Claims from Graphs: Correlation vs Causation
Evaluate evidence and avoid overclaiming from data
Claims from Graphs: Correlation vs Causation
Evaluate evidence and avoid overclaiming from data
Science - Grade 9-12
- 1
A scatter plot shows that students who sleep more hours per night tend to have higher biology test scores. A student claims, "Sleeping longer causes higher test scores." Is this claim fully supported by the graph? Explain.
Look for whether the graph shows an association only or whether it controls other variables.
The claim is not fully supported by the graph. The graph shows a positive correlation between hours of sleep and test scores, but it does not prove that sleep caused the higher scores because other variables, such as study time, health, or stress, could also affect scores. - 2
A graph shows that ice cream sales and number of swimmer rescues both increase during the summer months. Write a claim that is supported by the graph without confusing correlation with causation.
A supported claim is that ice cream sales and swimmer rescues both tend to increase during the summer months. The graph does not show that buying ice cream causes swimmer rescues. - 3
A researcher finds that neighborhoods with more trees have lower average summer air temperatures. Identify one possible causal explanation and one possible confounding variable.
A confounding variable is a third factor that may influence both variables being compared.
One possible causal explanation is that trees provide shade and release water vapor, which can cool nearby air. One possible confounding variable is neighborhood income, since wealthier neighborhoods may have more trees and different building materials or spacing that also affect temperature. - 4
A graph shows a strong negative correlation between the number of hours a phone battery has been used and the percent of charge remaining. Is a causal claim reasonable here? Explain why or why not.
A causal claim is reasonable here because there is a known mechanism: using a battery consumes stored electrical energy, which lowers the percent of charge remaining. The graph alone shows correlation, but background scientific knowledge supports a causal explanation. - 5
A study compares two groups of identical plants. One group receives fertilizer and the other does not. All plants receive the same light, water, soil type, and temperature. After four weeks, the fertilized plants are taller on average. What makes this evidence stronger for a causal claim than a simple scatter plot?
Think about controlled experiments and variables that are kept the same.
This evidence is stronger because the study includes a comparison group and controls several important variables. Since fertilizer is the main factor changed between the groups, the difference in plant height is more likely to have been caused by the fertilizer. - 6
A graph shows that countries with higher average chocolate consumption also have more Nobel Prize winners. What is one reason it would be inappropriate to claim that eating chocolate causes Nobel Prizes?
It would be inappropriate because the graph only shows a correlation between two country-level variables. Other factors, such as population size, education systems, research funding, and national wealth, could explain the pattern. - 7
A data table from a school is shown: Students who exercise 0-1 days per week have an average stress score of 8.1. Students who exercise 2-3 days per week have an average stress score of 6.4. Students who exercise 4-5 days per week have an average stress score of 5.2. Students who exercise 6-7 days per week have an average stress score of 5.0. What type of relationship is shown, and what causal claim should be avoided?
Describe the direction of the trend, then decide whether the study proves cause and effect.
The data show a negative correlation because stress scores tend to decrease as exercise days increase. The causal claim to avoid is that exercise definitely causes lower stress, because other factors such as sleep, social support, or free time may also be involved. - 8
A graph shows that as carbon dioxide concentration increases in a sealed plant growth chamber, the rate of photosynthesis increases at first and then levels off. The experiment controls light intensity, temperature, and water. What causal claim is supported?
The evidence supports the causal claim that increasing carbon dioxide concentration can increase the rate of photosynthesis up to a point under the tested conditions. The controlled variables make the causal claim stronger. - 9
A scatter plot shows no clear pattern between shoe size and score on a chemistry exam. What conclusion is best supported by the graph?
No visible trend usually means the variables are not strongly associated.
The best supported conclusion is that there is little or no correlation between shoe size and chemistry exam score in the data shown. The graph does not support a causal relationship between the two variables. - 10
A news headline says, "Energy drinks cause poor sleep in teenagers," based on a survey showing that teens who drink more energy drinks report fewer hours of sleep. Rewrite the headline so it is scientifically more accurate.
A more accurate headline is, "Survey finds that higher energy drink use is associated with fewer hours of sleep in teenagers." This wording describes a correlation without claiming that energy drinks are proven to cause poor sleep. - 11
A scientist wants to test whether a new water filter causes a decrease in bacteria in pond water. Describe an experiment that could provide stronger evidence for causation.
Include a control group, a treatment group, and variables that must be kept constant.
The scientist could collect equal samples of pond water with similar starting bacteria levels, filter half of them with the new filter, and leave the other half unfiltered as a control group. All samples should be kept under the same conditions, and bacteria levels should be measured before and after treatment. - 12
A graph shows that people who spend more time outdoors have higher vitamin D levels. Is it possible that time outdoors causes higher vitamin D? Is it proven by the graph alone? Explain both parts.
It is possible that time outdoors causes higher vitamin D because sunlight helps the body produce vitamin D. However, the graph alone does not prove causation because diet, supplements, sunscreen use, skin exposure, and season could also affect vitamin D levels. - 13
In a graph, city A has both high air pollution and high asthma emergency visits. City B has both low air pollution and low asthma emergency visits. What additional evidence would help determine whether air pollution contributes to asthma emergencies?
Look for repeated patterns, timing, mechanisms, and controls for other explanations.
Additional evidence could include data from many cities over time, measurements of specific pollutants, controls for smoking rates and healthcare access, and studies showing whether asthma visits increase after pollution spikes. Controlled or natural experiments would make the causal evidence stronger. - 14
A controlled lab experiment shows that increasing water temperature from 10 degrees Celsius to 30 degrees Celsius increases the reaction rate of an enzyme, but temperatures above 50 degrees Celsius decrease the rate sharply. What conclusion is supported?
The experiment supports the conclusion that temperature affects the enzyme reaction rate, increasing it over part of the tested range and decreasing it at high temperatures. Because the experiment controlled conditions and changed temperature, a causal claim is supported for the tested enzyme and conditions. - 15
A student says, "Whenever two variables have a strong correlation, one must cause the other." Write a response that corrects this misconception.
Mention at least two reasons a correlation might not be causal.
A strong correlation does not prove that one variable causes the other. The relationship could be caused by a third variable, coincidence, reverse causation, or a real causal link that needs more evidence. To support causation, scientists look for controlled experiments, timing, mechanisms, and alternative explanations.