Month 10 – February 2017

Month 10 – February 2017

I had this post sit for a while — more than one month. It was written on my last train ride from Nijmegen back to Austria and was intended to wrap up things from a larger perspective. I wasn’t sure, wether it turned out to be too emotional and thus decided to let it rest and take a look at it from some distance. After about a month now, I’m still fully in line with the text … so here’s the unaltered version.


So that’s basically it. After ten months, my fellowship has finally come to an end. Looking back, it has been quite a ride. In contrast to the former monthly entries, I here would like to take the grand perspective and elaborate on the most important outcomes of my stay in Nijmegen (personally, that is — the project itself has reached all its aims and even went beyond my expectations).

Over the last few years, I’ve been very reluctant to even think about the necessity to spend some time abroad. Sure, I knew it would be necessary in order to achieve my formal qualification aims (eventually leading to a permanent position) — still, I’d always been pushing even it away. Even when I finally wrote the proposals to apply for a fellowship, that whole thing remained sort of abstract. Only when the Schrödinger project got granted and I finally actually had to sign the form applying for unpaid leave for those ten months (basically closing the way back), I realized that my stay in the Netherlands with all its consequences was about to happen.

I remember taking the last look back on Science Park 3 at JKU, where my office was (and is) located. One week later, I was sitting on the train to Nijmegen. Fast forward ten months, I’m now sitting on the night train back home. These ten months surely have been the most influential ones in my way of becoming what I consider a design science researcher. Let me elaborate on this.

First, with the complete freedom in how to organize my work and what to focus on, I finally managed to develop a deeper understanding of how design-oriented research processes might work (or at least work for me). Having the time to deeply immerse oneself in a topic, spend vast amounts of time reading literature, and carefully crafting one’s arguments is a privilege that I haven’t had for years. Back then when I had the time, I retrospectively did not have the mindset to (a) value that freedom, and (b) set up a research project in a way that is methodologically sound and does not require argumentative ex-post stunts to establish a valid foundation for the research setup. I at least believe that my abilities have evolved over the years and that I have been able to apply and hone these skills in the last year.

Second, I rediscovered me being a computer scientist and the fun in actually building interactive systems. My project proposal had foreseen the development of a software component that would aid the process of reflecting on and modeling work processes. In quite a surge of inspiration (that, by accident, matched the project plan), I sat down in November and coded the first version of this system in a nearly two-week-long marathon. Being away from actual coding for nearly 7 years, I had completely forgotten the joy and sense of accomplishment one can get from getting some idea to actually work. Ever since November, there has not been a week where I did not devote at least one day to coding.

Third, in interacting with the teachers at Radboud and HAN, I had the opportunity to reflect on my own teaching and develop new perspectives and approaches — both, methodologically and content-wise. I have always had the feeling that interactive teaching formats are possible even for large groups of students (say, 70+), but actually never had the motivation and methodological background to actually try it out. That is, until I have been confronted with flipped classroom concept in Arnhem. This immediately resonated with me, and so I decided to implement it in the lecture I held back in Linz. Due to me limited availability, I was desperately looking for an appropriate format to hold a lecture in four full-day blocks for more than 90 students anyway. The flipped classroom seemed like a match — and it definitely was. It has be successful to an amount that I would never had expected (in terms of both, student feedback and results), which has lead to me to adopt it also for our introduction level lecture (with 100+ students).

Fourth, I’ve had the privilege to work (and chat) with great people from very diverse backgrounds, complementing my view on my work and the world in general in ways that I never would have anticipated. Without elaborating on this in detail, simply being confronted with fresh perspectives inevitably broadens one’s own mind and provides the opportunity to see the own routines and established mental models in a different light. In combination with the chance for intellectually stimulating discussions on topics that are my core interest, working with the people in Nijmegen and Arnhem probably has been the best possible way my Schrödinger fellowship could have gone.

Finally, I’ve developed some sort of self-confidence I did not know I could have before. Seeing that I’m finally able to put my thoughts on paper by demand, producing articles from available results without it being a month-long procedure, and — most of all — seeing that I’m still capable of developing non-trivial software (a chapter that I had already regretfully closed for myself) has shown me that — after all — there might be a reason why I’ve persisted so long. The combination with having experienced that I’m able to put to practice the skills and knowledge that I have acquired over the last years leaves me confident that I’m up to the challenges that are (hopefully) to come as a faculty member.

