Subscribe: News and activities
http://www.celtic-project.org/Project_blog!rss
Added By: Feedage Forager Feedage Grade B rated
Language:
Tags:
application  consumer  feedback  lti  provider  student  students  tool provider  tool  user  users  vle  webpa  workshop 
Rate this Feed
Rate this feedRate this feedRate this feedRate this feedRate this feed
Rate this feed 1 starRate this feed 2 starRate this feed 3 starRate this feed 4 starRate this feed 5 star

Comments (0)

Feed Details and Statistics Feed Statistics
Preview: News and activities

News and activities



This public blog is used to record the progress of the project - feel free to contribute feedback and suggestions etc. Click here for the RSS feed



 



Managing the LTI Lifecycle

Thu, 05 Jan 2017 14:48:17 GMT2017-01-05T14:48:17Z

Under the auspices of both IMS and the ceLTIc project, Simon and Stephen delivered another LTI workshop at the Blackboard Users' Conference in sunny Durham (UK) yesterday.  It sought to focus on the practical issues of deploying LTI tools in VLEs:

One of the main benefits to suppliers is that LTI is built into every VLE and so they only need to write a single connection.  However, one of the disadvantages to customers is that it is not possible to have the same depth of integration as is possible with a bespoke building block. Thus LTI connectors can require more careful management by staff to ensure they work well for users over time.

The main takeaways from this workshop were:

  1. how to best manage the lifecycle of LTI tools;
  2. questions to ask suppliers about how they support the lifecycle (or questions for suppliers to ask themselves about how best to support the lifecycle for their customers);
  3. latest news from the LTI community.

Thanks to Atomic Learning and Softchalk for allowing us to use their applications as case studies and to all the particpants for the interesting dialogue which took place.  The slides summarising the workshop are attached. LTI-durham17.pdf




LTI in Action

Wed, 07 Jan 2015 19:07:02 GMT2015-01-07T19:07:02Z

(image) The ceLTIc Project delivered once again at the annual Durham Blackboard Users' conference.  It fed both body and mind (and soul?) with LTI-branded M&Ms and an LTI in Action workshop which illustrated the latest developments with LTI and the open source building block using real-world examples from:

Many thanks to these organisations for their support.

If you would like us to run an LTI-related workshop for your own organisation, please contact Simon or myself.




LTI: Best Practice Guide

Wed, 13 Aug 2014 11:47:02 GMT2014-08-13T11:47:02Z

The ceLTIc project is pleased to announce the release of a Best Practice guide for LTI (http://ltiapps.net/guide/best_practice/).  This is largely based on the experiences gained from the sample LTI applications and connectors for existing applications which the project has released over the past few years.  Most of these have used the class libraries developed for PHP and Java which have sigificantly reduced the effort required to add LTI functionality to applications.  Whilst this guide has been released in a PDF format, it is intended to be a working document and a wiki-based version will be available shortly to which you are invited to add your own experiences, suggestions, comments and corrections so that it can continue to evolve and be truly representative of best practice across the LTI community.




LTI and "dummy" user accounts

Thu, 07 Aug 2014 22:27:45 GMT2014-08-07T22:27:45Z

Simon Booth and I had an interesting conversation today, as we often do!

Many of you may remember the days of WebCT and its student preview mode which allowed a section instructor to view their course as a student would see it.  They effectively became logged in as a dummy student and could participate in the available activities.  However, WebCT was clever enough to know that this was not a real user and ignored their data when it came to things like summarising the results of quizzes.  The April 2014 release of Blackboard Learn 9.1 has now introduced a similar feature; so what are the implications for LTI connections?

When an LTI launch is made in Student Preview mode, the user has a different user_id from that of the instructor.  However, there is no way for a Tool Provider to know that this is not a real student (unless the last name has been passed and it is inspected for a suffix of "_PreviewUser").  Hence the data for this user may end up being combined with the data for all the real students.  Moreover, if the recommended setting in Learn 9 is used, each time the Student Preview mode is selected a new user ID value is generated for the dummy account.  This means that the number of dummy students could end up exceeding the number of real students in a tool provider.  This may not only have implications for any summary data generated by the Tool Provider, but also cost implications if charges are made based on student numbers.

