Grade 8 Mathematics Unit 9: Bivariate Data & Sampling
This Grade 8 Mathematics unit, Bivariate Data & Sampling, has 4 lessons that cover TEKS 8.5C, 8.11A, 8.5D, 8.11B, 8.11C. Each lesson below gives the goal, key vocabulary, a worked example and practice questions with answers you can check.
Practice this unit with interactive questions
STAAR connection: Grade 8 Mathematics is a subject tested by the State of Texas Assessments of Academic Readiness (STAAR). Texas lawmakers have voted to replace STAAR with shorter tests beginning in the 2027–28 school year; the tests are still built on the TEKS, so practicing the TEKS stays useful.
TEKS covered in this unit
- 8.5C contrast bivariate sets of data that suggest a linear relationship with bivariate sets of data that do not suggest a linear relationship from a graphical representation
- 8.11A construct a scatterplot and describe the observed data to address questions of association such as linear, non-linear, and no association between bivariate data
- 8.5D use a trend line that approximates the linear relationship between bivariate sets of data to make predictions
- 8.11B determine the mean absolute deviation and use this quantity as a measure of the average distance data are from the mean using a data set of no more than 10 data points
- 8.11C simulate generating random samples of the same size from a population with known characteristics to develop the notion of a random sample being representative of the population from which it was selected
Lesson 1: Scatterplots and association
TEKS 8.5C 8.11A
What you will learn
I can construct a scatterplot of bivariate data and describe the association as linear (positive or negative), non-linear, or no association, and I can contrast data that suggests a linear relationship with data that does not.
I can describe a scatterplot using bivariate data, association, positive, negative and linear.
Key vocabulary
- bivariate data
- Data that pairs two variables for each individual, like hours studied and test score.
Examples: (hours studied, score); (temperature, cups of cocoa sold) - scatterplot
- A graph of ordered pairs of bivariate data, with one point for each pair.
Examples: Points (1, 55), (2, 60), (3, 68); Shoe size vs. pets - association
- The pattern in how two variables change together: positive (both rise), negative (one rises as the other falls), non-linear (a curve), or none.
Examples: Study time and score: positive linear; Ball height over time: non-linear
Worked example
A table gives hours studied (x) and test score (y): (1, 55), (2, 60), (3, 68), (4, 72), (5, 80), (6, 85), (7, 88), (8, 95). Construct the scatterplot and describe the association.
- Label the axes: hours studied on x, test score on y. Choose scales that fit all the data (x from 0 to 9, y from 40 to 100).
A scatterplot needs labeled axes and a sensible scale. - Plot each pair as one point: for (4, 72) go right 4 and up to 72.
The first number is horizontal and the second is vertical. - Look at the pattern: as x increases, y tends to increase, and the points stay close to a straight band.
Up to the right means a positive association; a narrow band means linear. - Conclusion: a positive linear association. Compare a ball thrown in the air: heights rise and then fall in a curve, which is non-linear.
A linear relationship follows a line; a curve does not.
Strategy: Axes, Plot, Pattern
- Axes: label and scale both axes.
- Plot every ordered pair.
- Pattern: describe direction (up, down, none) and shape (line or curve).
Check your understanding
The points of a scatterplot form a narrow band going up from left to right. Which describes the association?
- Non-linear association
- No association
- Negative linear association
- Positive linear association
Show the answer
Answer: D. Positive linear association
Up to the right means positive, and a narrow band means the points are close to a line.
A table shows temperature (°F) and cups of hot cocoa sold: (10, 90), (20, 78), (30, 66), (40, 50), (50, 40). Which describes the association?
- Non-linear association
- No association
- Positive linear association
- Negative linear association
Show the answer
Answer: D. Negative linear association
Sales fall as the temperature rises, and the values drop at a steady rate.
A table lists hours studied and scores: (2, 60), (4, 70), (6, 75), (8, 90). How do you plot the point for 6 hours?
- Go right 6, then up 6
- Go right 75, then up 6
- Go right 6 on the x-axis, then up to 75 on the y-axis
- Go right 6, then up 70
Show the answer
Answer: C. Go right 6 on the x-axis, then up to 75 on the y-axis
The pair is (6, 75): x = hours (horizontal) and y = score (vertical).