The FWF states in its funding guidelines that the aim of the Schrödinger Fellowship is „enable young scientists […] to work abroad at leading research institutions and on leading research programmes with the purpose of gaining research experience abroad during the postdoc phase; [and] as a result of such projects to open up new areas of knowledge, new scientific approaches, methods and techniques […]“. I can safely state that this has completely worked for me.

Advertisements

Best Paper Award at S-BPM ONE 2017

My paper „Business Process Elaboration through Virtual Enactment“ has received the best paper award at S-BPM ONE 2017 in Darmstadt. The paper summarizes the technical results of work package 3 and gives a first account on the empirical evaluation of the instrument. It is intended to provide the foundation for a more extensive journal article to be submitted later this year.

Thanks to FWF’s open access funding, you can download the paper from the ACM Digital Library for free under an CC-BY license.

Month 9 – January 2017

Month 9 – January 2017

The last but one month of my fellowship basically went as planned in the original project outline. Things start to wrap up now and I have started to concretely work on the follow-up activities to continue my research after the project. As anticipated in the project plan, I still devoted a large share of my time to further developing the Virtual Enactment tool. As always, some features that even won’t be that visible on the frontend, but are crucial for the core functionality, have required the most effort. Now, however, there seems to be a good base for further development.

Aside development, I have started to work with students back in Linz to plan the further evaluation of the tool in a more comprehensive study than the exploratory one I did back in November. Following a longitudinal research design, the study will presumably take place over the first half of the summer term. The journal article summarizing my work during the fellowship is planned to comprise these results and thus will be submitted sometime around June. To cover the more recent results, I plan to submit another paper to the T4SIS4T-workshop held at CAISE this year.

Regarding teaching, I have further elaborated my flipped classroom concepts, which I acquired during my fellowship. Following up on the great feedback for last term’s lecture on distributed information systems, I will completely revise the teaching concept and materials for my summer term lecture on process- and communication modeling, also preparing it for flipped classroom implementation. A share of my time in February will be dedicated to these preparations.

As for February, I expect some further development of the Virtual Enactment tool, some paper writing and quite some effort to close down all my affairs in Nijmegen. My fellowship will presumably end just as it started, sitting on the train and thinking about the months to come.

Month 8 – December 2016

Month 8 – December 2016

December is coming to an end a bit early because of the upcoming Christmas holidays. Month 8 of my fellowship largely went as planned was much less travel-intense than the previous months.

The main activities in December revolved around progressing workpackage 3, in particular revising the software tool based on the results of the evaluation study carried out in the end of November. As the tool grows in size and complexity, I have decided to use GitHub as both, a repository to manage the source code and a project management tool for organizing further development. The repository is available online at https://github.com/win-ce/VirtualEnactment and contains the inital stable release version and the current lines of research and development. In the course of moving over to GitHub, I also decided to open source the software under the GPL 3.0 license. In terms of development, I have revised the most critical features in terms of functionality and user interface — the elaboration interface is completely reimplemented in the current development branch, featuring a much more flexible approach to specifying the necessary information. Furthermore, I have focussed on on-boarding for novice users and have started to implement a scaffolding agent that actively introduces to the usage of the tool by incrementally activating and explaining the features.

Aside the implementation activities, I also have found time to write (or contribute to) three papers wrapping up the most recent research and development activities. The VirtualEnactment tool and the first experiences with its practical deployment are summarized in an article submitted to S-BPM ONE 2017. A working version is available on this website. Another paper co-authored with two of my students describes the use of the Adaptive Testing Platform already dealt with here for assessing BPMN modeling skills. It is also submitted to S-BPM ONE 2017 and available in a working version here. Finally, I have revised my contribution to Momentum Kongress in October for submission to the Momentum Quarterly Journal. It deals with how reformatory pedagogics, in particular the approaches of the Freinets, can be deployed in higher education settings.

