HSS.IC.B.6Common CoreMathStatistics and ProbabilityGrades 9-12
HSS.IC.B.6: Evaluating Reports Based on Data
In plain English: HSS.IC.B.6 is the Common Core statistics standard that asks students to evaluate reports based on data, such as news stories, ads and polls. Students judge how the data were collected, whether the study design supports the conclusion, whether the numbers and graphs are presented fairly, and whether the claim goes beyond the evidence. It is usually taught in Algebra II or a statistics course.
Evaluate reports based on data.
Common Core State Standards for Mathematics · Domain: Making Inferences and Justifying Conclusions (IC) · Cluster: Make inferences and justify conclusions from sample surveys, experiments, and observational studies Also written as HSS-IC.B.6 or S-IC.6 · Official standard
Students learn to read a report based on data the way a statistician does: by asking how the data were produced before accepting what the report concludes. They use a short checklist that pulls together the rest of the cluster: who was studied and how they were chosen, what kind of study it was, how large the chance error is, how the variables were measured, whether the numbers and graphs are shown fairly, and whether the conclusion stays within what the data can support.
Students practice on invented news stories, polls, ads and graphs. They compute what a report should have shown (rates instead of counts, the true ratio behind a stretched bar chart, the interval behind a poll's margin of error) and write short evaluations that say what a reader can and cannot conclude. The goal is judgment, not cynicism: some reports are sound, and students should be able to say why.
Learning Objectives
By the end of this lesson, students will be able to:
Identify how the data in a report were collected and recognize sources of bias such as voluntary response, convenience samples, undercoverage, nonresponse and leading questions
Decide whether the study design (sample survey, experiment or observational study) supports the report's conclusion, especially a claim of cause and effect
Use margins of error to judge whether a reported difference is larger than chance variation
Detect misleading presentations, such as truncated axes, counts used in place of rates, and percent confused with percentage points, and compute the fair comparison
Write a short, balanced evaluation of a report that states what can and cannot be concluded
Prior Knowledge Required
Students should already be comfortable with:
Why a sample must be representative for a generalization to be valid 7.SP.A.1
The purposes of sample surveys, experiments and observational studies, and the role of randomization HSS.IC.B.3
Interpreting an estimate with a margin of error HSS.IC.B.4
The difference between correlation and causation HSS.ID.C.9
Show two headlines written from the same invented survey, in which 46% of teens said they usually skip breakfast:
Warm-Up Prompt
"Headline 1: 'Nearly Half of Teens Skip Breakfast.' Headline 2: 'Majority of Teens Eat Breakfast.' Can both headlines be true? Which would you choose if you ran a cereal company? What else would you want to know before you believed either one?"
Both are true: 46% skip and 54% eat breakfast. The framing changes the message, not the data. Students' questions for the last part usually include "How many teens?", "How were they chosen?" and "Who paid for the survey?" Write those on the board: they become the first items of the checklist.
Direct Instruction20-25 minutes
Introduce the checklist that students use for every report in the lesson. Each question connects to an earlier standard in the cluster.
Who was studied, and how were they chosen? Is the sample random, or did people volunteer (voluntary response), happen to be nearby (convenience), get left out (undercoverage) or not answer (nonresponse)? Is the population in the headline the one that was sampled?
What kind of study was it? Only a randomized experiment can support a cause-and-effect claim. An observational study can show an association, which may be explained by other variables.
How big is the chance error? Look for the sample size and margin of error. Is the reported difference larger than chance variation?
How were the variables measured? Check the wording of survey questions and whether answers were self-reported.
Are the numbers shown fairly? Check graph axes, counts versus rates, percent versus percentage points, and whether a mean hides a skewed distribution.
Does the conclusion follow? Does the claim stay within the population and the design? Who produced the report, and do they benefit from the conclusion?
Voluntary response sample
A radio station's website poll gets 2,400 responses, and 81% say the city should ban gas leaf blowers. The station reports: "City Wants Leaf Blower Ban."
