Working with StatCrunch: Univariate Statistics Video Lecture | Statcrunch Tutorials for Beginners - Data & Analytics

hi this is bart poulson and i'm going to
demonstrate how to compute a series of
univariate statistics in statcrunch
first thing you need to do is get to the
website and sign in I'm going to use a
data set that already exists on
statcrunch so I'm going to come up here
to Explorer and click on the triangle
and select data I know which data set
I'm going to use so I come over to
browse all and I'm going to use Survey
f08 dot XLS I've used it before so it
comes up as soon as I type in s I press
search and it's this data set with the
barber on it what I'm going to do is I'm
going to compute some statistics for
height haircut study work and salary I
do this by going to stat summary stats
and columns the reason I'm using columns
is because these organize that words one
column per variable four rows it's one
row per person and I don't want to do an
individual analysis correlation and
covariance could be used for other
analyses but that's not what we're gonna
do right now so I'm going to use stat
summary stats columns a dialog box comes
up and I can select the variables that I
want and select one and if I hold down a
controller and option key I can get some
others I'll put study and work and
salary and by the way the variables come
up in the order that I click on them in
a moment I can show that's important
this one where it would let me select
for instance just women or just people
who work less than 20 hours a week and
this one let me organize it for instance
compare women and men or those people
who
are in different sections in the data
set I'm gonna click on next so I can
choose the statistics that I want now
this set over here and mean variance and
so on is the default set but those
aren't really the ones that I want and
so what I'm gonna do is I'm gonna come
over to the left side where it lists the
statistics and I'm going to start by
clicking on n now this is one where you
can click on more than one statistic but
you need to do them by holding down the
control or option key and they come up
in the order that you select them and
for me that's important the first thing
I want is the sample size so I know how
many scores were dealing with then I get
the mean and the standard deviation
because those are two of the most common
statistical ago a bunch of other ones
that we use like correlation and
regression and t-test and they'll
severances so on however I also like to
describe the entire distribution with
what are called the five number summary
that's the minimum score or then the
first quartile the second quartile or
the median the third quartile and then
the fourth quartile or the maximum and
so I'm gonna click on those in order I
start with the minimum then I go to the
first quartile then to the median then
to the third quartile then to the
maximum so you see I had to click on
them out of order to get them in the
correct order over here it makes life a
little simpler afterwards then oh by the
way this one percent toss if I wanted
things like the 5th percentile and the
95th so I would have the range for the
middle 90% I would enter them right
there
I'm not going to do that now I press
calculate and I get a table
that looks like this the column is the
name of the variable I have height and
study and work and salary the n varies a
little bit 75 for 76 and 69 because not
everybody answered every question come
down here for instance and you see that
there's somebody here who didn't answer
height that's the mean it's giving it to
way too many decimal places given the
data I
would reduce it to two decimal places
and I do that not in statcrunch but when
i paste it into either excel or word
standard deviation again too many
decimal places and then we have the
minimum which i sometimes call q0 the
the lowest quartile q1 which is the 25th
percentile the median also known as q2
or the 50th percentile means half the
scores are below and half or above q3
three quarters are below 25% or above
and the maximum what I'm going to do
right now is I'm going to export my
results as univariate statistics for
four variables okay and that's that one
what I do next is I would paste it into
word and modify it I'll show that to you
in just a moment