In terms of project administration, I have published all project results (read: articles and software) in the Zenodo Open Science repository. Zenodo is supported by CERN and funded by the European Union and provides long-term archival services for scientific results. One major advantage is that all artifacts published on Zenodo receive a DOI and thus become citable – I have already used this feature in my recent publications. Furthermore, Zenodo also integrates with GitHub, making it very easy to publish citable Software Archives. The archive for the project is available as a Zenodo Community.

In all, I’ve been wrapping up a very successful year. January will again see some further work on the software and a final evaluation of both, the tool and the methodology. Aside this, we have started to prepare longer-term cooperation activities with HAN university of applied sciences in order to continue collaboration also after my fellowship will have ended.

ScaWoMo results now available via the Zenodo Open Science Repository

Starting from today, all results of the project are being made available on Zenodo. Zenodo is an Open Science Repository hosted at CERN and co-funded by the European Union.

The project results are published via the ScaWoMo-collection on Zenodo. To start with, it contains the two papers already published at ICKM 2016 and PoEM 2016 that report on the results of WP2, and the working paper currently submitted to a conference reporting on the first results of WP3. It furthermore contains a citeable snapshot of the first stable release of the software tool that has been developed in WP3. The source code of the tool is also being made publicly available via GitHub.

Month 7 – November 2016

Month 7 – November 2016

The first half of November was still largely shaped by traveling, having been to Skövde, Sweden, to participate in PoEM and visiting Bamberg, Germany, for an invited talk given at the colloquium of faculty of information systems and applied computer science. After returning from my travels, I have been largely immersed into coding – for the first time in around seven years, I have actually done serious software development in implementing an initial version of the tool for scaffolded elaboration through virtual process enactment, which is the main result of work package 3. The prototype is publicly accessible at http://adaptivetesting.ce.jku.at/VirtualEnactment. Initial evaluation with end users has been promising and is currently being summarized for a publication to be submitted to S-BPM ONE 2017 (satisfying the requirements of deliverable D3.2).

In addition, I have finished the preparations of our study on why modeling approaches — although theoretically sound and successfully validated — still seem to have limited impact in daily practice in enterprises. The questionnaire is already online and will be distributed in the next few weeks. First pretests had shown some flaws in understandability of the questions, which were fixed in a new revision.

As a side activity, I have continued working on the adaptive testing platform for skill assessment in the field of conceptual modeling. This has brought a new revision of the testing platform, which is available at http://adaptivetesting.ce.jku.at/AdaptiveTesting2, currently still featuring the original accounting test. Aside the technical developments, we have summarized our work on item design for measurement of modeling skills and also anticipate to submit it as a paper to S-BPM ONE 2017. There, it should act as a call for participation to contribute to item design and validation, as the critical point in computerized adaptive testing is to construct a sufficiently large pool of calibrated items (i.e., items of known difficulty).

In the meantime, the flipped classroom course is coming to an end mid-December. In cooperation with a student doing research for his bachelor thesis, I have prepared an evaluation study of the concept and its implementation. Study design in based on the current state-of-the-art in flipped classroom evaluation and thus should lead to results that are worthwhile publishing, too.

December will see some tidying up of the virtual enactment tool based on the initial evaluation results. Most of the remaining time will be necessary to finish writing the papers mentioned above and the submission of my contribution on reformatory-pedagogics-based education in scientific work practices to the Momentum Quarterly journal.

Interactive Scaffolding for Work Process Elaboration

One of the core results of ScaWoMo is available online now. As part of my effort to provide interactive scaffolding during work process elicitation and alignment, I have developed a small tool that allows to interactively play through collaborative work processes and elaborate them on the fly. This process of playing through and elaboration is supported by interactive scaffolding agents, which observe the elaboration process and provide hints on what to consider when continuing.

The tool prototype is available at http://adaptivetesting.ce.jku.at/VirtualEnactment (never mind the URL – that simply was the only Tomcat-Server I had available at the moment with a publicly reachable IP address). The current version of the source code is available at https://github.com/oppl/VirtualEnactment.

I’m quite happy with the prototype and look forward to the first evaluation results. Any feedback is welcome – simply contact me at stefan.oppl@jku.at.