The 10 commandments of experimental data

Here is the original French version by Charles Nepote:
Les 10 commandements de l'expérimentation data

  1. À apprendre, pas forcément à réussir, tu chercheras
  2. Des hypothèses tu formuleras ou l’exploration tu assumeras
  3. Tes réussites comme tes échecs tu partageras
  4. À la diversité des profils et contributeurs tu veilleras
  5. Frugalité, agilité, simplicitétu chériras
  6. Un accès aux données tu obtiendras
  7. De l’intérêt des données tu ne préjugeras pas
  8. Dans des univers nouveaux, les données tu chercheras
  9. Face aux données, un esprit critique tu garderas
  10. À ces principes tout le monde adhérera

And my quick-and-dirty English translation:

  1. You will seek to learn rather than succeed
  2. You will formulate hypotheses or take an exploratory approach
  3. You will share both your successes and your failures
  4. You will be alert to diversity among candidates and contributors
  5. You will seek frugality, agility, and simplicity
  6. You will make your data accessible
  7. You will not prejudge the appeal or point of your data
  8. You will seek data in new fields
  9. You will maintain a critical eye when faced with data
  10. Everyone will adhere to these principles

Sounds like an excellent manifesto.  Feel free to improve upon my translation!

Conflicting priorities on information security

EBSCO have just released a White Paper “from our partner, OpenAthens”, The Evolution of Authentication and the Importance of Information Security.

The focus is very much on the information security of EBSCO’s subscription content.  There is no mention of user privacy, despite the fact how individuals want their data to be used is often in conflict with how corporations want to use this information.

Rather like the Leave campaign’s messages that voting for Brexit would be all gains and no losses, ignoring the complexity of complex decisions creates blind spots and vulnerabilities in systems and societies.  I would like politicians and corporations to stop patronising us with simple, comforting, false solutions and engage bravely and intelligently with difficult decision-making.

Observe what happens if you click on “Download your copy for free today to continue reading”:

Please fill out the form to receive your free copy of The Evolution of Authentication and the Importance of Information Security. Fill out the form and immediately receive the white paper. The fields requested are: Name, Email, Organization Name, City, Phone. All fields except Phone are required.

je dis ça, je dis rien

See also: EBSCO EDS and Single-Sign On, and Consumer democracy? (reference to Adam Curtis’ film Bitter Lake, describing how politicians create oversimplified good vs evil stories rather than confronting the realities of a complex world).

Journey to Full Text Finder – arrival in the Celestial City

Following my presentation Journey to Full Text Finder: A Pilgrim’s Progress at the EDS conference in July, here’s an update on how I got on with the migration from the old EDS to the new Full Text Finder (FTF) version.  Thanks again to John Bunyan’s The Pilgrim’s Progress (1678) for the inspiration for the title of this post.

On the whole, everything went smoothly, and I would particularly like to thank Seoud, Abid, and Adam at EBSCO for their help throughout the process. I have written this summary to help other people know what to expect from the process, and in particular the cascade effects on data checking and linking for which it is essential to set time set aside.

Side by side

Our migration to EBSCO Discovery Service Full Text Finder began with a visit from Seoud.  He talked me through the steps, and we agreed the timings.  We also discussed running the old and new EDS alongside each other, to allow students completing their courses in July and August to continue using the old version, while having the new version available for testing and experimentation before the beginning of the new academic year in September.

I asked a colleague in IT to create a redirect URL for EDS FTF which I then used wherever required, and this saved time later in updating URLs individually.

Data migration and checking

Following the data migration, our Periodicals Librarian spent time checking that our subscriptions in EDS FTF matched the old system, focusing on a few known trouble spots e.g. where we have single title subscriptions rather than a whole package. In some cases, a journal which existed from e.g. 1997-present but for which we only have access from 2015-present had been enabled for the full run, and not the years to which we have access.  This was particularly frustrating for titles for which our subscription is administered through EBSCONET, as EBSCO clearly have correct information in their system about our entitlements, but it was not being migrated or applied accurately.  Although this happened only for a small number of titles, there did not appear to be any pattern to predict which would be affected, and so all had to be checked.