As a result of this conversation, we are exploring with the rest of the IMS LTI community ways in which a tool provider may be made aware of the fact that a user account is just for a dummy user (which could equally be when an administrator enters a course without being enrollled in it).  A mechanism for disabling access to a tool provider by a dummy user account is also being proposed for the open source LTI Building Block as a precaution for cases where there could be cost implications.  We will keep you posted on developments.




A day in the 'Life of LTI'

Sun, 12 Jan 2014 17:38:41 GMT2014-01-12T17:38:41Z

As part of the 14th Durham Blackboard Users Conference Stephen and Simon run a pre-conference workshop on LTI. The title of the workshop (and this blog) picked up in the theme of the conference 'The Life of i'. The workshop itself was different from the previous workshop we ran at Durham in that we were not targetting developers only but rather developers, VLE administrators and teachers. The topics covered were chosen by the participants and were:

  • What LTI is and what benefits it can provide, whether you are a developer, a Learn 9 administrator, or a teacher
  • How to use the core functionality available within Learn 9 to configure and enable LTI tools
  • How to find tools which support LTI
  • The latest releases of LTI (2.0 and 1.2)
  • How you can extend LTI by adding your own services
  • The current roadmap for LTI

The slide stack we used is available here (in PDF format). The biggest challange with a workshop aimed at a wide range of users is keeping the interest of the audience throughtout the session. Hopefully we achieved this! 

This workshop will form the basis for the 'IMS LTI - Extending moodle for teaching' event at this year's Moodlemoot in Edinburgh. The biggest difference being that the core functionalilty will be from Moodle not Learn for obvious reasons!

As ever, thanks to the team at Durham for all their help in organising the workshop and the muffins remain as tasty as ever --- we'll definitely be back next year for some more muffins!




WebPA stats: July 2013

Wed, 31 Jul 2013 23:09:12 GMT2013-07-31T23:09:12Z

Last month (July) was also very quiet and marked the end of the WebPA service as part of the ceLTI:sharing project. Click on an image to view a larger version.

(image) (image)

Thanks to everyone who has helped make up one, or more, of the bars over the past year.  We hope that you have found LTI to be a great solution to your needs in this case, and see how it can help you with supporting other applications.




DevCon 2013

Tue, 09 Jul 2013 01:11:59 GMT2013-07-09T01:11:59Z

I gave a presentation today entitled Developing applications for Learn 9 using LTI at the Blackboard Developers' Conference in Las Vegas.

This session outlined the ways in which LTI tools can be made easily available within Learn 9 and referenced some ways which make it easy to experiment and develop such tools.

You can view a PDF copy of the slides from here.




WebPA stats: June 2013

Sun, 30 Jun 2013 20:08:31 GMT2013-06-30T20:08:31Z

Last month (June) was, not surprisingly, very quiet; click on an image to view a larger version.

(image) (image)




WebPA LTI examples at Newcastle University

Wed, 26 Jun 2013 14:48:06 GMT2013-06-26T14:48:06Z

We have been using a standalone version of WebPA at Newcastle University in a pilot form for the last couple of years.  Usage has been small scale, involving staff from 4 academic schools.

In semester 2 this year 5 academics have used WebPA via the LTI integration to deliver peer assessments via Blackboard. 4/5 were previous WebPA users, one member of staff used it for the frist time.

We've seen a number of advantages

  • there has been no need to generate additional usernames/passwords - students have signed into Blackboard and accessed the WebPA links
  • academic staff and support staff can easily access the same WebPA assessment via Blackboard - to monitor progress and share setup/admin
  • logs have been helpful to show student activity when non-submission penalties have been challenged  
  • transferring student grades to Blackboard provides an easy way to view final scores by campusid and student number

WebPA has been used to grade:

  • Biomedical Sciences - Stage 2 - a group poster presentation
  • A Stage 2 Chemistry Group Assignment
  • Stage 3 Supply Chain management (Agri-Business)

Some staff have chosen to use 100% webpa weightings while others have used smaller 10-25% weightings.  The majority have undertaken some moderation (especially if one group member received a very low score from their peers).

 




An example of WebPA at QMU in Speech and Language Therapy

Wed, 26 Jun 2013 08:56:52 GMT2013-06-26T08:56:52Z

The following provides an example of how WebPA has been used at QMU. This example was provided by Dr Joan Ma.

