Observations, measurements and variables
- Syllabus
- 2024
- Topic
- —
- Level
- —
Learning objectives
Measure, record and present only the precision supported
A measurement is appropriately precise when the instrument and reading method can genuinely support the digits recorded. A measurement needs a numerical value and unit; an observation should describe exactly what was detected without adding an unsupported explanation.
Use this sequence: choose an instrument with a suitable range and resolution → find the value of one smallest division → align the instrument correctly → read from the correct viewing position → record the value with its unit and only the supported precision.
Example: if 20 kPa to 30 kPa contains five equal spaces, each smallest division is 2 kPa. A pointer at the second division after 20 reads 24 kPa. For a vertical height, fix the metre rule with zero at the reference height, keep it vertical with a set square or plumb line, and read at eye level.
Record repeated results methodically in a table. Put quantity and unit in the heading, such as “power / W”, then enter only numbers in that column. Measurements from the same instrument should normally use consistent decimal places.
More digits do not automatically mean greater precision. Do not invent decimal places beyond the scale or display, attach units separately to every table entry, or read an analogue pointer from an angle that shifts its apparent position.
Identify variables by the job they do in the investigation
Variables are identified from the investigation aim and method, not from the type of physical quantity. The independent variable is deliberately changed, the dependent variable is measured as the response, and control variables are kept constant so other influences do not change the response.
| Variable role | Identification question | Gamma-penetration example |
|---|---|---|
| independent | What is deliberately changed? | number of lead sheets |
| dependent | What response is measured? | count measured by the detector |
| control | What else could affect the response and must stay constant? | source-detector distance and time used for each count |
A useful shortcut is to rewrite the aim as “the effect of X on Y”: X is usually the independent variable and Y the dependent variable. Then use the physics to find other factors that could also change Y; those become control variables. Keeping them constant makes the comparison fair because a change in Y can be linked to X.
A control variable is not the same as a control group, and it is not simply any quantity that happens not to change. It must be a relevant factor whose change could affect the dependent variable. If such a factor varies, the experiment cannot show which change caused the response.