Equation: Listeners chose to respond, and people with strong opinions respond more. The 81% describes only the volunteers; 2,400 responses do not fix the bias, so the claim about the city is not supported
Cause claimed from an observational study
In a survey of 500 teens, those who eat breakfast daily have a mean GPA of 3.4 and those who do not have 3.0. Headline: "Breakfast Raises Grades."
Equation: Difference 3.4 - 3.0 = 0.4 is an association only. Students chose their own habits, and variables such as sleep or family routines could explain both. An experiment would be needed for a cause claim
Truncated axis
A bar chart compares average commutes of 24 minutes (Eastview) and 27 minutes (Westfield) with the vertical axis starting at 22 (Diagram 1).
Equation: Bar heights are 24 - 22 = 2 and 27 - 22 = 5, so Westfield looks 2.5 times as long. Actually 27/24 = 1.125: only 12.5% longer
Difference within the margin of error
A poll of likely voters reports Candidate A 48% and Candidate B 45%, margin of error ±4 percentage points. Headline: "A Leads the Race" (Diagram 2).
Equation: A: 44% to 52%; B: 41% to 49%. The intervals overlap, and the 3-point lead is smaller than the margin of error, so the poll does not show who leads
Counts instead of rates
Town A (population 80,000) had 120 bicycle accidents last year and Town B (population 30,000) had 90. Headline: "Town A Is the Most Dangerous Place to Ride."
Equation: Rates: 120/80,000 = 1.5 per 1,000 residents and 90/30,000 = 3 per 1,000. Town B's rate is twice Town A's
After each example, ask students to name the checklist question that exposed the problem and to rewrite the headline so it is accurate, for example "Among 2,400 listeners who responded online, 81% favor a ban". Stress that the goal is not to reject every report: a randomly selected sample with a margin of error, a clear question and a claim that matches the design deserves to be trusted.
Guided Practice15-20 minutes
Evaluate one report together, question by question. Invented report: "After the city installed 40 new streetlights on Main Street, reported car break-ins there fell from 25 in 2024 to 15 in 2025, a 40% drop. Streetlights cut crime." First check the arithmetic: (25 - 15)/25 = 0.40, so the 40% is correct. Then work through the checklist: this is an observational before-and-after comparison with no comparison street; break-ins could have fallen citywide; counts of 25 and 15 can change a lot from year to year by chance; and other changes (more patrols, a new camera) could explain the drop. Pairs then propose a better design, such as comparing similar streets with and without new lights over several years, or randomly choosing which streets get lights first. Listen for students who reject the report because "the numbers are small" without explaining why small counts vary.
Independent Practice15 minutes
Students evaluate two reports on their own and write three to four sentences for each. (1) "A random sample of 50 seniors at Lincoln High found that 70% plan to attend college. Headline: 70% of U.S. Seniors Plan to Attend College." (The sample is random, but only from one school: the claim reaches far beyond the population sampled.) (2) "A store's sales rose from 1,200 to 1,500 units after a new display, shown on a chart whose axis starts at 1,100." (The increase is 300/1,200 = 25%, but the chart's bars have heights 100 and 400, so the second bar looks four times as tall. There is also no comparison with stores that kept the old display.)
Closure5-10 minutes
Exit ticket: Give one new short report and ask students to (1) rate how much they trust it on a scale from 1 (not at all) to 4 (fully), (2) give the two checklist questions that most affected their rating, and (3) write one sentence a reader can safely conclude from the report. Collect the tickets and share two well-reasoned ratings that disagree, to show that evaluation needs reasons, not only a verdict.
Differentiation Strategies
For Struggling Students
Give the six checklist questions as a table with a column for evidence from the report and a column for "problem or no problem"
Start with reports that have one clear flaw each before moving to reports with several
Provide sentence starters: "This report is based on ___, so it can / cannot show ___ because ___."
