Collecting data isn’t the most challenging part of Weatherblur or citizen science in general. In fact, with many of the student-driven investigations, you will end up with all kinds of data that your students collect as well as data from other schools and larger sources. In short – getting data is the easy part. The tough part is what comes next.

Helping students ask purposeful questions of their data is not only possible, but critical to helping them become critical consumers of data. The first step in helping your students is defining the categories of questions most frequently asked.

Check out this reference page, developed by Tuva.com, that outlines the six basic questions students can ask of their data.

Done? Okay – let’s see if we can dream up some of our own questions. Let’s say we have a dataset with the following attributes (characteristics):

- Can you come up with a question about
**distribution of points and variability within a group?***Reminder, distribution questions should include one categorical attribute and one quantitative measure.*

*Sample answer: What maximum daily temperatures during October are typical for Birmingham? This question has a categorical attribute of CITY (Birmingham) and one quantitative measure of Daily MAX TEMP.*

2. Can you come up with a question about **comparing two or more distributions**? *Reminder – include one categorical attribute and one quantitative measure.*

*Sample answer: How do maximum daily temperatures in the three cities compare?*

*This question has one categorical attribute (City, all three cities included) and one quantitative measure (Max Temp).*

3. Can you come up with a question about **covariation between two quantitative measures?** *Reminder – include TWO quantitative attributes.*

*Sample answer: What is the relationship between daily maximum temperatures and daily minimum temperatures? This question has two quantitative attributes: Daily Max Temp and Daily Min Temp.*

4. Can you come up with a question about **change through time (“time series”)**? Reminder – include TWO quantitative attributes.

*Sample answer: How did daily average temperatures change through the month in Sacramento? This question has two quantitative attributes, one of them (Date) is a time attribute, the other is Daily Avg. Temp.*

5. Can you come up with a question about **proportions within a group**? Reminder – include one categorical attribute.

*Sample answer: What proportion of the days during the month were clear, partly cloudy, mostly cloudy or overcast in each city? This question has one categorical attribute, Cloud Cover.*

6. Can you come up with a question about **spatial (geographic) distribution**?

*Sample answer: This is sort of a trick question (we know you are up for it!). You could ask a question such as Where are the three cities located? But this question could only be asked if Latitude and Longitude (both quantitative) for the cities were available and could be added to the dataset. The moral of the story? Don’t limit your students’ questions too quickly. There may just be a way to get that missing piece of information!*

**Phew!** Now you know so much more about the kinds of questions you can ask. Bravo! Now – the hard part. How do you teach your students about all of these different types of questions they can ask about data? At this point in their weatherblur journey, they are masters at asking SMART questions to help shape investigations. Now you need to help them ask purposeful questions of all their data.

Check out this lesson plan and accompanying student worksheet developed by Tuva. After introducing the vocabulary to students, pick a dataset from your investigation, one from the Tuva library, or quickly collect some data from the class itself, and see how your students can navigate this Mad-Lib style activity. This activity can be modified for a younger audience by altering the vocabulary or focusing on only a few questions that will be the most useful for your investigation.