Thinking about our common data moves with learners

November 16, 2022

There are many things that we can do with data…but let’s be honest time is a limited resource! And when we add these data skills into the mix with our content learning objectives and all of the other things we are asked to do on a daily/weekly basis, it can feel overwhelming.

That is not a good feeling. And certainly not what we want!

But if time is limited, I also wonder if we are using the time we do have as efficiently as possible. So maybe it is worth thinking about what we are spending our time with data doing.

What are the things we are having our students do with data?
What are we not having our students do with data?

There are no right or wrong answers here, but more a chance to pause and reflect for a moment, so that we can see if we can be more efficient with our limited time.

How are we spending our time with data?

Here are some recent responses from groups of science and math teachers around the country in terms of “What do your students do most with data?”

What do you notice across these groups (I mean, besides the fact that the category options are not exactly the same as we adjusted the question based on the audience we were working with :))?

Some things that pop out to me…

Many of us are having students “describe & analyze patterns”. Woohoo!
Lots of us are having students “make a CER” or “claim from data”. That tracks.
Lots of us are having students make a graph (but very few of us are having our students make multiple graphs). Hmm, interesting.
Few of us are having students looking at, discussing, or visualizing variability. Hmm, also interesting.
And very few of us are having students “think statistically” with data. Not surprising, but…

Meaning…even though these were responses from teachers at different styles of workshops, in different geographic regions, and who wear different hats in our K-12 education system there are a lot of similarities of what we are doing with our students with data.

In many ways this is great! Helping students to build their skills in making graphs (visual data), describing what is going on in the data (describe & analyze patterns), and making sense of what the data mean (interpret data to learn something) are key components of working with data.

They are also aspects that quickly come to mind when thinking about data as adults and are pretty clearly articulated in science and/or math standards.

However, a few other things come to mind too.

First, there are multiple “moves” that we do for each of these steps besides just a graph, a pattern fit/ID, and a CER.

Second, there are lots of other things that we do when working with data (all those other functions outlined here —>).

Now, do not get me wrong. I am NOT suggesting that our students do Every. Single. One. of these things each time they work with data. That would be “bonker-balls” (as my kids like to say).

But it does make me wonder: are we providing students opportunities to practice the range of skills that go into working with data within and across our classrooms? The data above — from various educators across grade levels and subject areas — indicates maybe not as much as we would like.

If you are wondering, what are the skills (or various steps / functions / tasks / moves) that go into working with data? A colleague and I put together a resource to help articulate what these are and how novices build these skills over time. You can check them out in the Building Blocks for Data Literacy.

How else can we spend our time?

Ok, so that is a LOT of things we can do with data…but again we have limited time!

So, where could we start? What could it look like to add one new aspect into our work teaching with data? Here I suggest 2 to consider to pick from to run with this winter. Let’s explore.

#1 - Incorporate Statistical Thinking into all Data Activities

We explore 5 easy ways to do this with ALL students — from our Pre-Kindergarteners up through our AP students — in our “Embrace Statistical Thinking” workshop in the Data Literacy Series…but let’s explore an easy one now that we can all start incorporating into any of our activities this week.

The first thing that we can help our learners think about in any activity they are doing with data is to consider “the sample”. We can NEVER measure every possible value of every possible variable/attribute of the phenomenon we are investigating…1) there is just only so much time and 2) we don’t know everything about the phenomenon that’s why we are investigating it ;). So we ALWAYS only have a sample of possible data.

Why is that important? Well if the sample is representative of what we are investigating then that means we can make different kinds of conclusions and inferences from the data then if the sample is not representative (in fact, this can cause lots of problems).

Ok, so how do we help our students with this? Unless we are teaching current events or social studies…I would not recommend using election data.

Instead, each time we have students do any of the following things with data this is applicable:

Collect their own data, or
Look at someone else’s dataset, or
Look at a graph someone else made.

Before we jump in and push forward to “get the task done” we can pause to ask the questions like…

What is our sample?
What/who is in the sample? What/who is not in the sample?
Is the sample big enough so that we feel confident we can say something about our question?
How was the sample selected?

What could this look like with an actual lesson?

Here is a graph from the “Fertilizing biofuel crops may release of greenhouse gases (Digital Data Nugget)”.

Before we ask students to make sense of the data for the provided question, we can instead encourage students to read the provided background text FOR INFORMATION ABOUT THE SAMPLE.

For example, students will learn that scientists in the experiment:

Looked at fields of switchgrass, a perennial grass native to North America, a promising biofuel.
Set up eight fertilization treatments (0, 28, 56, 84, 112, 140, 168, and 196 kg N ha−1).
In four replicate fields of switchgrass.
Total of 32 research plots.

Then we can ask them the questions above and get towards the ultimate question of “so what do we think, is this a good sample to get at our question?” The key here is that learners pause and consider the data they have before blindly jumping into making a CER or conclusion from the data.

Will they need help with answering these questions at first? Absolutely! But it is time worth spent.

#2 - Have students organize and process their datasets

Helping students better understand what is going in their data can come from opening up what kind of datasets our students are working with and what we ask them to do BEFORE we visualize the data.

The reality is that how we collect the data (typically in a wide format) is not often conducive for exploring or graphing our data (typically programs want it in a long format). Beyond the organization of the data values in rows and columns in our data tables or spreadsheet programs (**note check out the “Organizing Data to Explore It” page for more resources on this), there also often requires some decisions about the dataset. For example,

Which subset of the collected data do we need to investigate our question? Do we need to pool across different datasets to fully get at our question?
Are we trying to compare one group to another in our dataset? If so, how can we easily make that comparison?
Overall what do we see in a group or category from our data? What kinds of summaries or calculations will help us better investigate our question?

These kinds of decisions, and others, are common place for those who work with data on a daily basis. Based on their work with integrating data into various K-12 classrooms, Tim Erickson and colleagues call these kinds of decisions and the resulting work with the dataset “data moves” (in the similarly titled article “Data Moves” in Vol 12, Issue 1 of the Technology Innovations in Statistics Education available for free here).

This may sound weird, overwhelming, not something you have time for…but I think it is worth some consideration for a few reasons:

This can actively put learners in the drivers seat to make decisions about what to do with the data (e.g., they have to make decisions of what to do rather than just mindlessly execute tasks, which research indicates is what is a prerequisite for better retaining the skill),
Time spent working with the data before visualizing it gives you a broader and more in depth understanding of what is and what is not in the dataset (aka what seems like “time off task” actually gives our learners a better understanding of what is in the dataset), and
This is what professionals do with data (so if we are really wanting to help our students build their skills and develop their understanding of the practices of STEM…then this needs to be a part of it).

Ok, so if any (or all) of those resonated as something you are interested in creating in your classrooms, the next question is how can we get started with this (without it being a time suck)?

Glad you asked!

We put together a quick cheat-sheet resource for “Choose Your Own Google Sheets Adventure” to practice and explore some of these data moves. The intention is to provide a practice dataset and some suggested ways to try a variety of data moves to help spark some ideas of what you may want to do in your classroom.

The resource includes information about what each is, why we do it, how to do it in Google Sheets, and sample prompts for the provided dataset for each of these data moves:

Cleaning & trimming data
Creating cross references
Filtering
Sorting
Splitting data
Using equations
Conditional Formatting

What now?