 

Name of module Preparation for Practice 3C

 

Name of programme: BSc Speech and Language Therapy

 

Short description of how WebPA has been used

 

WebPA was used for peer evaluation in a group presentation assessment

 

Advantages of using WebPA

  • Encourages students to reflect on the work process
  • Marks reflect student contribution in a group situation
  • Encourages better team work as it is reflected in the final mark

 

Advice to colleagues using WebPA

  • Clear explanation to students before giving out the assignment
  • Look into the ratings carefully afterwards especially in cases of discrepancy



An example of WebPA at QMU in physiotherapy

Wed, 26 Jun 2013 08:48:10 GMT2013-06-26T08:48:10Z

The following provides an example of how WebPA has been trialled at QMU. This case study was provided by Dr Stella Howden.   Name of module: Contextualising physiotherapy: self, health and society   Name of programme: BSc Hons Physiotherapy   Short description of how WebPA has been used:   In this level 1 module students are assessed (summatively) through a group presentation. The module aims to promote exploration of self and how an individual’s behaviour can affect others. Other outcomes relate to learning to work collaboratively and participate in giving and receiving constructive feedback. To support the students to learn from one another and consider their self-appraisal of team working, WebPA is used at two points in the module.   The first stage is formative, where students are given a presentation task that closely aligns with the summative task. When students have completed that presentation they are given tutor feedback and a grade. They then use WebPA to rate themselves and others in their team. Five criteria are used (time management, communication, organisation, enthusiasm, problem solving). For each criterion students rate one another on a five point scale and are asked to provide some constructive written feedback that highlights the individual’s strengths and any areas for development.   The WebPA adjusted mark and all free text comments are emailed to students. Students are encouraged to reflect on this feedback as they move through the second stage of group working, culminating in a second round of presentations (summative), with WebPA feedback.   The WebPA ‘experience’ was evaluated after the formative stage. Student feedback was overwhelmingly positive. They valued the student comments the most highly, seeing this as an opportunity to see how others perceived their group working skills and behaviours.   Advantages of using WebPA Easy to use – students did not have any difficulties with understanding the system and rating one another. Automated calculation of the peer adjusted mark. Exporting written comments to Excel – ready for collation. Students can vote in their own time, in privacy. Advice to colleagues using WebPA I found that some students were better than others at generating supportive written feedback. After the formative stage, I ran a workshop on writing constructive feedback and the feedback quality improved on the second round. I would recommend doing this type of work-up before use of free text comments/feedback. From the formative feedback it seemed to be the most important element for the students. [...]



Rating with Outcomes!

Tue, 18 Jun 2013 16:27:43 GMT2013-06-18T16:27:43Z

Earlier this year we developed a simple Rating application to use in the developer workshops as a sample LTI tool provider (see earlier posting ).  It is available in both PHP and Java versions.

This application has now been enhanced to add code to illustrate the use of the Outcomes service.  The Outcomes service was added in LTI 1.1 and allows a single grade book column to be associated with each link to an LTI tool.  The outcome is a simple numeric value between 0 and 1, although many systems may display this as a percentage value (0% - 100%).  A number of options were considered for what this outcome should be for the Ratings application:

  1. average rating value;
  2. average rating relative to the class average;
  3. number of ratings made;
  4. proportion of items rated.

An average rating has little meaning unless the items being rated are similar in nature.  A value relative to a class average is likely to lead to values outside the permitted range of 0-1.  The number of ratings is feasible but is not comparable across different links to the Ratings application unless they all have the same total number of items to be rated.  Thus, it was decided to implement the Outcomes service to update the grade book with the proportion of the total items available for rating which a student has submitted a rating for.  For example, if a student has rating 3 of 4 items, their grade book entry would have a value of 0.75 (or 75%).  The value in the grade book column would change with the following events:

  1. a new item is added (or made visible) by an instructor;
  2. a visible item is deleted (or made not visible) by an instructor;
  3. an item is rated by a student.

In the cases of the first two events, the entries for all students in the grade book would need updating; in the third case only the entry for the student adding the rating is affected.  Using the LTI Tool Provider class libraries (available for PHP and Java) meant that this change required adding a couple of queries to obtain numbers of items and a method to call the Outcome service for the affected users; the latter method is 25 lines of code.

The new release of the Rating application has passed the IMS certification tests for LTI 1.1 and can be tested using the LTI Launcher for Windows application; see the Rating documentation site for further information.




