Time to Embrace What We Cannot Say on the Way to Helping Students Identify What They Can Say with Data

ask questions & consider possible outcomes interpret data to learn something

In classrooms, we (rightfully) often focus on what we can see in the data. We ask students to identify key features, describe overall patterns, and make claims. But an equally powerful—and often overlooked—question is:

What can’t we say from these data?

Helping students understand both the possibilities and the limits of data is essential to building true data literacy. In fact, this skill sits at the core of the Next Generation Science Standards’ call for students to distinguish “observations from inferences, arguments from explanations, and claims from evidence” (NGSS, 2013). It also appears in our ELA, social studies, and other subject area standards.

When students learn to recognize what their data cannot support, they become more careful thinkers, clearer communicators, and stronger reasoners with data.

Why “What We Cannot Say” Matters

Students frequently encounter data in polished graphs and structured problems that can be construed with a level of certainty that rarely actually exists in the world. Without explicit instruction, they may assume:

Every pattern demonstrates a clear cause-and-effect relationship.
Every data visualization tells a complete story, and only one story.
Every claim is easily identified and understood by all.

But real-world data rarely works that way.

Teaching students to pause and ask, “What is beyond the scope of these data?” helps them:

Separate observation from interpretation,
Distinguish correlation from causation,
Recognize missing variables or cases,
Avoid overgeneralizing, and much more :)

This is not about limiting student thinking. It is about strengthening and deepening it.

A classroom graph with two columns labeled “What We Can Say” and “What We Cannot Say,” image generated by Firefly, Flux.1 Kontext [max]/Adobe, Feb 20, 2026

Claims vs. Evidence

At the heart of data reasoning is the relationship between claims and evidence.

Students may make statements like:

“The experiment worked.”
“This group performs better.”
“The population is increasing.”

But strong data literacy asks:

What evidence supports that claim?
What, and/or who, is and is not in the dataset?
Do we have enough data to identify evidence?
Is the evidence sufficient?
Does the evidence support other claims too?

Encouraging students to articulate the connection between data and claim builds transferable reasoning skills that extend far beyond one content area.

If you haven’t yet explored it, our article: What Can We Actually Claim from Our Data? (Feb 2020, Data Literacy 101) dives deeper into helping students ground their claims in evidence.

https://www.nsta.org/science-scope/science-scope-february-2020/data-literacy-101

Observations vs. Inferences

One of the first distinctions students must master when they are interpreting data is the difference between observations and inferences.

Observation: What we can directly state from the data at hand…this is what is used in a data-based conclusion
Inference: A prediction or logical connection drawn based on reasoning beyond the data at hand

For example:

Observation: Gravitational force increased as the mass of the objects increased.
Inference: A similar relationship between gravitational force and mass likely exists throughout the solar system.

Observation: Test scores increased after a new curriculum was introduced.
Inference: The curriculum caused the increase.

The observations are made from the actual data and support the conclusion. The data does not, on its own, support the inference.

Explicitly modeling this distinction helps students build the habit of anchoring claims in evidence rather than assumption.

For more guidance on this, see the Next Generation Science Standards framework for science practices, which emphasizes evidence-based reasoning across grade levels.

Arguments vs. Explanations

Another subtle but powerful distinction that relates to helping students make meaning of data lies between arguments and explanations.

An argument defends a claim using evidence.
An explanation describes how or why something happens.

Students often blur the two.

When working with data, we want students to ask:

Am I defending a claim with evidence?
Or am I explaining a mechanism?

Understanding this difference strengthens academic discourse and supports critical thinking and reasoning practices across subject areas.

Dataspire’s approach to data literacy emphasizes these structural distinctions so students learn not just to analyze data—but to communicate about it clearly and responsibly.

Looking for a simple breakdown of the differences between arguing from evidence vs constructing an explanation? Check out the “CER – Arguing from Evidence or Constructing an Explanation?” Digital Resource.

Teaching the Limits of Data Builds Confidence, Not Doubt

Some educators worry that focusing on limitations might confuse students or undermine confidence.

In practice, the opposite happens.

When students understand:

What the data supports
What the data does not support
Why that distinction matters

They feel more confident in their reasoning and can begin to see that limitations of a dataset or data visualization is not a weakness of that source—it is natural, always present, and intellectual honesty.

Practical Classroom Moves

Here are simple ways to integrate this thinking into everyday instruction:

1. Add a “Cannot Say” Column

Whenever students work with data, include two prompts:

What can we say?
What can we not say?

This simple structure reinforces boundaries of the data at hand for describing and analyzing the patterns as well as interpreting the data in context.

2. Use Sentence Starters

Support academic language with frames like:

“The data suggests…”
“Based on this evidence…”
“These data do not enable us to conclude…”

3. Model Think-Alouds

Verbalize your own reasoning:

“I notice this pattern, but I can’t conclude causation because we don’t have data about…”

Students learn reasoning by hearing it modeled.

4. Revisit Claims Over Time

If new data is introduced, ask:

Does our claim still hold?
What changes do we need to make with these new data?
What do we still have questions about?

This reinforces that data reasoning is iterative.

Why This Matters Now

In a world saturated with charts, statistics, and eye-catching headlines, students must learn to critically evaluate data-based claims. Teaching them to embrace what cannot be said from a dataset or data visualization is just as fundamental to helping them responsibly articulate what they can say from those sources.

This is not just a classroom skill. It is a civic skill. A literacy skill. A life skill.

When students distinguish observations from inferences, arguments from explanations, and claims from evidence, they are not simply completing an assignment. They are practicing critical thinking and healthy skepticism. And that is the heart of data literacy.

Continue the Learning

If you’d like to explore this topic further, revisit our foundational article:

👉 What Can We Actually Claim from Our Data?
It provides practical examples and classroom strategies for strengthening evidence-based reasoning across disciplines.