Lesson 2: Trend lines and predictions
TEKS 8.5D
What you will learn
I can use a trend line that approximates the linear relationship between bivariate data to make predictions.
I can explain a prediction using trend line, slope in context and reasonable.
Key vocabulary
- trend line
- A straight line drawn close to the points of a scatterplot that approximates the linear relationship.
Examples: y = 5.5x + 50 for the study data; y = −1.3x + 103 for the cocoa data - interpolation
- Predicting a value inside the range of the data. These predictions are usually reliable.
Examples: Study data: predict at 4.5 hours; Cocoa data: predict at 35°F - extrapolation
- Predicting a value outside the range of the data. These predictions can be unreliable.
Examples: Predicting a score of 127 for 14 hours; Predicting cocoa sales at 90°F
Worked example
A trend line for the study data is y = 5.5x + 50 (x = hours, y = score). Predict the score for 5.5 hours and explain what the slope means.
- Substitute x = 5.5: y = 5.5(5.5) + 50.
A trend line is an equation, so predicting means evaluating it. - 5.5 × 5.5 = 30.25, then 30.25 + 50 = 80.25.
Multiply first, then add. - Check reasonableness: 5.5 hours is between 5 and 6 hours, and scores of 80 and 85 were observed, so 80.25 is sensible.
The value is within the range of the data (interpolation). - Slope 5.5: each extra hour of studying is predicted to raise the score by about 5.5 points.
The slope is a rate of change in the context.
Strategy: Substitute, Compute, Reasonable?
- Substitute the x-value into the trend line.
- Compute the prediction.
- Reasonable? Is x inside the data range, and is y possible?
Check your understanding
A trend line for study time is y = 5.5x + 50. What score is predicted for 4.5 hours?
- 55.5
- 24.75
- 77.5
- 74.75
Show the answer
Answer: D. 74.75
5.5(4.5) = 24.75 and 24.75 + 50 = 74.75.
In the trend line y = 5.5x + 50, what does the slope 5.5 mean?
- The best score is 5.5
- The score is 5.5 times the hours studied
- A student who studies 0 hours scores 5.5
- Each extra hour studied is predicted to raise the score by 5.5 points
Show the answer
Answer: D. Each extra hour studied is predicted to raise the score by 5.5 points
Slope is the rate of change: 5.5 points per hour.
A trend line for temperature (x, in °F) and cups of hot cocoa sold (y) is y = −1.3x + 103. How many cups are predicted at 35°F?
- 45.5 cups
- 57.5 cups
- −45.5 cups
- 148.5 cups
Show the answer
Answer: B. 57.5 cups
−1.3(35) = −45.5 and −45.5 + 103 = 57.5.
Lesson 3: Mean absolute deviation
TEKS 8.11B
What you will learn
I can determine the mean absolute deviation (MAD) of a data set of no more than 10 values and use it as a measure of the average distance the data are from the mean.
I can describe how spread out data are using mean, deviation and mean absolute deviation.
Key vocabulary
- mean
- The balance point of the data: the sum of the values divided by how many there are.
Examples: (4 + 6 + 8 + 10 + 12) ÷ 5 = 8; (3 + 5 + 7 + 9) ÷ 4 = 6 - deviation from the mean
- How far one value is from the mean. For MAD we use the distance (always positive).
Examples: |4 − 8| = 4; |10 − 8| = 2 - mean absolute deviation (MAD)
- The average of the distances of the values from the mean. A larger MAD means more spread.
Examples: 4, 6, 8, 10, 12: MAD = 2.4; 15, 15, 15, 15: MAD = 0
Worked example
Find the mean and the MAD of the data set 4, 6, 8, 10, 12.
- Mean: (4 + 6 + 8 + 10 + 12) ÷ 5 = 40 ÷ 5 = 8.
Add the values and divide by how many there are. - Distances from 8: |4 − 8| = 4, |6 − 8| = 2, |8 − 8| = 0, |10 − 8| = 2, |12 − 8| = 4.
Use absolute value so every distance is positive. - Add the distances: 4 + 2 + 0 + 2 + 4 = 12.