Overall WebPA usage stats

Fri, 14 Jun 2013 16:37:45 GMT2013-06-14T16:37:45Z

Here is a more complete summary of how WebPA has been used over the duration of this project.  With 1,880 active users and over 3,000 assessment submissions, I would conclude that the service has worked well for those involved.  (Note that these statistics are taken only from the production instance of WebPA and not the development/test server.)

WebPA Usage to 14 June 2013
InstitutionUsersCoursesWebPA LinksAssessmentsGroups per AssessmentSubmissions
StaffStudentsMinumumAverageMaximum
Aberdeen189138841111.6801,744
Edinburgh1211467736.012139
Keele10188333713.722291
Newcastle1143791214110.946653
Queen Margaret10167771114.616203
Total611,819333776   3,030

Definitions:

  1. User = a person who has launched WebPA at least once
  2. Course = a course in the VLE which has at least one WebPA link
  3. WebPA link = an LTI link to WebPA which has at least one associated assessment (in WebPA)
  4. Assessment = a WebPA assessment which has at least one submission
  5. Group = a WebPA group which has at least one member



WebPA and Moodle 2

Sat, 01 Jun 2013 10:59:42 GMT2013-06-01T10:59:42Z

It has always been a source of disappointment that the only institution using Moodle as its primary, centrally-supported VLE which expressed interested in participating in the ceLTIc:sharing project was unable to have the required module added to its server (for quite specific and understandable reasons).  The project has, therefore, focussed on how LTI might be used to integrate WebPA with Learn 9 in the first instance.  However, to demonstrate that the techniques being used are truly interoperable, we have also made the same functionality available in an adapted version of the open source LTI module for Moodle 2.  This includes support for the groups extension to the unofficial memberships service and an option to use login IDs as the LTI user_id parameter. The module has yet to be publicly released but is available on request (send an email to stephen @ spvsoftwareproducts.com) in the meantime; follow this link for documentation on using the module.




WebPA meeting, Loughborough (7/11/12)

Sat, 01 Jun 2013 09:12:47 GMT2013-06-01T09:12:47Z

This is a belated blog post about a ceLTIc/WebPA meeting & ‘train the traners’ session in Loughborough back in November 2012. The format of the day was a Specialist Interest Group (SIG) Meeting in the morning (which we were lucky enough to be invited to), followed by a presentation by Dr. Keith Pond (senior lecturer in the school of business & economics at Loughborough) discussing some of the research he has done relating to WebPA and teaching. In the afternoon we were given a practical training session on WebPA – a 'train the trainers' session designed to allow those of us less experienced with WebPA to find out more about the tool itself. The SIG was a useful meeting of minds and the discussion spawned some ideas that we are seeing the fruits of now. One of these was the WebPA academic workshop which the University of Edinburgh are hosting on 3rd June 2013. I’m sure there will be a blog post about that from one of our other partners in the next few weeks. We also discussed technical developments such as the group sync functionality, which Stephen Vickers has now built into the Blackboard Learn building block – a really useful productivity enhancement. Also on the agenda was help documentation, and in particular the sterling work that Ian McGilloway has been doing for Imperial College London. Luckily for us Ian, & Imperial, have been kind enough to open up their WebPA Help Guides to world – a really useful resource for the WebPA community. Lastly, but most importantly, the project’s future hosting options were discussed and we should hopefully be seeing some conclusion to those discussions in the coming weeks. Personally I found Keith Pond’s session extremely useful. Keith’s ideas around peer moderation seem to have been some of the drivers for the original WebPA project and his ongoing research into the various issues around using WebPA in an HE context is worth checking out (see some selected references at his Loughborough staff pages). Keith made some interesting points around good use of WebPA: It can be a good idea to get students to come up with their own moderation criteria. It helps reinforce things like: what do we mean by good cooperation; and how is that different from contributions? It helps in the process of getting students to reflect on group work and dynamics.A new(ish) feature in WebPA is the ability to add text comments. This came out of qualitative research (students wanted qualitative feedback, not just a number). The feedback flow to the student is in the form of generic comments from the group.But have students read feedback? And have they gone on to read the further information provided? Often not! Doing so might help towards completing the feedback loop.Interestingly students tend to write more about themselves than others. This aligns somewhat with the fact that students tend to score themselves higher in the numerical moderation - but the algorithm normalises for this effect.Surprisingly he found little, or no, cultural differences in student behaviour in a peer moderation activity (comparing WebPA usage with students at Loughborough & those at a Singapore campus). Finally we were given a demo/training session on the tool itself from one of the lead developers. For me, as someone who hadn’t used WebPA before, this was an extremely useful. Some of the more generally useful tidbits I thought were the ‘common gotchas’: late arrivals are problematic - group structure can only be changed up until first student submits (the only ways to accommodate late arrivals are to delay group making for as long as possible, or to create another assessment).;student switching groups – same solut[...]



