You’ve given your students information about the Higher Apps project, showed them the SQA examples, and pointed them in the direction of the best sites for downloading data. But they’re struggling to come up with a research question. Maybe they’ve found a great dataset, but don’t really know how to begin. Here is some advice on getting students on the right track.
Start by introducing students to datasets which have questions with them, so they get a feel for the kinds of questions they can ask. For example, this dataset on animal research has accompanying questions that will lead students through completing a t-test. Chapter 9 of the Leckie textbook has lots of datasets with questions too.
Forming Questions as a Skill
Teach forming a research question as an explicit skill. These can make excellent lesson starters. For example, choose a dataset from Open Intro and task the students with writing as many different research questions as they can. This dataset on births gives a good example, students could ask:
- Is there a difference in the mean weight of newborn babies between the sexes?
- Do a higher proportion of mothers who smoke have premature births than non-smokers?
- Is there a correlation between the age of the fathers in the study and the age of the mothers?
They’re are countless other questions this one dataset could generate. Good follow-up questions include:
- What type of graphical displays would you expect to see for that research question?
- Which hypothesis test would you use?
- Which question of yours was the most interesting? Which wasn’t interesting?
That final point is important. I’ve found that students who are struggling are tempted to “pick two columns and see what happens.” You could look for correlation between the age of mothers and fathers, but is this interesting?
Of course finding a research question for their higher apps project will be easier if the students have found datasets that actually speak to them. A dataset on weather patterns may be interesting for students interested environmental science, but not as relevant for students studying business. Encourage students to think about how the data they’re working with relates to their own interests, and questions will come.