Part 3 of 3 – Supercharge teacher effectiveness with learning stories that support a whole-of-student view.

From creating a clear intention to declutter data in Post 1 through to building a framework around data in Post 2, Post 3 describes a platform approach to “how you do it”. Avoid BIG DATA syndrome and building lots of complex charts. Teachers simply want better visibility into data.

When K-12 teachers instinctively access teaching and learning data to maximise their effectiveness, the elephant named DATA has truly LEFT the room!

Having access to and working with data, as a habit, is a primary goal for every school and teacher. Finding the time to retrain as a data analyst or scientist is completely something else.

In 2019 it’s terrible to think that data remains the elephant in the room in most teaching and learning meetings. Everyone knows data is out there, most have seen bits of it but pretty much everyone wishes they could see it all. ‘Data science’ as a practise has claimed an undeserved ownership over data, building a mystique around data collection and presentation. In reality, data scientists only have to train a computer to do some pre set moves. I think the dynamics and variables involved in teaching cognitive humans is far more involved and complicated. This is why teachers need to have access to data stories to support them in their context.

In fact, all of the hype around data science and analytics detracts from what data really gives teachers; a Dynamic Ability To Affect. Access to data must underpin a Teaching As Usual framework, devoid of jargon, latency and complexity.

For now it’s all about keeping our eyes on the real prize. The ultimate value of data in K-12 schools comes from the learning stories teachers use to affect and personalise learning. The easier it is for teachers to find stories in data, the more effective teachers are and the stronger the use of data becomes.

Helping teachers to see into teaching and learning stories is the ‘now’ conversation schools want to have. Getting teachers connected with their hypotheses and instincts is a priority. Personally I think 99% of teachers know there is data orbiting them with differing dimensions  of volume, velocity and variety. The core problem is that teachers don’t have a simple and time efficient way of accessing data and seeing the story it tells. What if this could change?

In Post 1 of this series, I started a conversation about the need to declutter our minds about data and only work with data that brings joy to teachers. In Post 2, I introduced the idea of six data domains, into which all K-12 teaching and learning data falls. As with all things in life, the devil lives in the detail. With Multiple Data Sources and Elements within each Source, the core of all data challenges and usefulness comes down to a single word, Alignment

Think back to the Rubik’s cube metaphor. Alignment describes the process by which the cube is solved. If data alignment is not robust, the momentum around it and support for its use ends quickly.

I now want to take a look at more detailed data alignment challenges schools face as they execute on ‘Delivering the whole of student view’. The guiding presumption from this point is that Schools want to organise their data Domains and Sources to deliver a ‘whole of student view’ across all teaching and student learning data.

You must develop an efficacy balance between learning data and teacher engagement, all in the context of learner stories.

This post is not meant to get technical and make an over-talked topic even harder. We all need to focus on strategies that declutter the way we think about data and in so doing improve visibility into it. There are two parts to working with data and both parts need to work together.                              

  1. If data is not well curated, the potential stories and usefulness will reduce.
  2. If the data exists and the stories can’t be seen, you miss out on the whole point of looking into data; to inform teaching.

Alas, the yin and yang is the collection of and visibility into data – described in one word again, alignment. Following along the theme and challenge of ‘alignment’ there are two steps to take.

Step 1. Create DATA Alignment

1.1 Align all learning data to the atomic level of a Student.

10 seconds: Make sure that all of your teaching and learning systems support a single unique Student reference. If you lose control of alignment of data to students you are disconnected with learning.

Bigger Discussion: Individual students reside at the atomic core of all K-12 data. The seemingly simple task of aligning students to data across data Domains and Sources still remains one of the biggest and most time consuming challenges schools face. The more applications schools use, the more student reference fractures appear.  Ideally, every student should have data that follows them by way of alignment to classes, groups and cohorts as well individually.

What to do: Mandate a standard Data Alignment Policy. Every application must integrate a unique student ID or school student email address as the key field against which all data is recorded. All learning data must also be ‘extractable’ and explainable from the various applications used. If your applications don’t support your student data alignment policy, find new applications. Simple. This includes diagnostic testing applications. This sounds harsh but face it. If you can’t see into the data, why bother collecting it?

