The motive of [[Design-based research|design-based research]], partially posits that the effort in the process of research should be in the abductive process of coming up with ways to improve learning (proto-theories) and reaching desired effects, and in testing them to adequate degrees in real-world naturalistic contexts, with fast moving iteration to gain local-theories that have translational value. Though the issue with derived local-theories is that they sit in an awkward contextual space that makes them difficult to find and use by those who could benefit from them - researchers and practioners. The notion of [[Memex|memex]] coined in 'As we may think' by Vannevar Bush in 1945, is a device that can store large amounts of information that can be accessed quickly and flexibility. Therefore, the memex would expand the available memory that the individual can use in their research and reasoning processes - motivating many technologies like the web and search engines. However, current [[Memex|memexs]] are not appropriate for the nature of knowledge derived from design-based research - the local theories. We don't really build knowledge in meaningful ways, and cannot fetch, share and distribute knowledge in appropriate ways to guide practice. That is teachers and researchers are expected to find and extract relevant insights to support their process through papers and search engines. Though teachers will not have the time to go through a vast array of papers and digest relevant information, then undertake the task of applying to the specific lessons and situations where it may be valuable. Rather, such dissemination of insight should happen in a more natural manner such as in the process of [[Learning design|learning design]] by having relevant local-theory presented at appropriate decision points. Unfortunately, the ideals of DBR still fail, if the dissemination and use is not effective. To contribute to theory, means that theory should be well enough documented, searchable, and usable in appropriate times ... which are qualities that computers could support. Though to even support the building of such a teacher memex for learning design, we first need a different process to acquire and represent these experimentation processes, such that the knowledge could be extracted into some form of [[Pedagogy modelling language|pedagogical modelling language]] that can be reasoned upon to provide relevant interventions for teachers. Before we start thinking about teachers, it could be easier to start from supporting researchers following DBR, that is the lower-hanging fruit of a researcher-memex. Where a memex has two sides, that is collecting the information and then interacting with the information. **I feel that this researcher memex should have the following properties:** - Collecting information - Collect the reasoning and rationality behind the how the proto-theory was derived. - The proto-theory could be collected in a semi-structured way through expression within a [[Conjecture mapping|conjecture map]], that maps onto a formal expression of the learning design through an [[Educational modelling language|educational modelling language]] like [[Orchestration graphs|orchestration graphs]]. - The data from the enactment should be collected. Not simply computational forms like log data, but ethnographic observations, student thoughts and so on. Additionally, this could also act as a data aggregator which simplifies the collection of evidence in a way that minimises admin-related tasks. - The researchers interpretation and analysis of the data. - Given that we collected formal structures and results, we should be able to **create theory representation in a semi-computation structure** (a [[Pedagogy modelling language|pedagogical modelling language]]), of how pedagogical knowledge (conjectures) and enactment in design (orchestration) map onto the appropriate results. - Interacting with information - Sharing of specific designs - Sharing of [[Pedagogy modelling language|PML]] - Easy replication and adaption of specific designs - Sharing of vast data that could be analysed to identify more generalisable results - ... (people can simply create different representations for interaction with this data) Ideally this should: - Shift our research effort to be in the rationality and abductive thought processes, whilst also building on prior work. - Support the execution, the testing, and general research process (**that is orchestrating research at a large scale!**). (see [[Research infrastructure for the learning sciences|research infrastructure for the learning sciences]]) - Allow for large scale research with naturalistic contexts, rather than simply online courses with super granular data. - Create direct links between pedagogy ([[Pedagogy modelling language|PML]]) and enactment in designs ([[Educational modelling language|EML]]). --- *This is the end of my comment, but following are some ramblings.* ## Ramblings ... Whilst some criticise the specificity of the results from DBR, it is in the nuances where the knowledge remains for education. Generalisable results often fail to have naturalistic validity. But the solution to this criticism is not to do some high level experimental evaluations, but to be able to support the derivation of contextual theories in a more rapid way, and support the dissemination of theories in a more accessible way. There is a relentless flow of practice occurring in every school around the world, but the pace of learning science research is so slow and often misdirected. I feel: - We spent too much time doing admin stuff. - The output format of our knowledge is rarely/never used. - We often drive what should be researched by what is easily evaluatable, reasoning is not valued enough as of present ([[Seymour Papert|Papert]] is not considered rigorous enough is current conditions :/) - We may be deriving knowledge, but what knowledge is worth deriving? - Ideas, things that work, designs, ... should be accumulable ... searchable etc. - We should have a platform for research should be open-source and highly extensible.