Open Education and the Role of Open Data in Supporting Learning

Why real data is useful in a classroom

Textbook exercises often use tidy, invented numbers. Real datasets are messier, and that is part of their value. Students meet missing entries, odd units, unexpected spikes and questions that do not have one answer. Working with them can make topics in math, geography, science and social studies feel connected to something concrete.

This article is about the classroom side: what learners and teachers do with open data. The separate question of how institutions and researchers publish data is covered elsewhere. As with any teaching method, evidence of how much it improves learning is mixed and depends on how well the activity is designed.

Where classroom datasets come from

Teachers and students can draw on several kinds of source:

• Weather and climate records published by public agencies.
• Population, census and household figures from national statistics offices.
• Environmental measurements, such as air quality or water readings.
• Transport timetables and traffic counts released by local authorities.
• Data the class collects itself, such as plant growth or reaction times, then shares with a clear license.

Before using any dataset, check the license and the documentation. These tell you whether you may share the work you produce, and what the columns actually mean.

A project in five steps

A simple structure works for many ages, with the difficulty adjusted to the group.

1. Ask a question. Something narrow, such as “Was last month warmer in the morning or the afternoon?”
2. Find and inspect the data. Open it in a spreadsheet and look at the columns, the units and any blank cells.
3. Clean it. Remove or mark obviously impossible values and note what was done.
4. Analyze and chart. Calculate summaries and draw a graph suited to the question.
5. Explain cautiously. Say what the data suggests, what it cannot show and what else would be needed.

Recording each step in writing is as valuable as the final answer, because it shows students that data work is a chain of choices.

A small worked example

The numbers below are invented to illustrate the idea; they are not real measurements. Suppose a class records five morning temperatures in degrees Celsius: 12, 14, 13, 15, 16.

The sum is 12 + 14 + 13 + 15 + 16 = 70. Dividing by five readings gives a mean of 14 degrees. Sorting the values gives 12, 13, 14, 15, 16, so the median, the middle value, is also 14. The range is the highest minus the lowest, 16 − 12 = 4 degrees.

Now add a faulty reading of 41 from a sensor error. The sum becomes 111 and, divided by six readings, the mean is 18.5 degrees. The median of the six sorted values (12, 13, 14, 15, 16, 41) is the average of 14 and 15, which is 14.5. The mean jumped by 4.5 degrees while the median moved by 0.5. This is a concrete way to see why people compare averages explained side by side, and why a single bad value deserves a closer look rather than automatic removal.

Skills that open datasets can build

• Graph reading, and choosing a chart type that fits the question.
• Spreadsheet basics, such as sorting, filtering and simple formulas.
• Judging sources. Who collected this, why, and how? These habits belong to information literacy more broadly.
• Noticing uncertainty. Real measurements have error, an idea explored under measurement uncertainty.
• Starting to program. Older students can load a file and compute summaries in Python, which is a natural step after python beginners material.

Pitfalls to plan for

Too much data. Large files can overwhelm a first project. Give students an extract of a few dozen rows before the full set.

Sensitive content. Avoid datasets that include identifiable individuals. If students collect data about themselves or classmates, think about what is appropriate and follow school policy.

Hidden bias. Data records what was measured, by whom and where. Gaps in coverage may reflect gaps in who was counted.

Overclaiming. Students often jump from a pattern to an explanation. Prompt them to consider other explanations and to separate a pattern from its cause, a habit explained under correlation causation.

Technical barriers. Not every household has a computer or reliable internet, so activities might need to work offline or in school hours. This links to wider concerns about the digital divide.

Sharing what students make

When students publish charts or short reports, they take part in the same open culture that the data came from. They should credit the source as the license requires, describe the method and be clear about limitations. If the class dataset is shared, a short description of each column helps the next person.

Working with open data does not guarantee deeper understanding, and results vary by group and task. But when the question is clear, the dataset is manageable and the conclusions are modest, it gives learners practice at something useful: asking questions of evidence and being honest about what the answer is.