Inspired by Fisher’s Iris data, this sample of flowers was created through simulation from a carefully designed model. From a student’s perspective, these flowers represent a random sample of flowers from a much bigger population of statistics flowers. The idea is that students get all of the 300 cards and need to measure different features of the flowers and determine other variables to create their sample data.
Designed variables are: type of statistics flower (tictastics, stistactis, or castistist), petal colour (red, orange, blue, green), number of petals, petal length, petal width and stigma diameter. The diagram below shows how the measurements should be taken by students:
I have made the sample size 300 to allow for categorical and distributional exploration e.g. What proportion of all statistics flowers have a black stigma? Does stigma colour appear to be linked to petal colour for statistics flowers? How could the number of petals for statistics flowers be distributed? But I appreciate that it would take a long time for students to measure 300 different flowers and record necessary data! Perhaps students could look at the flowers visually first, sort them by type of flower and see if they can detect any features that appear to differ (e.g. colour, petal length, etc.). Students could then measure some of the flowers and chuck this data into a graph for an initial view before being given access to the digital sample to do some more exploring. Remember these data cards represent a sample and the true population parameters, for example the mean petal length of all statistics flowers, are unknown to you and the students. It is not intended that these cards are used for “population bags”.
Here is the sample data set as a CSV file: flower_power
Here are the data cards as a PDF: flower_power
You will need to print these one to a page if you want the measurements in the CSV file match!