# Lesson 13: Interpreting Data

## Overview

Students begin the lesson by looking at a cake preference survey that allows respondents to specify both a cake and an icing flavor. They discuss how knowing the relationship between cake and icing preference helps them better decide which combination to recommend. They are then introduced to cross tabulation, which allows them to graph relationships to different preferences. They use this technique to find relationships in a preference survey, then brainstorm the different types of problems that this process could help solve.

## Purpose

In the previous two lessons, students used data visualization to help them make decisions about a single variable (what locker to choose, which topping to order, etc.). In this lesson, they learn how to find relationships between variables using cross tabulation in the responses to different survey questions. Determining how answer choices relate to each other will allow them to make predictions about users based on previous responses. In the final project, they will use this same type of analysis to help them to design an algorithm for their recommendation generator.

## Agenda

### Warm Up (10 mins)

### Activity (40 mins)

### Wrap Up (5 mins)

### View on Code Studio

## Objectives

### Students will be able to:

- Visually organize data to highlight relationships and support a claim.
- Use cross tabulation to find patterns and relationships in data

## Links

**Heads Up!**Please make a copy of any documents you plan to share with students.

### For the Students

- Interpreting Data - Activity Guide
- Interpreting Data - Resource

# Teaching Guide

## Warm Up (10 mins)

**Display:** Show students the table on Level 2 of Code Studio.

Discussion Goal

There is no one recommendation that is correct, but make sure students understand that although chocolate was the most popular cake flavor and cream cheese was the most popular icing flavor, only one person chose a chocolate cake with cream cheese icing.

It's not enough to look at the two answers in isolation. For example, if two cakes are chosen, chocolate cake with chocolate icing and carrot cake with cream cheese icing is much better than chocolate with cream cheese and carrot with chocolate. Looking at the relationships between answers helps to see which choices go well together.

Remarks

Here are some more survey results, but this time instead of looking at pizza toppings, we're looking at cake and icing flavors.

**Prompt:** If you could choose one cake with icing, what would it be?

**Prompt:** What if you could choose two cakes with icing?

**Prompt:** How does knowing both choices together help us better understand what sort of cake we should order?

Sometimes it’s not enough to look at just one type of data. You need to look at how different types of data relate together. Today, we’re going to look at one way that we can find relationships in data to help us solve problems.

## Activity (40 mins)

Teaching Tip

Even though each of these sheets is only one page, it's best to print them separately, so that students can look at the survey results while they fill in the chart.

Classes who complete these activities online may want to use two computers per group, one to display the survey results and another to fill in the chart.

**Group:** Put students in groups of 2-3.

**Distribute:** Give each group a copy of Interpreting Data - Activity Guide and Interpreting Data - Resource.

Remarks

For our cake and icing example, there were only eight results, so we could look at the answers and get a good idea of the relationships between them. In this survey, we have a lot more results, so we’re going to use a chart to count them up.

Read the instructions as a class, then direct students to look at the first table on the activity guide.

Discussion Goal

Students should use the chart to find relationships between the preferences so that they can differentiate between subgroups. They may note that although people who chose cats tended to choose art, people who chose dogs tended to like music.

Make sure that the class produces examples of the predictions working in both directions (pet to activity and activity to pet). For example, people who like video games are likely to prefer dogs.

**Model:** Display the first chart on the board, and model how to fill in the chart with the class. For each row of the survey results, add one tally to the chart.

**Prompt:** If someone likes cats, what activity is probably their favorite?

Allow students time to write down their answer, then check with their group before sharing out.

**Prompt:** How would your answer change if I told you that the person likes dogs instead?

**Prompt:** What is one more interesting relationship between favorite pet and favorite activity?

Teaching Tip

Students may be tempted to think of reasons that different preference are related. Remind them that there is nothing in the survey that helps them to understand **why** a relationship is true, only that the relationship exists.

Direct students to complete the worksheet in their groups.

The next chart relates pets and sports. Students fill out the chart, then find two interesting relationships between pet preferences and sport preferences.

The students repeat the activity for activity and sport preferences.

Discussion Goal

Although there is no "right" answer to this question, students should realize that finding relationships between preferences may help them to predict one preference from knowing another. This could be for a recommendation engine, to place ads, or to promote particular social media posts.

Before moving on to the reflection question, give students a chance to share our anything interesting that they learned about the relationships between the different preferences.

**Prompt:** How could knowing relationships between these types of preferences help you to address a real world problem?

## Wrap Up (5 mins)

### Journaling

Teaching Tip

This journal prompt looks forward to the end of chapter project, when students will use this process to solve a data problem of their choosing.

**Prompt:**

- What’s another data problem you could think of that you could use this method to help solve?
- What questions would you ask?
- What relationships would you look for?

- Lesson Overview
- Teacher Overview
- Student Overview

- Interpreting Data - Exemplar (PDF)

- Cake Survey
- Student Overview

# Cake Survey

For this survey, people got to choose a cake flavor and an icing flavor.

Cake Flavor | Icing Flavor | |

1 | Chocolate Cake | Chocolate Icing |

2 | Red Velvet Cake | Cream Cheese Icing |

3 | Chocolate Cake | Chocolate Icing |

4 | Carrot Cake | Cream Cheese Icing |

5 | Carrot Cake | Vanilla Icing |

6 | Chocolate Cake | Chocolate Icing |

7 | Chocolate Cake | Cream Cheese Icing |

8 | Carrot Cake | Cream Cheese Icing |

If these people had to agree on one cake and icing combination for a party, what should it be?

What if they were able to order two cakes with icing?

How does knowing both choices help us better understand what sort of cakes we should order?