For Advanced Students
Ask students to find a real news story that cites a study, locate the original study summary and compare the headline with what the study actually found
Ask students to explain why the lead in a single poll should be compared with more than the stated margin of error, since both candidates' percentages vary from sample to sample
Ask students to write two honest headlines for the same data set that leave readers with opposite impressions, and explain which one is fairer
Assessment Guidance
What to Look For
Strong evaluations name the specific source of a problem ("voluntary response: people chose to answer") rather than a general complaint ("it is biased"), and they separate problems with the data from problems with the conclusion. Check that students compute the fair comparison when a report uses counts, stretched graphs or percent changes, and that they recognize a sound report when they see one. Watch for students who treat a large sample as a cure for a biased sampling method, and for students who reject every observational study instead of saying what it can show.
02
Classroom Activities
3 Activities
1
Report Stations
25 minGroups of 3-4
Groups rotate through four stations, each with one invented report printed on a card. At each station, the group fills in the checklist, gives the report a trust rating from 1 to 4 and writes an improved headline. One report is sound, so groups must judge rather than assume every report is flawed.
Station Reports
Station 1: A sports drink company says 30 runners who chose to drink its product ran 5% faster in a race than 30 runners who did not. (Observational, self-selected groups, and the company benefits from the result)
Station 2: A survey of 1,000 randomly selected adults in a state finds that 62% support a later school start time, margin of error ±3 percentage points. (Sound: random sample, margin of error given, claim limited to adults in the state)
Station 3: "Burglaries up 50% in Pine Hollow!" The village had 4 burglaries last year and 6 this year. (The 50% is correct, but counts this small change a lot by chance)
Station 4: Shoppers at a mall were asked, "Don't you agree the city needs a new parking garage at this mall?" and 78% said yes. (Convenience sample of mall users and a leading question)
Procedure
Spend 5 minutes at each station and record the checklist answers on the handout
Give the trust rating and write one improved headline
After the rotation, the class compares ratings for each station and discusses any station where groups disagreed by 2 or more points
Modification for Distance Learning
Put each report on its own slide in a shared deck. Breakout rooms spend 5 minutes per slide and type their checklist answers and rating in a text box before moving on.
2
Fix the Graph
20 minPairs
Pairs receive two misleading graphs made from invented data, measure what the graph suggests, compute what the data actually say, and redraw each graph honestly on graph paper.
Graphs
Graph 1: Monthly phone sales of 50,000 in March and 55,000 in April, drawn as bars on an axis that starts at 48,000. The bars have heights 2,000 and 7,000, so April looks 3.5 times March; the true ratio is 55,000/50,000 = 1.1
Graph 2: A line graph of a school's attendance rate that shows only the three weeks after a new policy, with no weeks from before. Pairs discuss what data are missing and sketch what a fair graph would need
Procedure
For each graph, write the impression it gives in one sentence
Compute the true comparison: a ratio, a percent change or a missing baseline
Redraw Graph 1 with the axis starting at 0 and list what extra data Graph 2 needs
Swap redrawn graphs with another pair and check each other's scales
Challenge Variation
Pairs make their own misleading graph from an honest data set the teacher provides, trade with another pair and challenge them to find the trick and compute the true comparison.
3
Evaluate a Real Report
25 minIndividual, then pairs
Each student brings, or receives from the teacher, a recent news article, ad or school newsletter item that uses data. Students write a one-page evaluation with the checklist and then trade with a partner for feedback.
Procedure
Summarize the report's claim in one sentence and underline the data it uses
Answer each checklist question, writing "not stated" when the report leaves something out
Compute at least one fair comparison, such as a rate, a percent change or an interval from a margin of error, when the report gives enough numbers
End with a verdict: what a reader can safely conclude, and what further information would change the verdict
Partners check that every criticism names a specific checklist question and that the verdict follows from the evidence
Discussion Questions
Which information was missing most often in the reports the class brought in?
Did any report hold up well under the checklist? What made it trustworthy?
How would you rewrite your report's headline so it matches the evidence?
03
Diagrams & Visual Aids
2 diagrams
Diagram 1: The Same Data with a Truncated and a Full Axis
Both charts show average commutes of 24 and 27 minutes, drawn to scale. On the left the axis starts at 22, so the bars are 2 and 5 units tall and Westfield's commute looks 2.5 times as long. On the right the axis starts at 0 and the bars show the true ratio, 27/24 = 1.125.
