6. Epub 2018/01/18. pmid:29339942; PubMed Central PMCID: PMCPMC5764577.
  18. 18. Read KB, LaPolla FW, Tolea MI, Galvin JE, Surkis A. Improving data collection, documentation, and workflow in a dementia screening study. Journal of the Medical Library Association: JMLA. 2017;105(2):160–6. Epub 2017/04/06. pmid:28377680; PubMed Central PMCID: PMCPMC5370608.
  19. 19. Read KB, Surkis A. A 2015 Survey of Health Sciences Librarians Attitudes towards Research Data Management Education. 3 ed. Figshare2018.
  20. 20. Henkel H, Hutchison V, Strasser C, Rebich Hespanha S, Vanderbilt K, Wayne L, et al. DataONE Education Modules: DataONE; 2012 [cited 2018 January 5]. Available from:
  21. 21. EDINA, Library D. MANTRA: Research Data Management Training: University of Edinburgh; 2013 [cited 2016 November 18]. Available from:
  22. 22. Tibbo H, Jones S. Research Data Management and Sharing. Coursera2017.
  23. 23. Martin E, Goldman J. Best Practices for Biomedical Research Data Management Canvas Network: Harvard; 2018. Available from:
  24. 24. Read KB, Larson C, Oh S, Gillespie C, Yacobucci K, Surkis A. Research Data Management Training for Information Professionals Compass Learning System: NYU Langone Health; 2016. Available from:
  25. 25. Read KB, Surkis A. Research Data Management Teaching Toolkit: Figshare; 2018 [v6]. Available from:
  26. 26. Cook BG, Smith GJ, Tankersley M. Evidence-based practices in education. APA educational psychology handbook, Vol 1: Theories, constructs, and critical issues. Washington, DC, US: American Psychological Association; 2012. p. 495–527.
  27. 27. Mayer RE, Moreno R. Nine ways to reduce cognitive load in multimedia learning. Educational psychologist. 2003;38(1):43–52.
  28. 28. Mayer RE. Applying the science of learning to multimedia instruction. Psychology of learning and motivation. 55: Elsevier; 2011. p. 77–108.
  29. 29. MB Janssen-Noordman A, Merriënboer JJ, Van der Vleuten CP, Scherpbier AJ. Design of integrated practice for learning professional competences. Medical teacher. 2006;28(5):447–52. pmid:16973459
  30. 30. Sweller J. Human cognitive architecture: Why some instructional procedures work and others do not. 2012.
  31. 31. Bhanji F, Gottesman R, de Grave W, Steinert Y, Winer LR. The retrospective pre-post: a practical method to evaluate learning from an educational program. Acad Emerg Med. 2012;19(2):189–94. Epub 2012/02/11. pmid:22320369.
  32. 32. Revathi M, Vijayalakshmi B, Rajaratnam N, Chandrasekar M. Analysis of the Retrospective Self-Assessment Questionnaire of a Faculty Development Programme. South East Asian Journal of Medical Education. 2015;9(2):9–14.
  33. 33. McLeod PJ, Steinert Y, Snell L. Use of retrospective pre/post assessments in faculty development. Med Educ. 2008;42(5):543. Epub 2008/04/17. pmid:18412924.
  34. 34. Nimon K, Zigarmi D, Allen J. Measures of Program Effectiveness Based on Retrospective Pretest Data: Are All Created Equal? American Journal of Evaluation. 2010;32(1):8–28.
  35. 35. Federer L. The Medical Library Association guide to data management for librarians: Rowman & Littlefield; 2016.
  36. 36. Kafel D, Creamer AT, Martin ER. Building the New England Collaborative Data Management Curriculum. Journal of escience librarianship. 2014;3(1):7.
  37. 37. Daniel J. Making sense of MOOCs: Musings in a maze of myth, paradox and possibility. Journal of interactive Media in education. 2012;2012(3).
  38. 38. De Freitas SI, Morgan J, Gibson D. Will MOOCs transform learning and teaching in higher education? Engagement and course retention in online learning provision. British Journal of Educational Technology. 2015;46(3):455–71.
  39. 39. Rivard R. Measuring the MOOC dropout rate. Inside Higher Ed. 2013;8:2013.
  40. 40. Zhao S. Biomedical and Health Research Data Management for Librarians: National Network of Libraries of Medicine; 2018. Available from:
  41. 41. Surkis A, Read KB, editors. A Pilot Project to Facilitate the Development of Data Services at Health Sciences Libraries. Medical Library Association Annual Conference; 2018; Atlanta, GA.