As part of a activity about measurement in class, I challenged my students to estimate .35 and .7 the width of a sheet of paper on their own without and measuring. We then brought the groups together and compared their estimates.
When I measured .35 and .7, I asked them if they thought all the .7 estimates were closer to .7 than .6 or .8. Most of them were pretty good. But when I asked them if any of their .35 estimates would round to .35, few were close enough.
With this, we talked about how we can only estimate one digit further than we have marks and still be pretty consistent and set up our discussion of sig figs.