Diagram 2: Overlapping Poll Intervals
Each candidate's estimate is shown with its margin of error of 4 percentage points, drawn to scale. Candidate A's interval runs from 44% to 52% and Candidate B's from 41% to 49%. Because the intervals overlap from 44% to 49%, the 3-point lead is not convincing evidence that A is ahead.
04
Homework Assignment
~30 min
HSS.IC.B.6 Homework: Evaluating Reports Based on Data
Directions: For each report, use the checklist from class. Name each problem specifically (for example, "voluntary response" rather than "biased"), show any calculations, and finish with one sentence that states what a reader can safely conclude.
Part 1: How the Data Were Collected (Problems 1-3)
A magazine's website invites readers to vote on whether homework should be banned. Of 5,600 votes, 72% say yes. The magazine's headline reads "Most Americans Want Homework Banned." Evaluate the report.
Researchers followed 2,000 adults for 10 years. Those who owned a dog had a 24% lower rate of heart disease than those who did not. A news headline says, "Get a Dog to Protect Your Heart." (a) What kind of study is this? (b) Give two other variables that could explain the difference. (c) Could a study ever test the headline's claim? Describe the design and one practical problem with it.
A school mails a survey about the cafeteria to 800 randomly chosen families. Only 120 return it, and 90 of those rate the food "poor". The school newsletter reports that "three quarters of families think the food is poor." (a) Find the response rate and the percent of respondents who rated the food poor. (b) Evaluate the newsletter's claim.
Part 2: Numbers, Graphs and Conclusions (Problems 4-6)
A poll of 900 randomly selected likely voters finds 51% support for Measure 12, with a margin of error of 3.3 percentage points. A headline says, "Measure 12 Will Pass." Write the interval of plausible values and evaluate the headline.
A district newsletter shows a bar graph of graduation rates: 88% at North High and 91% at South High, with the vertical axis starting at 85%. (a) Find the heights of the two bars above the axis and the ratio they suggest. (b) Find the true ratio of the rates and the difference in percentage points. (c) Describe how to redraw the graph fairly.
(a) The share of students who bike to a school rose from 8% to 12%. One report says biking is "up 4%" and another says it is "up 50%". Explain how each report computed its number and write one accurate sentence. (b) City P had 330 bike thefts and has 150,000 residents; City Q had 210 bike thefts and has 60,000 residents. Which city has the higher theft rate? Show the rates per 1,000 residents.
Rubric
Criterion
Full Credit (2 pts)
Partial Credit (1 pt)
No Credit (0 pts)
Data Collection
Names the specific sampling or measurement problem and explains its likely effect
Names a problem without explaining its effect
No problem identified or wrong problem
Study Design and Conclusion
Decides correctly whether the design supports the claim, including cause and scope
Partly correct, such as noting association but not the population
Accepts or rejects the claim without reasons
Calculations
Correct rates, ratios, intervals and percent changes
One calculation error
Calculations missing or mostly wrong
Written Evaluation
Balanced verdict with an accurate rewritten conclusion
Verdict given but vague or one-sided
No verdict
05
Quiz: 20 Questions
Interactive, with answers
Instructions
Work through the questions in order. Your score updates as you answer, and Reset quiz clears everything so you or your students can try again.
Multiple choice: pick an option to check it. Short answer: write your answer, then reveal the model answer.
0 of 20 answered · 0 correct
Question 1 of 20 · Multiple Choice
A TV news show asks viewers to text YES or NO to "Should the state raise the gas tax?" Of 9,000 texts, 85% say NO. What is the main problem with using this result to describe the state's residents?
Answer: B
People who feel strongly, often against a tax, are the ones who text in, so the sample does not represent residents. Choice A is wrong: 9,000 is a large sample, but a large biased sample is still biased. Choice D rejects the number without a reason about how the data were collected.
