SPSS for newbies What do I do after I have entered my data Video Lecture

hi everyone well done for reaching the
end of my course and I hope you've
learned something and even more if
you've you've enjoyed bits of it or you
know you felt a sense of satisfaction
from learning something you know being a
first course in SPSS you shouldn't have
expected too much as I said I do this
course over an entire day and people
have lots of questions to ask me which
is why that takes a whole day you've not
been able to do that but hopefully this
udemy kind of platform you can ask
questions in the forum and I'll be happy
to answer them so long as they're around
about this yeah
not about your projects or something
like that which is not related to the
course material here so it's very
important to get a firm grounding in
data and the variables because more
complex methods later on you're going to
just we're going to automatically assume
you kind of know what type of variables
are out there and be very familiar with
normal handling of data okay so that's
why this has been very important the
next natural kind of question students
ask is I vented my data what do I do
next what kind of what kind of tools are
appropriate from for me you know for my
analysis well that really depends on
that answer depends on your hypotheses
so when you set out to do a project or
you know you're examining some data set
you've got some hypotheses in your mind
you know some statement you want to see
whether the data supports statements in
your mind so I'm going to do is write
down some some typical statements that
you might have you might have a
hypotheses and then point you to the
kind of right tool to use so knowing the
right tools depends on a your hypotheses
and B nature of the variables which you
now know scale nominal and ordinal
alright and students usually don't know
one or both of these questions so
obviously they have no idea what kind of
tools they can use and also they might
not not have learnt enough tools to know
that actually exist something exists for
their needs okay support let's do the
first one
there is a relationship between a and B
suppose you've got that as your
hypotheses know what a and B is
something well we've seen already in
this course that if a and B a
categorical categorical vs. categorical
we you can use the chi-squared test for
sociation if we've got something like
scale and scale to scale things we use
something called the Pearson's
correlation if I have something like
scale this is something that is ordinal
just so that's like yeah scale vs.
categorical ordinal we use something
called Spearman's Rho and that's
correlation as well if we have something
that scale versus something that's
nominal what we would do then is we need
to convert that scale interconnex
categorical so then it because garlic or
vs. categorical which can go back to use
the chi-square test an example suppose
the scale is age just scale you know any
old age so then we could kind of
categorize it by using the recode button
and forming groups for example we could
group age as up to 18 18 to 25 25 to 35
35 to 45 and over fortify for example
that's what I mean by converting scale
into categorical variable which you were
able to do that deals with the first
type of statement next common type of
statement is people want to assess
whether there's a difference but in
population means between a and B is the
difference between two groups that's
another way of saying it here we've got
the option of a t-test mann-whitney
wilcoxon why there are three is because
as I've told you before a hypothesis
test requires assumptions so this is
where we need to use test for some
assumptions for these tests an extension
to the difference in population means
across several groups right two or more
but usually we can do three or more
because if it's two then we can just use
one of these three here three more
groups we can use what is called ANOVA
method which stands for acronym for
analysis or there in
or can use a regression by the way here
we can also use regression but not
usually people don't bother okay so
that's we want to see if there's a
difference it means across several
groups in other words and though we're
saying you want to compare different
groups for one measure thing and the
thing that we're seeing is there's a
difference between the groups is scale
for example let's take an example from
education does the style of teaching
affect exams students exam score we
might so the thing being recorded here
that scale would be exam score and then
method of teaching would be the category
so here let's create three categories we
could have a teacher that use the method
of praising the class the opportunity
one that kind of punish the class and
one that is a control group so we can
compare then the teaching method on a
student's exam score if there is one
okay so that would be an example ANOVA
none over is very common in experiments
is actually built for where you've got
data from experiments so if you've
collected data from an experiment if
you're a psychologist if you've done
medical study that's more like it's
going to be an over and then the last
one is where you study cause and effect
type relationships that's to say that
you want to mix statement like oh you
know I've got something here that
depends on this and this and this factor
for example let's see let's take another
one from education that exam score
depends on time spent revising but want
under dependence time spent revising
could depend on other things it could
depend on then the type of subject is is
it a science of humanities it could
depend on your high school score how
well you did at high school and so on
okay
and regression is a very very flexible
tool so here the type variables it can
entertain all of them so you could have
a scale thing something which is scale
that depends on other factors which are
scale okay so my example scale thing my
scale is the exam score it depend on
could depend on something else which is
scale depend on yeah the time spent
revising you could have this thing that
is scale it could depend on mix of scale
and categorical things and that's the
first example actually gave you where
and finally you could even have
something which is categorical depending
on a mix of scale and categorical stuff
for example it could be like does
someone subscribe to a udemy video will
could depend on a whole number of things
the pricing of the course you could
depend on the type of a material is is
it you know something that falls in
social science or science and technology
education humor etc etc so edge
regression is very very powerful and you
know I've spent I run courses that last
entire year just on parts of regression
so regression is pretty much if you're
social scientist
probably you need regression if you do
an experiment type studies you can also
study it in the context of regression in
other words regression does a whole load
all right so now you've got the sense of
where you're going so what's next for me
what's next for me is if this is an
experiment uploading this course
depending how this goes the part two
will cover the difference in population
means your cover correlation it will
cover one way ANOVA so basic ANOVA and
it will go into an introductory level
until regression so it's a regression is
a is a very deep subject it's not just
to be covered in a ten minute video okay
so again well done and I hope you've
learnt something new and at least
got bearing on where you're heading next