Admin interfaces

Our test EDS FTF was up and running in June.  The old EDS was controlled through two admin interfaces – EBSCOadmin to control EDS itself, and A-to-Z Admin to control journals and databases A-to-Z lists.  The areas of overlap (and not) of the two admin mechanisms were sometimes unclear to me. The new EDS FTF is administered via EBSCOadmin, and it is great to have just one admin interface to drive this system.  As we had old and new EDS running simultaneously for a period, any changes would have needed to have been made to each system separately.  I decided that from the launch of new EDS FTF, any changes would only be made to the new EDS, and this did not cause any problems.

ebscoadmin

Databases A-Z

The new EDS FTF has access to journal titles via the Publications link, but it has no Databases A-Z feature.  I have been told by EBSCO that they did not include this feature, because only librarians wanted it, not students (and also perhaps because there is less demand for it in EBSCO’s primary market, the USA).  However, there are some essential databases which are not indexed by EDS (such as Westlaw) and our users must have a route to access these.  EBSCO have an A-Z solution which can be added to your EDS FTF but you have to request it, and it is basic (just a list of links, like in the old days…).  It also has its own admin interface.

a-z

Linking

Permalinks created in the old EDS are different from permalinks in the new EDS FTF.  EBSCO have redirects in place, but “there is no timescale of how long these will be in place”.  It is therefore necessary to create new permalinks anywhere these are used, such as reading lists.  We also had links to journal titles using a linking template that worked with the old journals A-to-Z, and these had to be re-created based on the new “Publications” journal-finding tool in EDS FTF.  This was more urgent, as no redirects would be in place for the old journals A-to-Z.  This can add considerably to the workload of staff who maintain reading lists.

permalink

Google Scholar

If your library is set up for “Library links” to allow your users to use the library’s link resolver with Google Scholar, your settings will need to be updated to reflect the new resolver within EDS FTF.  EBSCO told me that this update would be included in the migration.  When I contacted EBSCO for confirmation, they confirmed that our resolver in Scholar had been updated to Full Text Finder, but that it could take 1-2 weeks for the changes to take effect, suggesting that my enquiry had prompted a change in the settings rather than this happening without my intervention.

scholar

E-Resources FAQ

This is a collection of things I wish everyone knew about e-resources.  Whether this area is new to you or not, I hope you find something useful here; and do let me know about any points I’ve missed in the comments.

What are e-resources?

E-resources are also known as electronic resources and there are two main types: e-journals (or electronic journals) and databases.

Many e-journals are digital copies of print journal articles, but increasingly e-journal articles are published without a print analogue.

There are several kinds of databases

  • Bibliographic – this type of database is a collection of references to published literature.  It functions in a similar way to a library catalogue, but indexes details of articles rather than books
  • A&I (abstracting and indexing) – in addition to bibliographic details, this type of database also contains abstracts of the individual articles
  • Full text – a database which includes the full text of all the articles it has indexed
  • Data/statistics – a collection of numbers and facts which you can query in order to extract a particular dataset.  A database in the purest sense of the word.
  • Images – a database containing a searchable index of images and the images themselves

What does full text mean?  Full text refers to an e-resources that makes available online the whole contents of journal articles, not just the abstract or citation.  Full text articles are often subscription resources, requiring an individual or institutional account for access.

What is an abstract?  An abstract is a summary of a journal article, often published at the beginning of the article.

What is a platform? A platform is a website which hosts content or programs.  Examples include JSTOR and ISI Web of Knowledge (which hosts a number of databases including, confusingly, Web of Science).

What is SFX?  SFX is an OpenURL link resolver, which works by compiling a list of all the journals to which an institution (such as a university) is subscribed and linking to that content.  Primarily, it functions to allow you to search an institution’s subscriptions to see if you can access a particular e-journal, and which years are included in the subscription.  At Oxford University, SFX is locally branded as OU eJournals and is one of a number of resources whose contents are searchable via SOLO.

What is MetaLib?  MetaLib is a search system which allows you to search for resources, link to them, and (in some cases) search within them.  This is not possible for all resources, as they need to be compliant with a protocol called Z39.50 in order to be searchable.  At Oxford University, MetaLib is locally branded as OxLIP+ and is one of a number of resources whose contents are searchable via SOLO.

What is a paywall?  A paywall is a barrier to a website which requires you to authenticate to view the content.  Usually, this requires a paid subscription.  An important implication of this is that any content behind a paywall is not indexable by search engines and therefore will not appear in the search results.  Not everything on the Internet is known to Google.

There are several methods of authentication

Internet Protocol (IP) – the IP address of your computer identifies where you are in the world, and is also used by sites like BBC iPlayer which use your IP address to check which country you are in.  If you are using the university’s computing facilities on campus, the computer you’re using will have an IP address within the university’s main range, which is detected by the e-resource you are trying to reach and access will be granted.  Working “off-campus” means that you are off the university network, perhaps using your own laptop in a university library or working from your own home.  This means that your computer’s IP address is not within the institution’s IP range and you will need a different method of access.  VPN software is commonly used to solve this issue and it works by extending the institution’s network to your computer, thereby bringing it into its IP range.

