“What would you need for making a shoe fabric in your town?”. This is the first question I ask to my students when I start the statistics unit.
They answer several things: a plot for making the building, electricity, machines for making shoes, etc. But, and the end, I say “Yes. All of this stuff it’s needed, but what would you really need for switch on the machines?” and then, if there is no answer to that, then I put a clue “What do you do, first of all, when you go to shoe shop?”. Eventually, the answer becomes “The shoe number”.
And “how is the shoe number that you would manufacture first?”. And then obviously we have to determine what is the most frequent shoe number in our town. So we have to make a poll? But how to do that? I introduce the concept of poll, population and sample. Assuming that our classroom is non-biased sample, we list all the shoe number of my students1. This naturally introduces the frequency table (because we have a middle size sample) and the Mode.
This example of shoe number could be more useful. Depending on the course, then we represent the shoe numbers distribution and see what are the intervals of numbers with are the most frequent (in particular, the median).
This makes me introduce Statistics without having to be so magisterial and, with this, I realized that open questions are always good for introduce topics and engage students.
 Explore the MTBoS. A new exploration! 2015. https://exploremtbos.wordpress.com/2015/10/18/a-new-exploration/.
 Explore the MTBoS. Week 3 of the 2016 blogging initiative! 2016. https://exploremtbos.wordpress.com/2016/01/24/week-3-of-the-2016-blogging-initiative/.
and usually their parents and siblings too. If you would make a non-inmediate poll, you could send it as a homework.↩