Class Results for Fall 2014 

Eight people drew bikes, and before drawing them, indicated how
familiar they were with bicycles, and how well they understood how
bicycles operated.  On familiarity, the average rating was 2.125
(to remind you, 2 represented moderately familiar and 3 represented
very familiar).  On understanding, the average rating was 1.875
(where 2 represented understood moderately well how bicycles work
and 1 represented understood a little well how bicycles worked).

On the next class day, the drawings of the eight bicycles were 
presented and 13 people rated how well they thought the bicycles
would work if actually constructed.  The scale here went from -3
to +3 where -3 represented absolutely sure the bike would not work
and +3 represented absolutely sure the bike would work (again, the
modifiers for 2 and 1 were moderately sure and a little sure,
respectively).  Below, for each bike, are three measure of these
ratings.  The first is the mode, the second is the mean, and the third 
is the relative ranking.  Mode is the rating picked by most people;
mean is the average of everyone's ratings, and rank is a number from 1
to 8 based on the mean where 1 represents the bike with the highest mean
(the one people were most sure would work) and 8 represents the bike 
with the lowest mean (the one people were most surew would not work).

                 
Bike Drawing     Mode      Mean      Rank
    1             -3       -1.54       7
    2             -3       -0.85       4
    3             +1       +0.77       3
    4             -3       -1.23       5
    5             -2       -1.31       6
    6             -3       -1.85       8
    7             +3       +2.23       1
    8             +2       +1.15       2

 
So, there's a link on the web site where you can look at the bikes.

Look at each and look specifically for things that will make it not work
(excluding minor stuff like whether the wheels aren't perfectly round).
Is the seat missing, for example?  What about the chain or the pedals?
Is the chain in the right spot?  Will the pedals make the chain turn?
Are the wheels attached properly to the bike?  Can the front wheel be
turned?  Etc.

Try to relate these to the table values above.

And then, write me a paper about how these results are relevant to
whether imagery is accurate for an object you are quite familiar
with; why people say they know how something works when their
drawings of their images sometimes show things that can't work; 
and whether you think we would find different results if we gave people
drawings of a bike, instead of asking them to draw a bike (i.e., a
recognition test instead of a recall test where we can present drawings
in which, for example, the chain connects to the front wheel on some
bikes, to the rear wheel on others, or to both wheels on yet others).

When considering these issues, mention relevant experiments you learned
about, whether in the text or in lectures.