Secondary 2 Mathematics tuition in Punggol for students learning scatter graphs, correlation and lines of best fit where these ideas appear in their school Mathematics programme.
A scatter graph is not about joining every point.
It is about looking for a relationship between two variables across many paired observations.
At eduKate Punggol, our premium 3-pax tutorials teach students to describe the trend carefully, recognise outliers and avoid claiming that correlation automatically proves causation.
This guide supports our main Punggol Secondary 2 Mathematics Tutor hub and our data, statistics, graphs and probability guide.
Class size is limited to three students. Lessons are normally around 90 minutes weekly, with plotted data, interpretation, estimation and school-paper alignment.
Arrange a parent–student consultation with eduKate Singapore
Chat with eduKateSG on WhatsApp
A Scatter Graph Uses Paired Data
Each point represents two measurements from the same observation.
For example, one point might represent a student’s weekly practice time and that student’s test score.
The x-coordinate records one variable and the y-coordinate records the other.
The goal is to see whether larger values of one variable tend to be associated with larger, smaller or unrelated values of the other.
Positive Correlation
Positive correlation means that as one variable increases, the other tends to increase.
The points generally slope upward from left to right.
Examples might include height and arm span in a group of people, or practice time and performance in some settings.
The word tends matters. Not every individual point must follow the trend exactly.
Negative Correlation
Negative correlation means that as one variable increases, the other tends to decrease.
The points generally slope downward from left to right.
For example, journey speed and travel time for a fixed distance can show a negative relationship under suitable conditions.
No Correlation
If the points show no clear upward or downward pattern, the variables may have little or no correlation in the observed data.
No correlation does not prove the variables are unrelated in every possible situation. It means this dataset does not show a clear monotonic linear trend.
Worked Example 1: Describe a Trend
Suppose a scatter plot of weekly practice hours against test score shows most points rising from lower-left to upper-right.
A careful description is: there is positive correlation between practice hours and test score in this dataset.
A stronger claim such as “extra practice always causes higher scores” is not justified by correlation alone.
Correlation Does Not Automatically Prove Causation
Two variables can move together because one influences the other, because a third factor affects both, or because the pattern is partly coincidental.
A scatter graph shows association.
It does not by itself establish a causal mechanism.
This is an important interpretation habit far beyond school Mathematics.
Outliers Matter
An outlier is a point that lies noticeably away from the general pattern.
It may represent an unusual but valid observation, a measurement error, a recording error or a different underlying condition.
Do not erase it simply because it makes the graph look untidy.
Worked Example 2: Identify an Outlier
Suppose most points cluster near a rising trend, but one observation has very high practice time and very low score.
That point may be described as a possible outlier.
The next question is not “How do I remove it?” but “Is there a reason this observation differs?”
The mathematics may ask only for identification, but good interpretation keeps the uncertainty visible.
A Line of Best Fit Summarises the Trend
A line of best fit should pass through the central tendency of the point cloud, with a reasonable balance of points above and below.
It does not need to pass through every point.
If the graph shows a strong approximately linear trend, the line can be used to estimate one variable from the other within the observed range.
Worked Example 3: Interpolation
Suppose a line of best fit on a height-versus-arm-span scatter graph suggests an arm span of about 165 cm for a height of 163 cm.
If 163 cm lies within the observed height range, this is an interpolation.
Interpolation is generally more defensible than estimating far outside the data range because it stays within the region where observations support the trend.
Extrapolation Needs More Caution
If the observed data covers x-values from 10 to 30 and we use the line to predict at x = 80, that is extrapolation.
The relationship may change outside the observed range.
A straight-line trend inside the data does not guarantee the same pattern continues indefinitely.
Worked Example 4: Read an Estimate From a Line
Suppose a best-fit line passes approximately through (2, 50) and (6, 70).
Its slope is about (70 − 50)/(6 − 2) = 5 units of y per unit of x.
An x-value of 4 would therefore be associated with a y-value around 60 on this line.
The graph gives an estimate, not an exact observed data point unless a real point lies there.
Strength of Correlation Depends on How Tightly Points Follow the Trend
A strong positive correlation has points clustered relatively close to an upward trend.
A weak positive correlation still trends upward but with much more scatter.
Students should not confuse steepness with strength. A shallow but tightly clustered upward trend can be strongly correlated.
Five Common Scatter-Graph Errors
- joining data points in sequence as though the graph were a line graph;
- calling any upward-looking pair of points strong positive correlation without considering the whole dataset;
- confusing steepness with strength of correlation;
- using a line of best fit far outside the observed range without caution;
- claiming that correlation alone proves causation.
Why a 3-Pax Class Helps
One student may plot accurately and describe the trend too strongly. Another may understand correlation and draw a poor best-fit line. A third may do both correctly and overinterpret an extrapolation.
In a three-student class, the tutor can compare the same dataset through plotting, line placement and language.
An Illustrative 90-Minute Lesson
- Review coordinate plotting.
- Plot paired data as separate points.
- Identify positive, negative and no correlation.
- Discuss strength and outliers.
- Draw a sensible line of best fit.
- Use interpolation for an estimate.
- Compare interpolation with extrapolation.
- Finish with a correlation-versus-causation interpretation question.
Try Four Questions
- A scatter graph slopes generally upward from left to right. What type of correlation is suggested?
- A set of points follows a downward trend but is widely scattered. Is the correlation likely strong or weak?
- Why should a line of best fit not be forced through every point?
- What is the difference between interpolation and extrapolation?
Answers: (1) Positive correlation. (2) Weak negative correlation. (3) It represents the overall trend rather than connecting every observation. (4) Interpolation estimates within the observed data range; extrapolation estimates beyond it.
What Progress Should Look Like
- Paired data is plotted accurately.
- Positive, negative and no correlation are described correctly.
- Strength is distinguished from slope.
- Outliers are noticed without being discarded automatically.
- Best-fit lines represent the overall cloud.
- Interpolation and extrapolation are distinguished.
- Correlation is not overstated as causation.
Full Subject-Based Banding and School Scope
Students may take Mathematics at G1, G2 or G3 subject levels, and schools can sequence scatter graphs differently.
Use the student’s actual school programme to decide whether best-fit lines, strength or extrapolation are current. The central skill is disciplined interpretation of paired data.
Class Details
Format: Premium 3-pax small-group tutorials
Level: Secondary 2 Mathematics
Duration: normally around 90 minutes weekly
Location: 83 Punggol Central, Singapore 828761
Teaching approach: paired-data plotting, correlation language, outlier interpretation, best-fit estimation, guided and independent practice and school-paper alignment.
What Parents Can Bring to the Consultation
- recent scatter-graph or statistics worksheets;
- a marked school paper;
- questions where correlation language was imprecise;
- best-fit line or interpolation questions;
- the school’s current topic sequence.
Frequently Asked Questions
Does positive correlation mean one variable causes the other?
No. It shows an association in the data. Causation requires additional evidence and reasoning.
Does a steeper line mean stronger correlation?
No. Strength refers to how closely the points follow the trend, not how steep the trend is.
Why is extrapolation riskier?
It assumes the observed trend continues outside the range where data was collected.
When is tuition useful?
When the student can plot points but repeatedly misinterprets correlation, strength, outliers or best-fit estimates.
Helpful Reading and Next Step
Continue with data, statistics, graphs and probability, mean and frequency tables, and linear graphs and coordinates.
The objective is a student who can see a trend without claiming more than the data supports. Discuss your child’s current Mathematics work with eduKate Punggol.