Want to find out your IP address?  Just go to whatismyipaddress.com

Single sign-on (SSO) – logging in via SSO identifies you as a member of an institution (such as a university) and therefore allows you access.  A great advantage of SSO login is that your authentication can be pushed from one site to another via your browser, so you don’t have to keep logging in when you go to a different subscription site that accepts SSO authentication.

Username and password – the old school method.  Nowadays, this only really applies to a small number of really expensive resources, where tight budgets or low demand mean that a several-user subscription than whole-campus access has been purchased.  There may only be (for example) 5 usernames and passwords for the resource, and if all 5 are in use, you will need to wait until someone has logged out so that you can use that ID to log in afresh.

Also good to know

What is a session identifier?  Session IDs or tokens are commonly used in online shopping sites and data/statistics databases.  These types of sites combine a variety of information to produce the page you are viewing, rather than retrieving a pre-prepared HTML page.  The session ID is used to track the individual user’s actions during the course of their session on the site.  Your shopping cart contents or dataset only exists because you have selected and combined certain elements during the session, which will time out after an order is finalised, or the user logs out, or after a period of inactivity.

URLs which contain “session” or “sid” indicate a session ID, and are not persistent.  If you are attempting to link to a resource, check the URL: if it contains a session ID, the URL will not work when someone tries to follow it later on because the session will have timed out.

Some e-resources have embargoes which are periods during which access is not allowed (usually to protect the publishers’ interests, or in JSTOR’s words “protect the economic sustainability of our content providers”).  There are several types of embargo:

  • A rolling or moving wall – a fixed period of months or years.   For example, most journals in JSTOR have an embargo of 3 or 5 years, and as a new issue is published, its equivalent from 3 or 5 years before will become available on JSTOR.
  • An annual cycle – for example, all content before 1st January of this year is available.  This will add another year to the archive on 1st January of each year
  • A fixed date – for example, only content before 2005 is available

If you’re carrying out research in your subject area, make sure you don’t rely exclusively on resources with embargoes, as you will be missing current and recent material.

E-resources and copyright – keep your use legal!

Most e-resources publishers have a ‘fair dealing’ arrangement which allows you to print or save one article per journal issue.  Downloading an article happens when you view the article on screen, not just if you save it.  Please be aware that systematic downloading is not permitted under fair dealing arrangements and may compromise your institution’s access to the resource.  Also, remember that your access to e-resources is for your own research and learning only, and you may not email pdfs or other downloaded documents to anyone outside your institution.

See also: E-Resources – less frequently asked questions for the next part of the story…

Librarians as researchers: that’s a good IDEA

I really enjoyed this session, led by Miggie Pickton (University of Northampton) and Carolynn Rankin (Leeds Metropolitan University).

What does research look like?  Everyday research skills include: reading, watching, questioning, summarising, presenting, listening, choosing, organising, writing up, reflecting.  Many of us are already doing research, but maybe we just don’t realise that it is research!

Research is the professionalisation of everyday skills (Blaxter, 2008)

Library practitioners are often highly innovative in their practice and undertake research-related activity as a normal part of their working lives.

This new knowledge and understanding is often not recognised as research nor is it shared with the wider professional community.

We did an icebreaker exercise to meet each other and learn about the types of research activity we had each been involved in:

Name

Library service

What did they do?

Has had to provide evidence of service value
Has engaged with an external quality benchmark
Has and to collect statistics for annual reviews
Has run a focus group
Has written and/or presented a report to their organisation
Has helped a service user find resources for their research
Has contributed to a publication
Has explored ways of improving their service

Miggie and Carolynn introduced the framework for developing your research:

I=interest, issue, idea
D=develop, discuss, define
E=engage, elaborate, enact
A=advocate, advertise, apply

I=interest, issue, idea

  • Identify a project or research opportunity that interests you or meets a need
  • What do I want to know?
  • How could this help my practice or benefit my organisation?
  • What’s in it for me?

D=develop, discuss, define

  • Define the research question
  • What has been done on this before? Where is the evidence base? Where are the gaps?
  • Develop the project proposal – SMART objectives, appropriate methods

E=engage, elaborate, enact

  • Partnership and connections
  • Look for common goals
  • Win-win agenda
  • Who will you engage with and how?
  • This might be partners, colleagues, management, funders, policymakers

A=advocate, advertise, apply

  • Who needs to know about your work? Service users, managers, funders, policymakers
  • Where will your research make a different?
  • Effecting change within and beyond the library

S=Skills

Finally, the multiplier effect comes in when you add skills.