1.2 Alignment of Data Sources – consider 3 words – Volume – Velocity – Variety

10 seconds : Data comes from many systems, in a variety of formats, in large volumes, at different cycles and velocities. Understanding the Volume, Velocity and Variety of your Data Sources, and the Data Elements contained within, is really important for aligning  data with your teachers.

Bigger Discussion: Dealing with data Volume, Velocity and Variety is a challenge. Variety describes the richness of data in each Data Source. Volume and Velocity describe the quantity and pace at which data is collected in your school. When you have lots of varied data that contains critical storylines, you need to take a breath and think about how these data sources can align to what you are looking for.

NAPLAN or SAT’s for example have hundreds of available Data Elements per student across five domains of testing – this represents data Volume and Variety at its best. Luckily NAPLAN and SAT’s are only once a year. There are also very simple Data Sources like Attendance data where the Variety and Volume rates are low but the Velocity at which teachers need to know is instant.

What to do: With varied data surrounding schools, the best option is to understand all of your key data supply lines, their similarity, the volume and variety of information they carry and the velocity at which the supply lines of data move. Work out what you really need to know! Quick moving data builds learning stories near to real-time. Lumpy legacy data is just that, Be brutal with data, consider only what can add value to a story for a teacher in class. If you get that right, most other data demands fall into line.

1.3. Align Metadata across Data Sources within Data Domains

10 seconds : All data recorded across K-12 schools share common ‘metadata’ attributes like Year, Class, Subject and Term. Sometimes these attributes are not obvious, contained or captured in base Data Sources. That translated means, you must make it easy for teachers to find consistent comparison data between different Data Sources across Domains and the way you do that is via metadata.

Bigger Discussion:  
Remember the 6 Data Domains I discussed in Blog 2? Every Data Source will fit logically into one of these Data Domains. From within these Domains stories are built.

Within Data Domains, distinct Data Sources will have differing scales, ranges of scores and skills tested. Try to align as many common metadata points as possible. For example, snapshots like NAPLAN, SAT , PAT and ALLWELL are all Data Sources in the Diagnostic Domain sharing common metadata like Year, Cohort, Test Name, Subject, Date, Band (possibly). This data is fundamental to building consistent inquiry models across Data Sources.

Then there are less structured Data Sources with more data diversity, much of it being text based commentary. These Sources are more transactional in nature, are more variable across Year groups and come from different source systems with differing volumes and velocities. Pastoral and Formative Assessment platforms, (as an example), all add daily data volumes across many students at multiple year levels in a school, especially year 9!  Every time data is recorded about a student in a different year group, for example, data changes.

What to do:  Aligning Data Sources to Data Domains and then defining Data Elements is the task. It’s a great process to get you in touch with your inner data and actually best done by non-data people. The Elements like Score, Grade, Percentage, Curriculum link, Skill and Sub Skill should be aligned between different Data Sources. Some Data Sources also hold unique Data Elements that cannot be overlooked and need to be blended in.

Step 2. Create learning stories with a mantra of connecting with teachers rather than charts

The biggest challenge remains. How to simply and quickly connect audience-centric storytelling to leaders, teachers and students?  The intention is as simple as the Marie Kondo mantra of life, happiness and meaningful surrounds. Declutter our lives, time, thinking (and family…yikes) by removing the ‘stuff’ that really doesn’t make a positive difference.

For schools, that means move incomplete data and complex charts out of the way. Focus on the connection you are trying to make with your leaders, teachers and students. Complex presentation layers do not correlate with greater connection to your audience. Teachers find no joy in data if there is no teaching and learning context.

Brainstorming with Teachers and Leaders always helps declutter what really matters to them. We spent time in a ‘Kondo’ style session with some great teachers. The floodgates opened and in minutes we had 10 points. “Can we….”

  1. See all important data in one place without having to interpret a chart?
  2. Read the highlights of the data summarised in simple sentences?
  3. Simply link into my classes and students?
  4. Bring Pastoral and academic data into view for any student or cohort?
  5. Cherry pick data from different Data Sources and view them side by side?
  6. Set alerts to monitor data events that draw my eye to look more into?
  7. Quickly identify interventions and differentiation?
  8. Follow hunches into comparisons across results from different data?
  9. See a whole-of-student view across all sources for handover year on year, term on term?
  10. Engage Parents and Students with the same full learning story.

