Categorical data classifies objects into groups by a shared attribute (e.g., favorite color, type of pet). Students collect survey data, organize it in tally charts or frequency tables, and display it as bar graphs or picture graphs. They answer questions by reading and comparing categories in the data.
Conduct class surveys on age-appropriate topics and have students record responses in tally charts. Display data in multiple formats and compare what questions each format answers best.
You already know how to make tally charts — recording each item in a category with a mark and bundling every fifth mark as a diagonal line through four. Now you're putting that skill to work in a complete data process: deciding what question to ask, collecting the responses, organizing them, displaying them visually, and reading the display to answer questions. This full cycle is the foundation of statistics at any level.
Categorical data is data that puts things into named groups, not on a number scale. "What is your favorite fruit?" produces categorical data: apple, banana, mango, orange. (Compare this to "How tall are you?" — that's numerical.) For categorical data, you count how many responses fall into each category. The count for a category is called its frequency. A frequency table is just a organized list of categories and their frequencies, and a tally chart is the easiest way to build one as you collect data.
Once you have frequencies, you can display the data visually. A bar graph uses the length of bars to represent the count for each category. A picture graph uses symbols (one symbol = one response, or one symbol = some fixed number of responses). Both formats let you compare categories at a glance — which category has the most? the fewest? how many more does one have than another? These are the questions data displays are designed to answer quickly.
The most important habit in this process is matching the category label to the correct count. "Apple" is not a number — it's a label. The number 7 next to "apple" tells you how many people picked apple. When reading a graph, always check both the label and the scale before drawing any conclusion. Getting these two elements confused is the most common source of errors in data interpretation, at this level and beyond.