Draw a Boxplot Using StatCrunch Video Lecture - Statcrunch Tutorials

so this data represents the 2009 crime
statistics by state we're going to draw
a box plot of the violent crime number
of violent crimes in each of the states
to do this we select graphics and then
we select box plot graphics then box
plot we will draw a box plot of violent
crime click Next and I personally like
to draw the boxes horizontally rather
than up and down vertically and we're
also going to use fences that we just
talked about to identify outliers they
will show up in the graph we will label
the x-axis violent crimes title at
violent crimes by state
click create graph and there your
violent crimes by state and so you can
see the outliers are marked by these
dots these dots and if you wanted to
know which states correspond to those
dots you can click your mouse button and
draw a square over the observation it
becomes magenta and then within the
actual data set that particular
observation won't light up that's
California with the most violent crimes
now do you think California is the most
violent crimes
which state is the most populous yeah
they're more people and so it's really
not fair to cast blame above California
having being violent because it's so
violent because of the sheer size of its
population so reporting number of
violent crimes is not a fair comparison
instead what should be reported is crime
rate number of crimes per population
okay and so we could actually create a
violent crime rate go to data compute
expression and the expression that we're
going to compute is violent crime
divided by population
I'm clicking add column as I do this hit
compute and there I get it this is a
number of crimes per population
I could have flip-flopped this and done
population per violent crime and say
number of individuals per violent crime
as well and now if I do graphics box
plot violent crime per population
draw horizontally use fences and this is
basically crime per capita is what we
call it per capita
defiling crimes per person basically
and now there's our data and what's
really neat if you look back you can
actually in this one you can see where
California lies now now they're not an
outlier at all interesting huh
so let me deselect that from my data set
and here I have an outlier so let's see
what that one is now go back here see
where it's highlighted in the data set
Washington DC notice that we're given
the data by region you see that or by
geographic division but let's just focus
on region one other thing that we could
do let me deselect it one other thing we
could do is draw box plots of violent
crime for population but we could group
by region so a separate box plots going
to be drawn for each region the Midwest
the south so on and so forth and so that
would allow us to do some sort of a
comparison within the different regions
of the country so I'll just click Next
and do these two and then title is going
to be violent crime by region right of
the country crime per capita is the axis
label and create the graph and so now
you got this for different regions west
south north east Midwest here's an
outlier in the in the South which is
still District to Columbia and another
interesting thing is well what do you
all these different things mean within
the box plot the reason it's called a
box plot is because of this box
sometimes this is called a box and
whisker plot the this observation right
here for the West would be the smallest
value within the data set so the min the
left edge of the box is the first
quartile q1 the middle bar in the box is
the median or q2 the rightmost edge of
the box is q3 the third quartile and
then this would be the largest
observation the max
and so you can see that for the West the
Northeast and the Midwest the median is
roughly the same right but the median
appears to be higher for the South
Region so more violent crimes per capita
in the south than in the other three
regions it would have seen from this
graphic