SPSS for newbies Weight Cases button for grouped data Video Lecture - SPSS

hi everyone today I'm going to introduce
you to the wait cases button a button
that's incredibly useful but you find
that you might not see it in an
introductory course in SPSS to know why
it is useful let's present you with two
sets of data represent the same thing so
say I have some question and I've got
two categorical variables here gender
sex gender and choice choice that for a
particular kind of thing that I asked
them so for a case one the person who
completes a questionnaire was male chose
choice one okay second female chose
choice one as well
notice that some of these things repeat
themselves a male choosing choice one
appears more than once out of this list
of seven people I had two guys that
chose choice one look at the second
representation of this data set it's got
the same columns except for it's got one
additional columns and that one digital
columns I'm calling frequency just
counts the number of times that these
combinations happen so a male choosing
one I said appeared twice out of seven
so notice here male choosing choice one
frequency twice about a female choosing
choice one where I'm claiming it's three
so is it female choosing choice one
that's one that's two that's three so it
comes to re
so this is like the where each role
lists the list the individuals whereas
this one here groups the results so
we've got an additional column here and
you can see that if you had many many
many the sample size was large so had
many many many complete this survey then
you know your hands would be kind of
falling off typing all this stuff in
whereas this one is much shorter you
look at the combinations and you just
count them and put them into a separate
column this one we can call this
aggregated data you know I mean
aggregation meaning summing kind of and
or you could call it group data and then
so this aggregation is you'll see it
when you deal with categorical variables
it is continuous variables in here
like age is less likely you're going to
get exactly the same combination that
appears more months depending what you
continuous variable is so that you more
likely cease when your variables are all
categorical so I've pointed out the
difference between these two so what
well this one the group data aggregated
data is easier much faster to type in
and so you know if I had a choice
between the two you might prefer to use
facial analysis on this one of course
the respect of which table you use which
form you use you should get the same
results shouldn't affect the analysis
obviously just like this one is easy to
type in if we see another angle it might
be that you've just given data set
that's aggregated and that's so you have
to work it and you have no choice on the
matter so how do we do that well let's
take the unallocated data first say I
want to do some categorical analysis
based on these two say M crosstabs so
what you do is go
analyze descriptive statistics here I'm
not interested in actually doing the
analysis but one of them okay let's look
at the table here which simply just
tells us the number of individuals fall
into each of these categories so here
you got go male choice one two of them
female choice one three of them that's
the table
now let's replicate the same thing using
the aggregated the the aggregated data
here you go how do I do it well what not
to do is we do not go to we do not run
the analysis we don't go analyze analyze
crosstabs and do this look but let me
just do it so you don't do what I'm just
about to show you I'm showing you here
the reason for that is because it
doesn't take account of the number of
times each combination occurs so if you
only do male gender versus the choice
can you see it's just these numbers are
completely different it's not taking
account number of times that combination
is repeated so here we've got male
choice one is one female choice one is
one well that's not the case what we
have to do when we're using aggregated
data is to use a weights command that's
what this whole video is about to get
there we'll go to data weight cases by
default we do it's not doesn't weight
cases we want to weight cases by
frequency
I live call frequency which simply tells
it the number of times that a
combination appears the row combination
then I got okay and now switched on now
for any analysis I do whatever I do is
based on the weighted cases so let's now
do it analyze descriptive statistics
now lo and behold we can see that the
numbers now make sense so mail twice
female three times and that's it so we
don't for any other choice of things I
do if I go on to fit anything of
anything else or compute anything else I
don't have to keep going to date
wait cases they'll just remember that
all you have to remember is that if
there's any time that you want to switch
the thing off for some reason just make
sure you undo it so go to data
wait cases and then do not wait cases
they'll take you back to the original
and waited case this case okay so that's
it wait cases