For the past several months, I’ve been working on a project in the primary and secondary (K-12) education space – my first in that domain. In a lot of ways, I’ve found the experience to be quite unique when compared to some of the other industries that I’ve worked in, and I wanted to write this as a follow up to a Solution Brief that BlueGranite did on the subject, and to share some of my thoughts and experiences stemming from my time working in the education field – as it’s been a journey that I’ve found to be singularly interesting, often challenging, and ultimately, very rewarding.
To help set the stage a bit with some technical specificity – this project is about building a data warehouse in SQL Server from the ground up, incorporating three primary domains of data from within a fairly large school district with approximately 30,000 students: student data, financial data, and HR data. With the data structured and consolidated in one place, Microsoft’s Power BI is then used to provide insights to staff throughout the district using sophisticated analytical logic, powerful aggregation capabilities, and centralized web-service hosting and deployment for ease of collaboration and consumption.
Holistic Process, Structured Data Are Key
Something I’ve written about in the past (if I can be permitted to go ahead and climb atop my soap box right out of the gate) is that properly structuring data is absolutely paramount for an analytics project – and education is no exception; if anything, it only underscores the importance. This isn’t to say that the data we worked with started off unstructured (though it can in some cases, as with sentiment data drawn from survey feedback), or that insights can’t be drawn from raw data, just that the state and structure of our source data is seldom ideal to an analytics practice – both in terms of what is most performant in the analytics tools being used, but also with regard to what is most conducive to imparting meaning and insight to our data consumers.
So how is such an important thing attained? Aside from the technical particulars of what constitutes a “proper” data structure for analytics – which represents a body of nuanced information far greater than what can be placed into a blog format – I’d like to focus more on the holistic process that needs to be established as a prerequisite to building an analytics solution. In summary, that process is about garnering understanding, which itself is contingent upon something far more interpersonal and organic than technical: the establishment of a dialogue between those who work within the day-to-day processes which generate our intended source data, those who are tasked with collecting and structuring that data, and those who will become consumers of the analytical insights which are the essential output of the project. This represents no small feat of coordination, and often defies some manner of technological “silver bullet” for remediation.
In education, a lot of data is generated by human beings entering data about other human beings; there is margin for error, bias, and assumption. There are contextual qualifications, such as assessment scoring, that do not fit neatly into universal schemas. So it is absolutely critical that the processes which generate source data are understood fully. This is the foundation to any analytics solution, and like anything, the soundness of the foundation represents the stability of everything built on top of it. As an analytics practice matures, the necessity becomes more apparent than ever. Advanced analytics, like those on offer in Microsoft’s Azure AI suite, absolutely require a high degree of confidence in the data inputs that they draw upon. The process by which data is structured for analytics is also one in which such confidence is built and standardized across an organization.
Breaking Down Siloes Boosts Shared Goals
If those tasked with building and maintaining an analytics practice draw the foundation of their understanding from those that oversee the generation and curation of source data, then we can say that the building plans are drawn from the perspectives of those that will be consumers of the data. This is the source of our big picture, and the sophisticated logic that crystalizes it. This is the perspective that provides the basis for breaking down the silos in which our source data is often stored, and the impetus for finding value in the exercise itself. In education, every facet of a school district’s operation bends ultimately to a single common goal: improving student outcomes. This means that a truly holistic view involves not just the data about students and education directly, but also about the district’s function across the board. This includes everything from how budget is allocated to program management efficacy – from staffing of open positions to credentialing and qualifications of those staff. In terms of structuring data for analytics, this particular process is often referred to as “conforming”, or tying disparate sources of data to a single common form for the sake of deeper, more complete insight.
As a technologist, I’ve always been drawn to the way in which new tools help facilitate old tasks. Or, more specific to Business Intelligence, how new technologies can ease the turning of the gears of discovery and implementation which drive an analytics practice along its path to greater maturity and, by extension, greater insights. However, as a consultant, I am also fascinated by the more organic nature of problem solving, and the elemental facets of such which seem to defy the passage of time – largely I think because they stem from our strengths and weaknesses as human beings. There’s a certain resonance to that when working in the education space, which is fundamentally about advancing human understanding – an endeavor which requires a certain measure of due diligence, an open dialogue among those with something at stake, and a spirit of breaking down barriers if it is to succeed.
We Build Solid Foundations
BlueGranite can help your organization implement a holistic, thriving analytics practice, too. Whether you’re considering an Azure-based modern data warehouse, or how to put AI and machine learning to work for your enterprise, we can help. Contact us today to learn how we can implement a strategic analytics framework tailored to your needs.