WebPA stats: May 2013

Fri, 31 May 2013 23:28:02 GMT2013-05-31T23:28:02Z

The stats for last month (May) are shown in the graphs below; click on an image to view a larger version.

(image) (image)




WebPA Event

Wed, 15 May 2013 08:37:55 GMT2013-05-15T08:37:55Z

Along with The University of Edinburgh, the project is organising a WebPA event in Edinburgh on Monday, 3rd June (2013). Start 11.00; End 16.00.

Whilst the ceLTIc project’s primary focus is the use of LTI to share WebPA amongst a number of partner institutions this event is focused on the pedagogy of WebPA. The event is free to attend and lunch will provided.

The final agenda is:

11.00 – 11.10 Welcome: Simon Booth (ceLTIc)/Jo Kingsley (Edinburgh)
11.10 - 11.45 Practitioner Reflections on WebPA: Mike Pretious (Queen Margaret)
11.45 – 12.20 Experiences of using WebPA – practicalities and issues: Neil Gordon (Hull)
12.20 – 12.55 Experiences of peer assessment using WebPA: Sara Preston/Laraine D’Antin (Aberdeen)

12.55 – 14.00 Lunch

14.00 – 14.35 The benefits of LTI-enabled WebPA: Stephen Vickers (Consultant)
14.35 – 15.10 WebPA and the School of Veterinary Studies: Sharon Boyd (Edinburgh)
15.10 – 15.45 Delivering Student feedback via WebPA: Shireen Lock/Ian McGilloway (Imperial)

Thanks to all the speakers.


The presentations from the event can be viewed here:

http://listenagain.stir.ac.uk/media/keep/webpa/listenagain.php

 




LTI saves lives

Mon, 06 May 2013 21:49:20 GMT2013-05-06T21:49:20Z

Or "A day in the life of an LTI developer" ...

Many LTI-enabled tools have been created by taking an existing application and adding LTI support to it.  For example, this is what we have done as part of this project with WebPA and WordPress.  Each integration poses it own unique challenges, particularly in respect to how contexts, resource links, users and roles in the VLE (tool consumer) are mapped to objects in the external application (tool provider).

I recently performed a simple LTI bypass to the e-Reflect tool and have written it up as a case study entitled "LTI saves lives" to illustrate the nature of the process and how this relatively simple operation can bring new life to the patient.  Before the operation, e-Reflect was only capable of supporting a single institution per installation and users had to be pre-provisioned; afterwards it was transformed into a multi-institution application with the definition of users and roles being delegated to each VLE.  Not bad for a day's work!




WebPA stats: April 2013

Wed, 01 May 2013 01:37:54 GMT2013-05-01T01:37:54Z

The stats for last month (April) are shown in the graphs below; click on an image to view a larger version.

(image) (image)




Evaluating apps: the quick way with LTI

Tue, 30 Apr 2013 22:24:36 GMT2013-04-30T22:24:36Z

One of the issues being explored by the ceLTIc:sharing project is how LTI can be used to make it easy for teachers to try out a new application.  I am pleased to bring you news that we now think we have made this as easy as it can possibly be.  Here are the solutions which we have experimented with:

  1. Dynamically generate a consumer key and secret allowing access to a site for a specified trial period; this was first announced with the sample Rating application.  These credentials can be entered into your VLE and allow you to launch into the application and try it out.  Launches will automatically fail at the end of the trial period.
  2. Using the LTI Launcher for Windows application enter a provisioning URL for an LTI tool; this will dynamically request a temporary set of credentials and automatically add the tool to the application.  This is similar to the previous solution except that it does not require the use of a VLE and it also automates the request for the credentials to avoid them having to be transcribed manually.
  3. If you have the LTI Launcher for Windows application installed on your machine, then just click on an Open in LTI Launcher for Windows button on a tool's website.  This further simplifies the previous solution by removing the need to enter a provisioning URL for a tool.
  4. Select one of the available tools from the Provision menu available from within the LTI Launcher for Windows application; this will automatically provision the tool without the need to find its website.

