Material Insights · Understanding the Material World

How science knows: observation, models, measurement and testing: Key terms, quiz and sources

All courses / Understanding Science / Module 1: How science knows: observation, models, measurement and testing Wrap-up

How science knows: observation, models, measurement and testing: review

Key terms

Observation
A recorded fact gathered by the senses or by instruments.
Hypothesis
A proposed explanation that makes a prediction which a test could prove wrong.
Scientific method
The logic of testing ideas against nature, not a fixed list of steps every project follows.
Falsifiable
Able to be shown wrong by some possible observation.
Control group
The group that does not receive the change, used to show what would have happened anyway.
Independent variable
The one factor an experimenter deliberately changes.
Dependent variable
The quantity measured to see the effect of the change.
Replication
Repeating a study, by the same team or others, to see whether the result holds.
Blinding
Keeping participants or measurers unaware of who received which treatment.
Fact
An observation that careful, independent observers agree on and can repeat.
Model
A simplified stand-in for something complex, used to explain or predict.
Scientific law
A compact statement, often an equation, of a regularity found in nature.
Theory
A broad, well-tested explanation that ties many facts and laws together.
Pseudoscience
Claims that look scientific but avoid being tested or corrected.
SI
The International System of Units, built on seven base units tied to constants of nature.
Scientific notation
Writing a number as a figure between 1 and 10 times a power of ten.
Order of magnitude
A factor of ten; numbers differing by one power of ten differ by one order.
Random error
Scatter in readings that falls on either side of the true value and averages down.
Systematic error
A bias that shifts every reading the same way and does not average away.
Precision
How closely repeated readings agree with one another.
Accuracy
How close a result is to the true value.
Correlation
Two quantities moving together, which alone does not show that one causes the other.
Technology
The practical use of knowledge to make tools and systems.

Try it in the table

  • Open the Science table node 'Place-value numerals with zero' and write 4,070 and 4.07 x 10^3 side by side. Say in one sentence what the zero does in each.
  • Read the node 'Probability and statistics' and use its ideas to estimate the margin of error for a poll of 1,600 voters, then for 100 people. Which poll would you trust, and by how much?
  • Look at the node 'Germ theory of disease' and list the observations that a competing 'bad air' theory could not explain. What prediction separated the two?
  • Open 'Special and general relativity' and find the eclipse prediction. Redo the comparison of 1.75 and 0.87 arcseconds and say why a 1% measurement could tell them apart.
  • In the engineering course, read 'What engineers do' and 'Thinking in estimates'. Write three differences between a scientist's question and an engineer's for the same bridge.
  • Weigh a 1 kg bag of sugar five times on a kitchen scale, lifting it off between readings. Work out the mean and the spread, then ask whether the scale or the bag is more likely to be wrong.

Final quiz

Which feature makes a hypothesis scientific?

A hypothesis earns its place by risking failure. An idea that fits every possible result tells us nothing, however popular or hard to disprove.

A seed company compares its seed in one village with a rival's seed in another village. Why is the comparison weak?

Two differences are mixed together. Alternating strips in the same fields would give both seeds equal conditions.

Newton's gravity is still used to plan most spacecraft paths even though relativity is a better theory. This shows that

A better theory extends the range of the older one. Where gravity is weak and speeds are low, the old equations agree with it to high accuracy.

Which statement uses the word 'theory' as scientists do?

In science a theory is a well-tested explanation. The other statements use the everyday sense of a hunch or wrongly treat it as a lesser law.

The Earth's circumference is 4.0 x 10^7 m and the speed of light is 3.0 x 10^8 m/s. How many times could light go round the Earth in one second?

Divide: 3.0 x 10^8 / 4.0 x 10^7 = 0.75 x 10^1 = 7.5 times.

A thermometer reads 1.5 degrees too low at every temperature. Averaging 1,000 readings will

Averaging removes random scatter, not a fixed offset. Only comparison with a trusted standard exposes a systematic error.

A poll of 400 random voters has a margin of about

One over the square root of 400 is 1/20, which is 5 percentage points. To halve it, you would need four times the sample.

Cities with more fire engines have bigger fires. What is the most reasonable reading?

Large fires call out more engines, and big cities have both. A hidden third factor or reverse cause explains the pattern, so an experiment or mechanism is needed to claim cause.

Which of these is a question that science alone can settle?

The pH can be measured and repeated. The others rest on values, taste or priorities, which evidence informs but cannot decide.

A claim of a new room-temperature superconductor has been shown by only one lab and cannot be repeated by others. Best response?

Replication is the community's safeguard. The claim is neither proved nor disproved; it is waiting for independent checks, which is how errors are usually found.
Next module: Motion and force: from Aristotle to Newton →

Sources for this module