Question 2 of 20 · Multiple Choice
A study of 3,000 adults finds that people who drink green tea every day live longer, on average, than people who do not. A headline says, "Green Tea Adds Years to Your Life." What is the best evaluation?
Answer: D
People chose whether to drink tea, so other lifestyle differences could explain the longer lives. Only a randomized experiment supports a cause claim. Choice A confuses sample size with study design: a large observational study still cannot show cause.
Question 3 of 20 · Multiple Choice
To estimate how many hours per week the adults in a town exercise, a researcher surveys people leaving the town's largest gym. What is the main flaw?
Answer: A
People at a gym are easy to reach but are not representative: they probably exercise more than other adults, so the estimate would be too high. Choice D would create a different, voluntary response problem. Choice B is wrong because a margin of error cannot correct a biased sample.
Question 4 of 20 · Multiple Choice
A survey asks, "Do you support the new park, which will give children a safe place to play?" and 88% say yes. What problem should a reader notice?
Answer: B
The phrase "a safe place for children to play" pushes respondents toward yes, so the 88% probably overstates support. A neutral question would simply ask whether they support the new park. Choice D misnames the study: asking questions is a survey.
Question 5 of 20 · Multiple Choice
A town of 18,000 people reported 45 cases of flu in a week. What is the rate per 1,000 residents?
Answer: D
45/18,000 = 0.0025, and 0.0025 × 1,000 = 2.5 cases per 1,000 residents. Choice C reports the count as if it were a rate. Choice B divides in the wrong order (18,000/45 = 400), and choice A misplaces the decimal point.
Question 6 of 20 · Multiple Choice
The percent of adults in a county who work from home rose from 30% to 36%. Which statement is accurate?
Answer: C
36 - 30 = 6 percentage points, and 6/30 = 0.20, a 20% increase relative to the starting value. Choice A mixes up the two ideas: the change is 6 points but 20% in relative terms. Choice B uses the relative change as if it were a difference in percentage points.
Question 7 of 20 · Multiple Choice
A poll reports Candidate X at 42% and Candidate Y at 37%, each with a margin of error of ±6 percentage points. A headline says, "X Has a Clear Lead." Which evaluation is correct?
Answer: B
Plausible values for X run from 36% to 48% and for Y from 31% to 43%. The 5-point lead is within the chance variation of the poll, so the race could be tied or Y could lead. Choice D goes too far: the poll still gives useful information about the range of support.
Question 8 of 20 · Multiple Choice
A bar graph shows values of 60 and 66, but its vertical axis starts at 57. About how many times as tall does the second bar look compared with the first?
Answer: A
The bars rise 60 - 57 = 3 units and 66 - 57 = 9 units above the axis, so the second bar looks 9/3 = 3 times as tall. The true ratio is 66/60 = 1.1, choice B, which is what an honest graph would show. Choice C is the difference 66 - 60, and choice D is the height of the taller bar alone.
Question 9 of 20 · Multiple Choice
A survey about internet use is conducted by calling only landline telephone numbers. What problem is most likely?
Answer: D
Anyone without a landline has no chance of being chosen, so part of the population is not covered, and those people may use the internet differently. Choice C does not apply: the researchers choose whom to call. Choice A may also happen, but it is not the problem built into the method.
Question 10 of 20 · Multiple Choice
A company mails a satisfaction survey to 2,000 randomly chosen customers and gets 150 replies. What is the response rate, and why does it matter?
Answer: A
150/2,000 = 0.075 = 7.5%. With so much nonresponse, the customers who replied chose themselves, and they may differ from those who did not. Choice C is a common error: the original random sample does not protect against bias created by who responds.
Question 11 of 20 · Multiple Choice
Which report gives the strongest evidence that a new reading program causes higher reading scores?
Answer: C
Random assignment makes the groups comparable, and a difference beyond chance can then be credited to the program. Choice A compares schools that chose for themselves, which may differ in resources. Choice D has no comparison group, so other changes that year could explain the rise.