At present the following applications can be evaluated using the above solutions:

If you have written an LTI Tool Provider application and would like it added to this list, then please contact us.

A screencast illustrating these solutions in practice can be found on the LTI Launcher for Windows website.  If you can think of a way in which experimenting with an LTI tool can be made even simpler we would love to hear from you - the comments option below awaits your feedback!




WebPA stats: March 2013

Mon, 01 Apr 2013 10:54:03 GMT2013-04-01T10:54:03Z

The stats this month (March) include a few corrections to the values for February. Click on an image to view a larger version.

(image) (image)




A student progress dashboard for WebPA?

Fri, 29 Mar 2013 09:28:13 GMT2013-03-29T09:28:13Z

(image) One of the conversations currently going on within the IMS community is about Learning Analytics and LTI.  The interest here has a number of strands but starts with the relatively simple notion of whether it might be possible to have some form of dashboard which allows students to get visual feedback on their activities in tools which are external to the VLE (e.g. from IMS LTI tool providers).  Technically this does not seem to be very difficult; it is quite similar to the dashboard option added to the LTI Building Block for Learn 9 (see the posting entitled "WebPA dashboard").  Whilst the posting did not generate any comments from anyone interested in applying the concept to WebPA, I am hoping that a "Student Progress dashboard" may find greater resonance with readers.  If so, please add a comment with your thoughts on what would be useful to include in a visualisation for WebPA.  The intention is that this should be an image and relatively easy to interpret without too much instruction.  A mockup of what such a dashboard might look like is shown in the image (click on it to view a larger version).

Thanks in advance.




WordPress and user scope

Sat, 23 Mar 2013 12:32:28 GMT2013-03-23T12:32:28Z

One of the activities of the ceLTIc project has always been building example LTI integrations for different tool providers to help illustrate the issues which arise when implementing the LTI specification.  As outlined in an earlier blog post (entitled Selecting an appropriate user ID) a developer must make an appropriate choice for how to identify individual users.  This was a relatively straightforward choice for the LTI connector for WebPA because the application was also updated at the same time to support LTI.  In this case, users are identified from a combination of the consumer key and the user_id parameter.  This means that each user account may have access to more than one WebPA module, but the application code prevents users from changing to a different module except by launching from the appropriate resource link in the VLE. When it came to an integration for WordPress (to be released soon, watch this space!) life was a little more tricky.  This is an application for which the codebase is developed independently of the project, so we needed to find a solution which would fit within the existing way in which user accounts and blogs operate.  So which of the 4 choices discussed in the earlier blog post would make sense for WordPress (considered in reverse order): Resource link - this is what is most likely to be expected from an LTI connection to a tool provider; each link in the VLE would link to a separate blog and users would only be able to access this blog from the link.  It does mean, however, that WordPress may have multiple user accounts created for each person (one for each blog).Context - this choice could be seen as an acceptable compromise.  It limits the creation of user accounts to one per context per person.  Users may be able to access other blogs from one launch, but at least these will be for a course (context) in which they are enrolled.  It may even be that an institution may only ever want to have a single blog per course anyway.  It may also be possible to make it hard to exploit the access to other blogs by hiding the relevant links in the interface.Consumer - perhaps allowing users to access other blogs they have had access to in the past (but may no longer have) is not considered a significant issue - this could depend upon what WordPress is being used for.  In which case, this choice would allow a single user account in WordPress to be created for each user account in the VLE and simplify the management of the system.Global - when the WordPress instance is only to be accessed using a single consumer key (i.e. from a single VLE) then this choice is equivalent to the Consumer choice (see above).  It has the added benefit of potentially being able to create user accounts with the same ID as users enter for other systems.  This could be useful if direct login access is also to be allowed to WordPress.  Thus, if a user has a login name of "rob" and this is also the value of the user_id parameter provided in an LTI launch request, then a WordPress user account with an ID of "rob" can be created (or used, if it already exists).  If the LTI user_id parameter is something more opaque (as per the LTI specification) then it may still be possible to find a way of associating this value with the user's login name.  However, this choice would not normally be appropriate if launch requests are to be accepted from more than one VLE; it runs the risk of different users in each VLE having the same [...]