Remember that running right alongside this wish list, there are six levels of audiences (School, Cohort, Class, Teacher, Students and Parents) that you have to plan for. My advice is to use the same structure and data aggregation logic across all layers of audiences.  Rubik cube logic, again!

For a teacher looking across the cube, the cube represents their class or Cohort. For a Student, the cube represents their individual learning story across all Domains for all years.

All of a sudden, diverse, multi-data source inquiries that end up looking like these, are all by your design.

What do you do with all of this data? How do you get stories moving?

10 seconds : Focus on the challenge of getting stories and data into the hands of your teachers, simply.  Teaching and Learning is a team sport played daily, not a science or IT project. Consensus now suggests there is limited impact in giving teachers multiple charts, many more system touch points and dumps of learning analytics. I think the mission to be accomplished is to actually deliver learner insights from data, rather than place another data fed elephant in the room.

What to do:  With a robust structure around data – where are you going to keep it and how will you mobilise it?  K-12 schools have LOADS of data. Right about now is when technical people start talking up terms like Data Lake, BI tools and data Warehouse, all the way back to good old Excel.

You have three main options really. All are directly related to bravery and investment limits along with getting an extensible result that works for you.

  1. You can call in Data ‘analytics’ experts to define all your data. Keep in mind that you will probably be teaching them about your data Sources and Elements! Usually this starts the iterative process of asking how you would like your charts to look to solve hypotheses for leaders, teachers, students and parents. Understand the all up hours and license costs of BI tools before you start. Bravery Level – HIGH . Financial Investment needed – HIGH, Result – ALL UP TO YOU
  2. You can start configuring a Data Lake and all of the engineering that goes with that yourself. Many times the data analytics people from Option 1 will suggest or do something like this. Ask yourself the question “Are you a school or an IT bespoke shop?” Bravery Level – HIGHEST. Financial Investment needed – HIGH, Result – ALL UP TO YOU
  3. You can work with a supplier that has a turnkey approach to Data Sources and Data Elements in K-12 schools. Repeatable data integration done for a school rather than by a school will produce a result that can be used by teachers instantly. Literatu Learning Ledger is a turnkey platform for K-12 Schools. Remember there is no unique IP in common data sources. IP comes from doing something of value with the data. Bravery Level – VERY CONTROLLABLE,  Financial Impact – LOW, Result INSTANT

There are 10 important questions to ask yourself and others, as you look to build data stories from the data you have. What you have to maintain is Alignment and currency of data. Don’t’ let data stories go off with a lack of refresh, build and iterate them across the school one by one. Look for feedback and always try to answer the ten questions teachers asked us, above.

In terms of picking a partner to help unfold data in your school, you have to think about these 10 core questions.

  1. If we have 10 Data Sources today, what happens when I blend another one? How do we rewrite and blend all the charts and rules?
  2. Can data loads be automated and standardised so that we don’t have an IT budget blowout? Is there a service that can help do this reliably?
  3. We don’t really know how teachers will consume the data we have or how we ideally want data to be presented. Can our teachers build dashboard concepts share them and discover what works ourselves? What if we really don’t like charts?
  4. Can we aggregate Data Elements on demand, from any Data Source ?
  5. Do teachers have to know how to use data analytics packages to inquire into data? Is there an extra license we need to purchase to do this?
  6. Can teachers include their spreadsheets as valid Data Sources?
  7. Can we see data across weeks, terms and years, making handover of Students easier as they move ahead?
  8. Can we look across Cohort and House level data views?
  9. Can Students and Parents log into the same system as Teachers?
  10. Can we develop learning stories in a narrative format – rather than charts?

If you cover off these questions, you will have made a good start. Managing data sources is one thing, building a data culture comes next. What comes first? Show data, talk about what you see, find what you need and repeat. Until you do this, you won’t be able to guide a culture with any supporting layers of inclusion. Last I checked, culture was not formed from policy but rather an unconscious competence and repeatable behaviour.

Show data, talk about what you see, find what you need and repeat. When all of this happens by default, across most of your teachers, you will build a data culture to support teaching and learning decision making for years.

Mark Stanley is CEO and Founder of Literatu.

Literatu Learning Ledger builds a whole-of-student view across multiple data sources for K-12 schools, globally.