To demonstrate how the results of a poll or other scientific study can be biased by selecting special types of people to respond or by asking only certain questions.
The main goal of early statistics with young students is to, “…make them informed consumers…of data.” (Benchmarks for Science Literacy, page 226.)
At the K-2 level, students will have had practice with sorting and classifying data. They will have done this with many tangible materials, such as rocks, buttons, shells, leaves, etc. Hopefully, students will have had opportunities to sort and classify tangible materials according to a number of different attributes and will have had some practice in communicating to others the groupings that they made.
For grades 3-5, this type of information gathering can be used for posing more challenging questions. Now students will begin to explore that sometimes, a representation of a group does not provide an accurate picture of the whole group. This can be a difficult concept for students to understand. Keeping things simple will help students begin to sort through the idea that data can be manipulated.
At this level, you can encourage students to look at data from a few different angles; identify the largest group in a field of data (a foundation for introducing the concept of “average”); talk about how knowing the largest group can be useful; and recognize that the largest group does not necessarily represent the whole group of data.
In this lesson, students will take a poll to gather information about homework time. With your guidance, students will then look at their information in ways that encourage them to think about the concepts listed above. This process will help them to consider the statement in the Benchmark for this lesson, “How much a portion of something can help to estimate what the whole is like depends on how the portion is chosen.”
To help students become engaged in the concept of bias sampling, you might take a quick poll of the class about something fun.
For example, tell the students that you would like to pretend that you are a newspaper reporter. You have come to this class to find out what foods the students like best. Choose three types of food (we’ll use enchiladas, pizza, and candied yams for this example) and ask students which food they like best. Count together to identify the food that received the most votes. (Let’s imagine it’s pizza.) Now talk with them about how you plan to write your article about how “Ms. Vargas’ class likes pizza best!”
Encourage students to think about how this article summarizes the class vote that you just took.
You might ask:
- Why did pizza make the headline?
- Did everyone vote for pizza?
- What happened with the votes for enchiladas and candied yams?
You can expect that students will be familiar with the “majority rules” idea. In the Development of this lesson, you can facilitate more critical thinking about how this majority came to be chosen.
Now it is the students’ turn to be reporters. They will conduct a survey about how much homework time is appropriate for them to have. Students will work in three groups to conduct the survey: one group will interview students, another will interview parents, and the other will interview teachers.
Before students begin their survey, lead a class discussion about how information is collected.
Consider these questions with your class:
- Why might it be important to interview students about homework time? Parents? Teachers?
- Do you think that each group will answer the questions the same way?
- Why not interview just one of the groups?
- What does interviewing three groups of people do?
- If you were the principal and were making the final decision about how much homework would be assigned to your class, should you only interview the teachers? Why or why not?
The goal of this lesson is to get students to think about bias as related to polls and studies. Help them think through how selecting only certain types of people to participate can bias a scientific study or survey.
It is also important to lead a discussion about the role questioning plays in gathering information. Choosing one set of questions over another affects the kind of information one can solicit from a poll or study. Also, challenge students to consider why it is meaningful to ask the same questions to respondents when taking a poll or doing a study. You might give students a quick example to help them think about how questions are linked to answers. (For example, why do you ask the same questions on tests?)
Other discussion questions could be:
- If you were working with a team of people and everyone went out to ask people questions about homework time, would you want everyone on your team to ask the same questions or different questions? Why?
- What would happen with your information if everyone asked different questions?
- What could you learn if everyone asked the same questions?
When students have grasped the importance of asking like questions to obtain specific information, introduce the Homework Survey student sheet. Students should take a class vote to determine which question will be asked for each item. You can encourage them to discuss the strengths and drawbacks of each proposed question. When they determine the final questions, they can write them on their student sheet, leaving space for recording participants’ responses.
Now have students work in three groups to interview students, parents, and teachers. Each student should complete her/his own Homework Survey student sheet so that everyone has an opportunity to ask questions and record answers.
When students have finished with their surveys, have them return to their small groups (the student, parent, or teacher group). Ask them to work together on the Report for the Principal student sheet.
Have each group read its letter to the principal to the class. When each group has had its turn, you can facilitate a discussion about these reports that helps students reflect on how the findings were different from group to group.
You could ask:
- What did you notice about these reports?
- What was similar?
- What was different?
- You each asked the same questions, but got different answers. Why?
- Each report made a recommendation to the principal. What would happen if the principal received only one of these reports?
- Should the principal find out how many students, parents, and teachers there are in the school and take only the survey information from the largest group?
- What could you do to make your group’s report to the principal reflect more opinions?
- How could you change your letter to more accurately represent the information you gathered?
Remember that the main idea you want them to consider is that information can be biased when only certain types of people are participating in a study or survey.
For another Science NetLinks lesson in which students analyze data, see What Can Data Tell Us? In this lesson, students look at questions that can be posed and answered by examining data distribution. They also look for circumstances that might bias the results of a study.
You can help students think about how bias sampling relates to other things like medical studies or trials. First talk with them about what is a medical trial, keeping it simple by describing it as an experiment that helps doctors know what medicines work the best.
- What would happen if only certain groups of people participated in the medical trial?
- If doctors need to know about a cold medicine, do you think they should ask people who do not have a cold to participate in the trial? Should they only ask people who have a cold?
- If doctors need to know about allergy medicines, do you think it makes a difference whether the participants in the trial have allergies or not?
- Do you think it makes a difference whether medical studies or trials are done only on children or adults?
- When do you think it would be important to conduct tests only on a specific group?
- Would it ever be unfair to exclude certain groups from medical tests or trials?
You can also encourage students to think about these issues as they relate to other things like the manufacturing of toys or kinds of cereals that companies make. Again, the idea is for them to begin considering how information can be biased and how this bias affects other things. This will help them become “informed consumers of data.”