describe the association in a scatter plot as positive, negative or none, linear or nonlinear, and find clusters and outliers.
Phase 4: Build
Do Now. Scientists and engineers at a federal lab wrote a handbook on how to study data. Read what it says scatter plots are for, then predict.
A plot shows hours of screen time after 8 p.m. on the x-axis and hours of sleep on the y-axis. What pattern do you expect?
The reading opens over the lesson; close it to come back to exactly where you were.
Read the whole cloud of points. Do not focus on one dot. Step back and ask NIST’s questions: is there a pattern, which way does it go, is it straight or curved, and does anything stand apart?
Direction. A positive association means y tends to increase as x increases: the points rise from left to right. A negative association means y tends to decrease as x increases: the points fall. No association means the points show no up or down trend.
Form and features. If the points lie close to a straight line, the association is linear. If they bend, it is nonlinear. A cluster is a group of points close together. An outlier is a point far from the overall pattern.
| Student (sample) | P | Q | R | S | T | U | V | W |
|---|---|---|---|---|---|---|---|---|
| Screen time after 8 p.m. (hours) | 0.5 | 1 | 1 | 1.5 | 2 | 2.5 | 3 | 3.5 |
| Sleep that night (hours) | 9.5 | 9 | 9.25 | 8.5 | 8 | 7.75 | 7 | 6.5 |
Worked example 1: Leah reads the Screens and Sleep data. Leah plots the sample table. As screen time goes from 0.5 to 3.5 hours, sleep goes from 9.5 down to 6.5 hours. The points fall from left to right, close to a straight path. Her statement: “The sample shows a strong negative linear association: students with more screen time after 8 p.m. tended to sleep less.” Notice the word tended. It describes a pattern, not a rule for every person.
Worked example 2: Omar finds an outlier. A new sample student K reports 9 hours of sleep and a quiz score of 62. In the sleep and score data, students who slept about 9 hours scored about 88. Student K’s point sits about 26 points below the pattern, far from every other point. Omar marks it as an outlier and writes a question beside it: “Was K sick that day, or was the score recorded wrong?” He does not erase it. An outlier is a question, not a mistake to hide.
Clusters tell stories too. If one group of points sits at 6 to 7 hours and another at 8.5 to 9.5 hours, ask what the groups have in common, such as students with early sports practice.
Picture or sketch each plot before you answer.
Write the description for your panel.
Plot the Screens and Sleep sample on grid paper. Write a two-sentence description using direction, form and strength, and the word tended.
Add sample student K (9, 62) to your Sleep and Scores plot. Circle it. Write two questions a data team should ask before deciding what to do with it.
Use the reading. Pick two of NIST's questions and answer them for the Sleep and Scores sample.
A pattern is about the group. An outlier is about one point.
Exit ticket: sketch three tiny plots that show positive, negative and no association. Label each one.
Pacing: warm-up and reading 7 minutes, mini lecture 9, guided practice 13, independent 9, closure 2.
Likely misconceptions: reading negative as bad; calling any rising curve linear; deleting outliers without a reason; treating “tended to” as “always”.
Supports: a three-panel anchor chart (positive, negative, none) with arrows. Stretch: invent a 6-point data set with a negative nonlinear association.
Answer guide: warm-up: points fall; hot cocoa: negative; outlier (4, 20) (the pattern 2x + 1 gives 9); positive; false; nonlinear; 8.5 down to 7 hours. Shoe-size item: “Shoe size and height show a positive linear association: taller students tended to have larger feet. One point is an outlier, with a very large shoe size but an average height.” Omar’s example: the pattern gives about 88 at 9 hours, and 88 − 62 = 26 points below.
National Institute of Standards and Technology, NIST/SEMATECH e-Handbook of Statistical Methods: 1.3.3.26 Scatter Plot · U.S. government work, public domain
The National Institute of Standards and Technology, or NIST, is a federal science lab. Its online statistics handbook says the purpose of a scatter plot is to check for a relationship between two variables. The variable on the vertical axis, Y, is usually the response. The variable on the horizontal axis, X, is one we suspect may be related to it.
The handbook lists questions a scatter plot can answer. Are X and Y related? Are they related in a straight-line, or linear, way? Are they related in a curved, nonlinear, way? Does the spread of Y change as X changes? Are there outliers, points far from the rest?
NIST calls the scatter plot the most heavily used graph for finding relationships between variables. Its example gallery shows plots with no relationship, a strong positive straight-line pattern, a strong negative straight-line pattern, curved patterns and a plot with an outlier.
The handbook also says how scatter plots are usually drawn: a mark at each data point, with no line connecting the points. That way nothing appears on the screen except the data. Our dashboard will follow the same rule.
Source: National Institute of Standards and Technology, NIST/SEMATECH e-Handbook of Statistical Methods: 1.3.3.26 Scatter Plot (itl.nist.gov)