What you need to know
- Unit 3Check a reading that is inconsistent with the others, and comment on improvements, including extra apparatus that would reduce error.
An inconsistent point is not automatically deleted, and it is not a significance test. Improvements have to be realistic for the method you were shown.
By the end, you should be able to…
- Identify a reading that is inconsistent with the rest of the evidence.
- Say how you would check it before rejecting it.
- Propose an improvement that reduces a named source of error.
Build your understanding
Find the point that does not belong
Compare the reading with the other repeats, or with the pattern on the graph. An illustrative set of times, 38 s, 41 s, 40 s and 92 s, makes 92 s the inconsistent value. A point that sits far off an otherwise steady line is the graph version of the same idea.
Do not call every variation an anomaly. Repeats are expected to differ a little. Describe the inconsistency: it is much larger than the other differences, or it is the only point off the line. Then say what you would do, not only the label.
Check before you discard
The specification says check the reading. Read the scale again, check the unit, check that the correct row was copied, and repeat that trial if the material is still available. A transcriptionMaking an mRNA copy of a gene, using one strand of DNA as a template, in the nucleus. error is a different problem from a biological oddity.
If the repeat agrees with the rest, you can set the original aside and say why. If you cannot repeat it, you may still plot it, or show a mean with and without it, and avoid letting one point carry the conclusion. Silently deleting it hides the evidence the examiner can see.
An improvement names the error
“Use better apparatus” is too vague. Name the weakness, then the change. Subjective colour matching can be improved with a colorimeter and a blank, if the mixture is suitable for a light reading. An uncontrolled water-bath temperature can be improved with a thermostatically controlled bath and a thermometer. A gradient taken from a tiny triangle can be improved by using a large triangle on a clearly ruled graph.
The improvement must be possible. Suggesting a mass spectrometer for a school colour test is not realistic. Suggesting three more temperatures inside a narrow range is realistic when the problem was a missing optimum. Match the suggestion to the error you identified.
Make your reasoning visible
Use the numbers you were given
Point to 92 s, or to the coordinate that misses the line. A comment that could fit any experiment is too general.
One improvement, one reason
Write “so that …” after the apparatus. The reason should be the error you named, not a new aim.
Avoid this misconception
The claim: Any result you do not like is an anomaly and should be crossed out.
The correction: Check it against the other readings or the line, look for a copying or unit error, and repeat it if you can. Do not hide a point the data still contain.
Check your understanding
Try each question before opening the suggested answer.
Practice question
Retrieval and explanation · suggested answer
Illustrative repeat volumes are 8.4, 8.6, 8.5 and 13.2 cm³. Which value needs checking, and what would you do?
Show suggested answer
13.2 cm³ is inconsistent with the other three, which cluster between 8.4 and 8.6. Check the burette reading and the record, then repeat that titration. Do not delete 13.2 until it has been checked.
Practice question
Retrieval and explanation · suggested answer
A graph of rate against temperature rises smoothly except for one point far below the line. The student redraws the line through that point. Why is that a weak response?
Show suggested answer
One unchecked point then dictates the line. The point should be checked and, if it remains, should not by itself determine the relationship suggested by the other points.
Practice question
Retrieval and explanation · suggested answer
Colour standards were judged by eye and the student is unsure of the match. Suggest one realistic improvement and the error it reduces.
Show suggested answer
Use a colorimeter with a blank, if the solutions are clear enough for a valid reading, so the result is a number rather than a subjective colour match. Simply repeating the same visual match would reproduce the same judgement.
Take the idea further
Apply the same specification knowledge to an unfamiliar problem. Any extra assumptions are supplied in the question.
Stretch question
In an unfamiliar respirometer, the fluid moves 2 mm, 3 mm, 2 mm and 18 mm in four repeats. The soda lime and the organism are already set up as the question describes. Propose how to treat the 18 mm reading, and one apparatus check that could explain a systematic movement even with no organism. State an assumption.
Show a hint
Separate one odd repeat from a leak or temperature change that would affect every run.
Explore the reasoning
Check the 18 mm reading and repeat that run. If the repeat matches 2 to 3 mm, exclude 18 mm from the mean and say so. A control tube with no organism, or with glass beads in place of the organism, checks whether temperature or a leak moves the fluid on its own. The assumption is that the other conditions stay the same between the experimental and control tubes.
Recall this lesson
Lesson recall questions
Question 1 of 5: Compare the range of a 0–110 °C thermometer with its resolution in an enzyme-temperature investigation.
Unit 3 implementation and measurements
Practice question
Retrieval and explanation · suggested answer
Compare the range of a 0–110 °C thermometer with its resolution in an enzyme-temperature investigation.
Lesson: I1: Readings, range and significant figures
Try this from memory first.
Beyond the classroomA result can be reported with its doubtMeasurement · uncertainty · Enrichment, not required recall
Measurement institutes explain that a measured value has an associated uncertainty. Publishing the spread, or the reason a point was queried, is more honest than presenting every digit as exact. You are not expected to use a national calibration method in the examination.
Back to the spec: Unit 3 implementation: check an inconsistent reading, and suggest a realistic way to reduce the error. Hiding the point is not the same as reducing the error.
National Physical Laboratory: the SI unitsQuestion
Why is “we ignored the odd result” weaker than “we repeated the odd result and it was not reproduced”?
Explore a possible response
The first version gives no evidence that the point was wrong. The second shows the check the specification asks for. A point may still be reported if the repeat confirms it.