Case study 1 |
Confidence Intervals |
|
School or Department |
School of Medicine |
|
Institution(s) involved |
University of Southampton |
|
Contact + Email |
Dr Trevor Bryant (T.N.Bryant@soton.ac.uk) Lindsay Wood (lindsay.wood@ncl.ac.uk) |
|
Date |
1st March 2010 |
|
Tags |
Medicine, medical statistics, research methods postgraduate, repository usability |
|
Questions |
Explanation and further information |
|---|---|
|
1. What is the curriculum context of the resource or resource collection? |
Powerpoint presentation on Confidence Intervals used with postgraduate (PhD & MSc) medical students. |
|
2. What were the aims and objectives of the resource or resource collection? |
The aims were for students to understand how to interpret, calculate, use and apply confidence intervals. |
|
3. How was the resource or resource collection implemented? |
PDF file of a PowerPoint presentation including examples taken from the Medical Literature. |
|
4. What technologies and/or e-tools were needed to deliver this? |
Microsoft PowerPoint and subsequent PDF creation software. Visual Understanding Environment (http://vue.tufts.edu/) was used to construct decision tree maps for guidance package advice. Open Labyrinth (http://sourceforge.net/projects/Open Labyrinth/) was used to create an online application to deliver the decision tree maps. SurveyMonkey (http://www.surveymonkey.com/) was used to survey interested parties and collection data on their methods used in pedagogy and resource discovery. EdShare was used as a institutional repository to which the resource was originally uploaded (http://www.edshare.soton.ac.uk) JorumOpen (http://www.jorum.ac.uk/ ) was used as a repository to which this learning resource was uploaded to. |
|
5. What guidance and/or support did you develop? |
Categorisation guidance was followed. The categorisation technology should auto detect file format and size. The categorisation system should use very clear language and give examples of what is expected. Drop down menus, that included controlled vocabulary, with the option for ‘Other’ field and free text are needed. Auto-completion should be maximized (i.e. use of Ajax) to fill in terms from those already in the database. Clinical, pre-clinical, F1/F2 are the most useful terms to use instead of years of study. The system should encourage upload by being as usable as straight forward as possible. Data captured during categorisation will need to be stored and should flow through the guidance packages, without the need to repeat data entry. Preliminary patient consent guidance was followed. As this resource does not involve patient data, limited use of the guidance package was made. IPR/Copyright guidance was followed. The resource has a single point of IPR belonging to the author, a University employee, as dictated by the individual’s contract. Institutional Policy guidance was not available. There was no institutional policy on the release of OER. The time scales required to establish new institutional policies and the administrative burden tracking down ownership of IPR for ex-employees whose exact contractual status at the time of authoring a learning resource is a barrier to the timely release of OER. A ‘traffic light’ approach is being considered, which classifies a resource on an OER readiness scale, whereby the degree of compliance with the guidance packages is weighed against the likely risk of litigation. It is suggested that in lieu of an Institution Policy, all involved parties will be contacted and asked if they object to the resource being made open. Internationalisation guidance was not available. Pedagogy/QA guidance preliminary survey was completed and detailed feedback given on usability and which of the questions provided the most valuable information to be applied as metadata. Resource Discovery/Re-use guidance was not available. The preliminary survey was completed. Resource Upload guidance was followed. The resource was assigned a CC Attribution Non-commercial No Derivatives License. |
|
6. Uploading and hosting resources. |
The resource was successfully uploaded as an OER to the JorumOpen repository by a Senior Lecturer and eLearning staff member. The JorumOpen repository system does not meet requirements. The metadata collection is minimal. A future repository will likely need lots of metadata to discriminate between resources if there are many of resources returned in a search. The repository needs to be very easy to use and not expect depositors/end users to learn how to use the system. All usability approaches to encourage resource uploads need to used. There needs to be a batch upload system and metadata data needs to be pre-populated. It is not clear of the purpose why FE and HE are differentiated in JorumOpen. The environment should provide search and deposit options separately. The CC button icons (external links) should be separated from the radio buttons to avoid confusion. The terminology used is unclear. For example, the user needs to learn by trial and error what action ‘Save as draft’, ‘Publish’ and ‘Exit’ perform exactly. JorumOpen should validate URL links to prevent addition of ‘dead links’. There is no template system to allow for rapid addition of content. All warnings should all be in red giving a consistent style. A takedown time of 48 hours in too long, there needs to be instant removal by the resource uploaded. This is critical for patient based material. There needs to be additional re-use/download statistics. In summary, JorumOpen has more usability issues compared to modern repository sites. |
|
7. What are the key outcomes of the resource or resource collection? |
Clarification of resource authors’ IPR based on contract status. A case-by-case approach to making resources open in an Institution where a policy does not exist. |
|
8. What follow-up activity will be/has been carried out as a result of the resource or resource collection? |
None. But the depositor would like to be able to know who has accessed and used the resource. |
|
9. What are the lessons learned from the resource or resource collection? |
It will not be sustainable to update resources in multiple repositories. Updating in an Institutional repository is the preference. Tracking down and resolving IPR issues and approving a novel Institutional Policy is very time consuming. A solution might be the proactive release of OER resources by contacting interested parties and inquiring if there are any objections, together with a rapid take down policy. The issue of what justification there is for a centralised repository and how competitive it is compared to other open sharing sites (Institutional repositories, YouTube, Slideshare, etc) was raised The risks are perceived as releasing an OER that someone objects to. Following toolkit guidance should avoid this and a robust take down policy will minimise the impact. Also, how to maintain and protect Institutional branding on share resources. The benefits are to promote the University and to promote the individual resource creator. |