Question 12 of 20 · Multiple Choice
An ad says, "3 out of 4 dentists surveyed recommend Brand Z." The fine print says 4 dentists were surveyed. What is the best evaluation?
Answer: B
3/4 = 75% is computed correctly, but with only 4 dentists, one different answer would change the result by 25 percentage points, and the dentists may have been chosen by the company. Choice D is wrong: 3/4 does equal 75%, and the problem is the evidence, not the arithmetic.
Question 13 of 20 · Multiple Choice
A small company reports that its 10 employees earn an average salary of $89,500. Nine employees earn $40,000 each and the owner earns $535,000. Which number better describes a typical employee's salary, and why?
Answer: C
The mean is (9 × 40,000 + 535,000)/10 = $89,500, but nine of the ten people earn $40,000. In a skewed data set, the median describes a typical value better. Choice D divides the owner's salary by 10, which is not a measure of center of all ten salaries.
Question 14 of 20 · Multiple Choice
A randomized experiment on 60 college soccer players found that a new stretching routine reduced muscle injuries. A fitness website reports, "Stretching Routine Prevents Injuries for Everyone." What is the main problem?
Answer: D
The experiment supports a cause-and-effect conclusion for players like those studied, but the website extends it to everyone. Older adults or non-athletes may respond differently. Choice A reverses the logic: randomized experiments are the design that can show cause.
Question 15 of 20 · Short Answer
A shoe company's press release says: "In our study, 40 runners who chose our new cushioned shoe reported 30% fewer knee aches than 40 runners who wore other shoes." List three problems with this report and suggest a better study.
(1) Self-selection: runners chose their shoes, so the groups may differ in training, weight or history of injury. (2) Self-reported response: runners who chose the new shoe may expect fewer aches. (3) Conflict of interest: the company benefits from a positive result. A better study would randomly assign runners to the new shoe or a standard shoe, measure knee pain in the same way for both groups and, if possible, not tell runners which shoe is the new model.
Question 16 of 20 · Short Answer
A school's average daily attendance rose from 90% to 93%. Find the change in percentage points and the relative percent increase, and write a sentence a newsletter could use accurately.
Change: 93 - 90 = 3 percentage points. Relative increase: 3/90 ≈ 3.3%. Accurate sentence: "Average daily attendance rose by 3 percentage points, from 90% to 93%."
Question 17 of 20 · Short Answer
Hospital R treated 2,500 patients and had 50 infections. Hospital S treated 800 patients and had 24 infections. A report says Hospital R has the worse infection problem. Evaluate the claim with rates.
Hospital R: 50/2,500 = 0.02 = 2% (20 per 1,000 patients). Hospital S: 24/800 = 0.03 = 3% (30 per 1,000). Hospital R has more infections only because it treated more patients; its rate is lower. A fair report would also check whether the hospitals treat similar patients, since sicker patients have more infections.
Question 18 of 20 · Short Answer
A poll finds that 58% of a city's residents support a new bus line, with a margin of error of 3 percentage points, from a random sample of 1,100 residents. Is it reasonable for a report to say "A majority of residents support the bus line"? Explain.
The interval is 58% ± 3%, or 55% to 61%. Every value in the interval is above 50%, so it is reasonable to say a majority of residents support the bus line, as long as the sample was random and the question was neutral. This is an example of a report that holds up under the checklist.
Question 19 of 20 · Short Answer
A report claims that students who take music lessons have higher math grades, based on a survey of 400 students. Describe how you would design a study that could test whether music lessons cause higher math grades.
Take a group of students who do not already take music lessons and randomly assign half to receive music lessons for a school year and half to another activity of similar time (so both groups spend the time). Measure math grades or test scores for both groups before and after in the same way. Compare the groups and use a simulation to decide whether the difference is larger than chance. Random assignment removes the self-selection problem: families who choose lessons may differ in other ways.
Question 20 of 20 · Short Answer
A company's sales doubled from one year to the next. Its ad shows a picture of a shopping bag for each year, and the second bag is twice as tall and twice as wide as the first. Why is the picture misleading? How many times larger does the second bag look?
