Secondary 4 scatter graphs become easier when students separate association from certainty. This Mathematics tuition guide for Punggol families explains positive and negative correlation, outliers, lines of best fit, interpolation, extrapolation and cautious interpretation through original examples.
A student may recognise an upward pattern but write “x causes y”. Another may draw a line of best fit through the first and last point rather than through the overall trend. The mathematics is partly graphical and partly about what the graph does—and does not—justify.
At eduKatePunggol, Secondary 4 Mathematics is taught in focused groups of up to three students in 1.5-hour lessons near Punggol MRT. This guide supports the wider Secondary 4 Mathematics year plan. Match practice to the student’s actual subject level and school programme.
A scatter graph compares paired data
Each plotted point represents a pair of values from the same observation.
For example, one point might represent a student’s hours of revision and test score, or a day’s temperature and electricity use.
The horizontal and vertical axes must be read before interpreting the direction of the pattern.
Positive correlation means the variables tend to increase together
If larger x-values generally occur with larger y-values, the scatter shows positive correlation.
If larger x-values generally occur with smaller y-values, the correlation is negative.
If no clear directional pattern appears, there may be little or no correlation.
Worked example 1: describe a pattern without overclaiming
Suppose a scatter graph of revision hours against score rises from left to right.
A suitable description is:
There is positive correlation: students with more revision hours tend to have higher scores.
This does not prove that revision time alone caused the score. Other factors may also matter.
Strength depends on how tightly the points follow the trend
A strong correlation has points clustered relatively close to the trend. A weak correlation has more scatter around that direction.
Do not judge strength only from whether the graph slopes up or down. Look at how dispersed the points are.
Outliers deserve attention
An outlier is a point noticeably separated from the main pattern.
It may represent a genuine unusual observation, a measurement issue or another factor not captured by the graph. The graph alone does not automatically tell us why the point is unusual.
A line of best fit represents the overall trend
A good line of best fit should pass through the centre of the scatter with a reasonable balance of points above and below it.
It does not need to pass through every point. In fact, a scatter graph usually exists because the points do not lie exactly on one line.
Worked example 2: estimate from a line of best fit
Suppose a line of best fit for x between 2 and 10 passes near the points (4, 30) and (8, 50).
The approximate gradient is:
(50 − 30)/(8 − 4) = 5.
An approximate line through (4, 30) with gradient 5 would give y ≈ 35 when x = 5.
Because the line itself is fitted to scattered data, the estimate should not be presented as exact unless the context provides an exact model.
Interpolation is usually safer than extrapolation
Interpolation estimates within the observed x-range.
Extrapolation extends the trend beyond the observed data.
The further we move beyond the known range, the less certain it is that the same relationship will continue.
Worked example 3: distinguish interpolation and extrapolation
If observed x-values run from 10 to 50:
- estimating y at x = 35 is interpolation;
- estimating y at x = 80 is extrapolation.
The second estimate depends on an assumption that the trend continues well beyond the collected data.
Correlation does not automatically prove causation
Two quantities can move together because one influences the other, because both are influenced by a third factor, or by coincidence within the observed data.
In an examination answer, use language such as “there is a positive association” or “higher x-values tend to be associated with higher y-values” unless the context justifies a stronger causal statement.
Graph scale can change the visual impression
A compressed axis can make a trend look flatter; a stretched axis can make differences appear dramatic.
Always read the numerical scale before judging the pattern.
This connects to the Secondary 4 statistics guide, where scale reading also matters for cumulative-frequency curves and box plots.
How we diagnose scatter-graph mistakes
Axis error: x and y quantities or scales are misread.
Correlation error: direction and strength are confused.
Best-fit error: the line is forced through endpoints instead of the overall trend.
Prediction error: an extrapolated value is treated as equally reliable as an interpolation.
Interpretation error: correlation is written as proof of causation.
Why the three-student format helps
In a group of up to three students, one learner can describe the direction, another can assess strength and another can explain whether an estimate is interpolation or extrapolation. The tutor can see whether the language matches what the graph really supports.
What a 90-minute lesson could look like
An illustrative lesson could use ten minutes for axis and scale reading, twenty minutes on correlation, twenty minutes on outliers and best-fit lines, twenty minutes on predictions and twenty minutes for independent graph interpretation, error review and continuation work.
Repair, stabilisation and extension
Repair: use clear plots with obvious positive, negative and no-correlation patterns.
Stabilisation: mix strength, outliers, line-of-best-fit estimates and interpolation/extrapolation.
Extension: compare plausible interpretations and ask students to state what additional evidence would be needed before making a causal claim.
Try a short independent check
- A scatter rises from left to right with points tightly clustered. Describe the correlation.
- Observed x-values are 5 to 40. Is prediction at x = 20 interpolation or extrapolation?
- Observed x-values are 5 to 40. Is prediction at x = 70 interpolation or extrapolation?
Answers: strong positive correlation; interpolation; extrapolation.
What progress should look like
- axes and scales are read before the trend is described;
- correlation direction and strength are distinguished;
- outliers are identified without inventing causes;
- best-fit lines represent the overall scatter;
- interpolation and extrapolation are distinguished;
- causal claims are not made from correlation alone.
Punggol class details and consultation inputs
eduKatePunggol Secondary Mathematics tutorials run for 1.5 hours in groups of up to three students near Punggol MRT. Confirm current class availability, fees and meeting arrangements directly.
Bring the student’s subject level, examination year and recent data-interpretation questions. Original graph markings are useful because line placement and scale reading can be inspected directly.
Frequently asked questions
Does positive correlation mean one variable causes the other?
No. Correlation describes association. Causation requires additional evidence.
Must a best-fit line pass through every point?
No. It represents the overall trend of the scatter.
Describe only what the data supports
Return to the Secondary 4 Mathematics year plan for the wider SEC runway. For broader data spread and quartiles, use the statistics guide.
Read the axes, describe the trend and keep predictions inside the evidence where possible. Families can WhatsApp eduKatePunggol with recent school work to discuss a suitable next step.