These are the orange segments in the lab. - MAD = 12 ÷ 5 = 2.4. On average the values are 2.4 units from the mean.
Interpret the answer in words.
Strategy: Mean, Distances, Average
- Mean: add and divide.
- Distances: subtract the mean and drop negative signs.
- Average the distances to get the MAD.
Check your understanding
What is the MAD of the data set 3, 5, 7, 9?
- 6
- 4
- 1
- 2
Show the answer
Answer: D. 2
Mean = 6. Distances: 3, 1, 1, 3. Sum = 8, and 8 ÷ 4 = 2.
What is the MAD of the data set 2, 2, 5, 9, 12?
- 6
- 3
- 3.6
- 18
Show the answer
Answer: C. 3.6
Mean = 30 ÷ 5 = 6. Distances: 4, 4, 1, 3, 6. Sum = 18, and 18 ÷ 5 = 3.6.
Team A scored 8, 9, 10, 11, 12 and Team B scored 4, 7, 10, 13, 16. Both have a mean of 10. Which statement is correct?
- Team A’s scores are more spread out: MAD 1.2 compared with 3.6
- Team B’s scores are more spread out: MAD 3.6 compared with 1.2
- Team B has a smaller MAD because it has larger scores
- The teams have the same spread because the means are equal
Show the answer
Answer: B. Team B’s scores are more spread out: MAD 3.6 compared with 1.2
Team A: (2 + 1 + 0 + 1 + 2) ÷ 5 = 1.2. Team B: (6 + 3 + 0 + 3 + 6) ÷ 5 = 3.6. A larger MAD means more spread.
Lesson 4: Simulating random samples
TEKS 8.11C
What you will learn
I can simulate generating random samples of the same size from a population with known characteristics and use the results to see that a random sample tends to be representative of the population.
I can explain a simulation using population, random sample, sample size and representative.
Key vocabulary
- population
- The whole group you want to know about.
Examples: All 100 marbles in the jar; All students at a school - random sample
- A sample chosen so that every member of the population has an equal chance of being picked.
Examples: Draw names from a hat; Use a random number generator - representative sample
- A sample whose results look like the population. Random samples tend to be representative, and larger samples are more reliable.
Examples: A sample with 58% blue from a 60% blue population; A random sample of 50
Worked example
A jar has 100 marbles: 60 blue and 40 orange. Simulate random samples of 10 marbles and describe what the results show.
- Population: 60% blue. A perfectly representative sample of 10 would have 6 blue.
Know the true characteristic before you sample. - Draw random samples of 10 several times (use the lab): counts of blue might be 4, 6, 7, 5, 6, 8.
Each sample is chosen by chance, so each one is a little different. - Sample percents are 40%, 60%, 70%, 50%, 60%, 80%: they vary, but most are close to 60%.
Random samples tend to be near the population value. - Now draw samples of 50: the percents cluster much closer to 60%.
Larger random samples are more reliable. Combining samples also helps: 36 blue out of 60 marbles is 60%.
Strategy: Know, Sample, Compare
- Know the population percent.
- Sample at random many times with the same size.
- Compare the sample percents with the population percent.
Check your understanding
Which method gives a random sample of 20 students from a school?
- Ask the students sitting at one lunch table
- Ask only your friends
- Ask the first 20 students who arrive
- Give every student a number and pick 20 numbers with a random number generator
Show the answer
Answer: D. Give every student a number and pick 20 numbers with a random number generator
Only the random number method gives every student an equal chance. The others are convenience samples.
A population is 60% blue marbles. Which sample size gives results most likely to be close to 60%?
- A random sample of 10 marbles
- A random sample of 5 marbles
- A random sample of 50 marbles
- All sample sizes give exactly 60%
Show the answer
Answer: C. A random sample of 50 marbles
Larger random samples vary less from sample to sample, so they are closer to the population percent more often.
A jar has 200 beads and 25% are red. A random sample of 20 beads has 7 red (35%). What does this show?
- The jar must have 35% red beads
- Random samples can differ from the population percent, but they tend to be fairly close
- The population percent changed
- The sample was not random
Show the answer
Answer: B. Random samples can differ from the population percent, but they tend to be fairly close
Sample percents vary from the true percent. A single small sample can be off.