Doubling both the height and the width multiplies the area by 2 × 2 = 4, so the second bag looks about four times as big even though sales only doubled. A fair picture would double only one dimension, or use bars of the same width with heights in the ratio 1 : 2.
0 of 20 answered · 0 correct
06
Frequently Asked Questions
10 Questions
What does HSS.IC.B.6 mean?
HSS.IC.B.6 means students can judge whether a report that uses data, such as a news story, poll, ad or research summary, supports its conclusion. Students check how the data were collected, what kind of study produced them, how large the chance error is, whether the numbers and graphs are fair and whether the claim stays within the evidence.
Is HSS.IC.B.6 taught in Algebra 1, Algebra 2 or Statistics?
It is usually taught in Algebra II, alongside the other standards in the cluster on inferences from sample surveys, experiments and observational studies, and in statistics courses. The Common Core places it in the high school Statistics and Probability category. Because it pulls together HSS.IC.B.3, HSS.IC.B.4 and HSS.IC.B.5, it works well as the final lesson of the unit.
What questions should students ask when they evaluate a data report?
Six questions cover most reports: Who was studied, and how were they chosen? What kind of study was it? How big is the chance error? How were the variables measured? Are the numbers and graphs shown fairly? Does the conclusion follow, and who produced the report? Students should answer "not stated" when the report leaves something out, because missing information is itself part of the evaluation.
Is a large sample always better than a small one?
A larger random sample is better, because it has a smaller margin of error. But size does not fix a biased method. An online poll with 50,000 volunteers can be less trustworthy than a random sample of 1,000, because the volunteers are not representative. Students should always ask how the sample was chosen before asking how big it is.
What is the difference between percent and percentage points?
A change in percentage points is a simple difference between two percents, and a percent change is that difference divided by the starting value. If support rises from 20% to 25%, it rose by 5 percentage points, which is a 25% increase (5/20). Reports sometimes use whichever number sounds more dramatic, so students should compute both.
Can an observational study ever support a claim of cause?
Not by itself. An observational study can show that two variables are associated, but other variables may explain the association. Researchers can strengthen the case by controlling for known confounding variables and by repeating the study in different groups, and some cause claims in medicine rest on many observational studies when experiments would be unethical. For this standard, students should say that a single observational study shows association and that a randomized experiment is needed to show cause.
What mistakes do students make when evaluating reports?
A common error is to call every report "biased" without saying why, or to reject all observational studies instead of saying what they can show. Many students also think a large sample removes bias, confuse random sampling (who is studied) with random assignment (who gets which treatment), and accept graphs without checking the axis. Asking students to name the specific checklist question behind each criticism addresses most of these problems.
Where can teachers find reports to use in class?
Local newspapers, school and city board reports, product advertisements and press releases from government statistical agencies are good sources, because they come from different kinds of producers. Pair a news article with the study or data release it describes when possible, so students can compare the headline with the original findings. Invented reports, like the ones in this lesson, are useful for isolating one flaw at a time.
How is HSS.IC.B.6 tested, and does it appear on the SAT?
Yes, in a related form. The digital SAT's Problem-Solving and Data Analysis domain includes evaluating statistical claims, for example deciding whether a study supports a conclusion about cause or about a larger population. State and classroom assessments often give a short report and ask students to identify a flaw, choose the conclusion the data support or explain how to improve the study.
How does HSS.IC.B.6 connect to other standards?
It uses the whole cluster: HSS.IC.B.3 for study types and randomization, HSS.IC.B.4 for margins of error and HSS.IC.B.5 for judging differences between treatments. It also draws on HSS.ID.C.9 (correlation is not causation) and on the data display standards, since many misleading reports rely on a poorly drawn graph. Outside math, the same skills support science lab reports and research writing in English and social studies.
07
Related Standards
6 standards
These standards connect to HSS.IC.B.6: prerequisites to review first, parallel standards at the same level, and next steps that build on it.
Before this lesson
7.SP.A.1Prerequisite
Understand that generalizations from a sample are valid only if it is representative