Selecting an appropriate user ID

Sat, 23 Mar 2013 11:19:01 GMT2013-03-23T11:19:01Z

Most tool providers will want to record activity against individual users and hence will need some way to uniquely identify each user when a launch request is received.  So what parameters should be used as a unique user identity value?  Well, the answer to this question is that it depends! The PHP and Java LTI Tool Provider class libraries implement 4 choices (or scopes) for creating unique IDs: Global: just use the user_id parameter;Consumer: combine the consumer key with the user_id parameter;Context: combine the consumer key with the context_id and user_id parameters;Resource link: combine the consumer key with the resource_link_id and user_id parameters. Each of the choices is useful in different situations. Since the user_id parameter is only guaranteed to be unique for the same tool consumer, the Global choice only makes sense if the tool provider is dedicated to a single tool consumer (e.g. a local installation for a University) or if it is known that the sets of user_id values from the different tool consumers do not intersect.  This is not a choice I would recommend; even if only a single tool consumer is expected to be supported, you never know if this might change in the future - for example, an alternative VLE is made available within the University. The Consumer choice guarantees to provide you with a unique ID for each user within the tool provider system.  However, its use does mean that the same user ID is used for launches from all the resource links added in the tool consumers.  Thus, there is a possibility that users following one resource link, may be able to access content linked from another resource link, thereby bypassing the authorisation process provided by the tool consumer.  For example, a student may have been unenrolled from a course in the VLE and, therefore, no longer have access to the resource links in that course.  If launching from another course allows them to access the end points of these resource links, then the tool provider has effectively provided a back door for users to access resources which they should have been disconnected from.  Some solutions to this problem may be available to you.  For example, if you control the code for the tool provider application, you can prevent users from switching to a different resource after a launch (this is how the LTI connector for WebPA works).  Alternatively, if the tool consumer supports the unofficial memberships service, the list of users for each resource link can be updated on a regular basis (e.g. nightly) so any impact will only be short-lived. The Context choice is similar to the Consumer choice (see above) except that it implements an ID value which is only unique within a context (e.g. a course).  If all resource links within a context are always available to all users enrolled in that context, then there is no issue about providing a back door method for accessing resources.  However, if access to some resource links might be restricted to a subset of the enrolled users (e.g. via a conditional release mechanism) then this choice could provide a way for users to bypass the authorisation implemented by the tool consumer (for which similar solutions as discussed above could be implemented). An implementation of LTI should recognise that when a launch request is received, all this tells the tool provider is that the user (as identified by the user_id parameter) from the specified tool consumer is enti[...]



Moodlemoot Presentation— Dublin

Sun, 10 Mar 2013 13:53:28 GMT2013-03-10T13:53:28Z

Simon attended the Moodlemoot in Dublin (http://moodlemoot.ie) and give a presentation along with the ceLTIc:developers LTI workshop — thanks to Gavin Henrick (http://www.somerandomthoughts.com/blog/) for organising the Moot and adding the workshop to the hackfest. The ceLTIc LTI poster also made a couple of appearances at the conference: once in the posters and once in Gavin’s LTI presentation — thanks Gavin!

The presentation was about both parts of the ceLTIc project but concentrated on the sharing part as the use of LTI and WebPA we felt was of more interest to those attending the conference, especially as those interested in the technical aspects of LTI could attend the workshop in the hackfest.  It was a challenge to fit the talk in the seventeen minutes or so allocated but these shorter time periods do concentrate the mind. As always the presentation included a demonstration of how easy it is to set up an LTI link and this was done as a video — it also helpful knowing how long the demonstration lasts when fitting everything into seventeen minutes!

The workshop had 11 attendees which was 11 than I expected given that Martin Dougiamas and other chief Moodlers were  part of the hackfest and in the room next door! The development environment worked on all the machines present (including Linux) except the tablet — now there’s a challenge for the next workshop. Having sorted out the iPad then the iPhone/iPod would be next or should we start with Android!?

LTI was a significant theme at the conference with a number of interesting presentations on the topic and Martin Dougiamas also gave a positive endorsement in his plenary session. The project’s plans to take-over the World remain on course!