This session was practical and energising, and it started me thinking about the many ways I could apply these ideas to my work.

Knowledge management

Here is a summary of a recent Mind Tools article about knowledge management that I found interesting.

What is knowledge?

Data is a specific fact or figure, without any context; information is data that’s organized; knowledge builds on the information to give us context (remember this from library school?).

There are two different types of knowledge, explicit and tacit:

Explicit knowledge includes things that you can easily pass on to someone else by teaching it or writing it into a record. This kind of knowledge can be captured in a staff handbook or workflow.
Tacit knowledge is less concrete. It may relate to the best way to approach a certain person for their help or co-operation, or how to unjam the photocopier. This type of knowledge is usually acquired by experience.

Why is knowledge management important?

Sharing information within a team or department means that when a person is away on holiday or off sick, or moves to a new job, their knowledge persists in the organisation.
As well as time and cost savings, an environment which fosters the sharing of ideas can help increase innovation, build trust and improve relationships.

Implementing knowledge management

There are two different ways of managing knowledge: using technology-based systems, or using softer systems.
Examples of technology-based systems include a co-authored staff handbook or wiki. It is easy to access this information, but it takes effort to keep it up-to-date. Good for capturing explicit knowledge.
Examples of softer systems are shadowing or mentoring. Better for sharing tacit knowledge.
A successful knowledge management strategy should try to use both approaches.

Tips for implementing knowledge management systems

  • Identify tacit knowledge, then brainstorm ways of sharing it
  • Start with a small team to avoid information overload
  • Some people may feel uncomfortable about sharing their hard-won knowledge, so make knowledge exchange part of the organisational culture and find ways to reward people for sharing it freely
  • Make the processes of capture or sharing easy. Easier participation makes for increased involvement and success

Gather ye data while ye may

Many of the tasks librarians do are now becoming better-known outside the profession:  reader services (managing lending and dealing with enquiries), technical services (cataloguing), making decisions about purchases and deselection, teaching information literacy, administering budgets, working on policies, marketing the library and its collections and services…

However, gathering data is a really important part of running a library, and I would like to give it some attention today.

Collecting information about how a library service works is important in order to evaluate its effectiveness and inform its future direction.  Much of this information is captured in numerical form.

At the moment, I am creating data snapshots of the following:

  1. How fast returned books are re-shelved.  Measuring how many books library staff can re-shelve in an hour allows me to plan my team’s time and adapt to busy periods by increasing shelving hours when I know that rates of return will be high (at the end of the academic year, in particular).
  2. Usage of print periodicals: in my library, we take around 40 academic journal titles in print form (as well as online).  By asking library users to return a paper slip each time they use an issue of a print periodical, informed decisions can be made in the next budget year about the cost/benefit of maintaining each print subscription.
  3. Demands of enquiry work on library staff.  In my library, we do not have a reference/enquiry desk, but the library office door is always open and students are encouraged to come and ask us for help.  We also respond to queries via email and phone calls.  The absence of an enquiry desk can give the impression that the library does not handle enquiry work, so keeping a record of the types of enquiries and time spent answering them is useful when compiling a summary of the work done by library staff each year.

Gathering data is a time-consuming activity, so I aim to create data snapshots by measuring certain factors during particular intervals, for example 2 weeks out of a given term.  Although it is difficult to choose typical periods to monitor, and there will always be some confounding factors, I find that it helps the team to focus on the data-gathering projects for short periods rather than having to record data on everything we might measure every day.

For data which can be extracted from the library management system, I do record a variety of statistics throughout the year, for example:

  1. Checkout and checkin stats by membership group and material type – these numbers show how many items of different types (e.g. books; DVDs) are borrowed and returned by different segments of the library user community (e.g. undergraduate students; academics)
  2. My budget! I keep a master spreadsheet of all expenses and a very few incomes relating to the library budget.  Each amount is given a cost code which indicates its sub-budget e.g. Books, Periodicals, Library Materials.  Within the Book cost code, each amount is further broken down by subject area.  All of this information is then included in a pivot table in Microsoft Excel, which I use to see the balance remaining in each cost centre.

One of the things I enjoy most about my work is tidying up a messy collection of information in Excel and producing a clear summary of what it means.

Ned, I take your point that by only measuring what is happening in our physical library spaces, we are not taking account of changing behaviour and increasing use of libraries’ online facilities.   However, the role of my library as a physical space is still its most important aspect, so I’m starting with that for now.