SPSS for newbies: questionnaire data entry- filter questions Video Lecture | SPSS: For Beginners - Data & Analytics

entering data for filter contingency
screening questions okay so this
continues on a series of questionnaire
data entry beginning with my first video
on questionnaire data entry and
continuing on with Likert scale input so
I'm going to assume that you've watched
both these videos right so filter
question otherwise known as screening
otherwise known as contingency or might
be known by other things these types of
questions look like this so during your
stay in hospital did you have an
operation or procedure yes or no notice
this is an nominal nominal scale if you
answer yes you go to question 42 if you
answer no you go to a question 49 so in
other words this questionnaire directs
you to different parts of the
questionnaire depending on how you
answer and this is so this type of
question is known as a filter question
then you might wish to use something
like this so you might want to wish
something like this use something like
this if you want to find out more about
a subset of view of the of the
respondents who answer a particular
question see the people go on to answer
42 questions that the answer will be
irrelevant to the person who answers
know on this okay
so a person answers yes here that they
did have an operational procedure then
42 would be relevant did a member of
staff explain the risks of the operation
no way you could understand but if you
didn't have an operation obviously
question 42 would not be relevant so
maybe you need something like this in
your questionnaire whether it been
marketing or psychology or anything
something something all right then we'll
show you how to set up your your SPSS
window to accept data like this all
right this is question 141 and this is
42
so let's do 41 first go to SPSS I go to
variable view I'll give it a name
well it's question 41 so let's just 41
decimal is going to set to 0
label all right the sentence of the
thing that of the full question so okay
during your state did did your operation
okay and the answer is yes or no one for
yes
add and two for no let me add and okay
I'm going to assume here they've got no
missing values they show you in the
other two videos how to deal with
missing values clearly this is a is a
nominal measure so we'll change it there
okay then we'll go into question 42
right and we'll set up in a similar way
I won't do 449 because it's a same
principle just keep this simple so 42
and the answer is yes completely
understand yes to some extent and no
alright so we will call it one two three
question 42 so one completely - meaning
to some extent three four no did not
understand and we haven't exactly
completed this yet but I'm just going to
come back to this later on I'll tell you
why
all right so we've pretty much set up
except for this the question 42 which is
not completely done yet but I'll come
back to this okay so let's say the first
person and said yes all right so in
which case question 40 to be relevant
did you how much of the rest of you
understand let's say completely so
that's one as well okay question and
yeah and second person say no so in this
case that's really relevant so we don't
answer anything that extra person guys
know of costume 41 goes to question 49
does not answer question 42 so let's
leave that blank for now third person
say yes let's say to some extent for per
say yes to some extent let's say first
they say yes non let's say question sir
person's success no so again that's
irrelevant all right okay so these dots
do I leave them blank it's best not to
leave them blank so what do I do
well see these blanks mean it's not
applicable to the person because they
answer the note on question 41 so we
should set up a n/a not applicable so
I'm going to choose it here to do it
here values I'm going to code it for and
it's a not clickable
okay four and four and if we don't like
to see the code want to see the actual
labels will click on this icon here one
two a there you go personally yes
completely understands no not applicable
turns question 4 to 2 and so on okay so
this is pretty much what it looks like
now how do I analyze something like this
well if we look at question 42 say I
just want to look at the frequency table
display frequencies say c6 valid
responses because we've got six
respondents here and a frequency so
completely understand one yes yes okay
that's easy to understand to some extent
right out 40 question 42 to people and
to some extent so that's two that's good
right no that was one person all right
and not applicable to people so that's
what it says that adds up to hundred
percent right that's one thing you could
do but suppose your so you would say
that you know out the respondents er
about 33% did not answer this question
42 because it is not applicable to them
and and so on but suppose you want to
make the sentence of the type out of the
people who had an operation in other
words out of the people answered
question 41 said yes to 41 bla bla bla
then what you want to do is you want to
change this by filtering out only the
people who said yes how do you do that
you go to select cases so you've go to
go to data and select cases down here or
quicker to use this icon right here
because it's to one select cases okay
and we'll click if and then question one
take that over equals one so I did this
beforehand so let me kind of cancel that
to show you how I do that so do you and
your stats question 41 take that over
press the equal sign one that means out
for the people answered yes to question
41 continue okay and watch what happens
you can see that there's a strikeout on
case two because that person answered
note of 41 and the strikeout on case six
that person and so no 241 so and then
we've got new conical filter which shows
us what has been selected what data has
been selected so basically all cases 1
to 6 except for 2 & 6 and now if we redo
the analysis analyze
descriptive stats frequency and just say
ok again look how the output changes so
now before it was six that was valid now
is only for it's valid because it struck
out two so now we can say our sentence
would be out of the people who had an
operation half 50% says here said they
understand stood to some extent the
risks who are undertaking alright so
that's how it changes your sentence
changed it to out of everyone - out of
the people who answered yes to question
41 and that's a that's bae that's
basically it but not some filter
questions and not juste so much by
undergrads I've ever seen but hey you
might be the first one in your college
to use it okay good luck