Weekly domestic (US) box office revenues for the 8 Harry Potter films

Teachers Information

This activity is based on data sourced from University of Florida - Department of Statistics miscellaneous data page
Claire Laverty shared this link on the NZ Stats Teachers Facebook page.

I have imported the data into CODAP and have set up a CODAP document for students to work with/from.

If you or your students have never used CODAP before I recommend having a look at this CODAP document to get you started. Additional support on tables and graphs can be found here. I recommend you look at these before you launch into this activity with your students.

The document has the table of data, an initial graph ready for data and a text box outlining the description of the data. The data is US weekly domestic box office revenues for the 8 Harry Potter films.

Harry Potter on CODAP

Things to explore
  • the variables one at a time, by dragging and dropping them into the graph
    • what is the investigative question we are exploring each time, e.g. What is the distribution of the number of theaters that are showing the Harry Potter movies?
      • I wonder if the distribution of the number of theaters varies depending on the film number? (How might this one be answered?) - put theaters on x axis, films on y axis and then change films to categorical, you do this by clicking on the axis label film and from the dialogue box select treat as categorical, this gives dot plots for each of the 8 films, what would the box plots look like, click on the ruler and select box plots, this also allows for mean, median and other values to be added
      • I wonder if there is a relationship between the number of weeks and the number of theaters open, I suspect that the more weeks open the less theaters.
        • I wonder if this varies for the film number? drag the film number into the middle of the graph, this will give a varigated colour scheme, treating as numerical - to treat as categorical, click on film in the legend space and select treat as categorical
  • look for relationships between variables by dragging different variables to the x and y axis
    • remember to change from a number e.g. film number to categorical you click on the variable in the graph and select treat as categorical
  • make up new variables by creating formula
    • e.g. I wonder what is the total cumulative revenue for each film...
      • click on the ruler by the table, select new attribute in films
        • give attribute a name e.g. total_cumerev (no spaces)
        • type formula max() and then in the brackets type (cumerev) - you will notice that these start to pre-populate as you type
        • graph the data what does it show
    • e.g I get that the revenue decreases, but does the revenue per day per theater decrease at the same rate
      • something happens around week 7 since the release, there is a jump in the revenue per day per theater - why might that be? Explore release dates for Harry Potter films in the US, think about the timing of these and what is seven weeks later. What do you think is happening?

When finished students can save their results to a local file or to google docs. They can also share with you by selecting share on the "hamburger" menu. The three little lines at the top left.

I wonder if there is NZ data... so a bit of searching found Box Office Mojo and a New Zealand view

Lord of the Rings - series revenue data US domestic


I have also downloaded and adjusted the US domestic information for the Lord of the Rings movie series which can be explored. Check out Lord of the Rings on CODAP.