Académique Documents
Professionnel Documents
Culture Documents
1 | P a g e
Phenomenology ....................................................................................................................... 38
Other
Qualitative
Methods.................................................................................................. 38
QUANTITATIVE
METHODS ...........................................................................................................39
Pearson’s
Correlation
Coefficients .................................................................................. 39
Spearman’s
Rank
Correlation
Coefficient .................................................................... 39
Dependent
Samples
or
Matched
Pair
t-test................................................................. 40
Independent
Samples
t-test................................................................................................ 40
Mann-Whitney
U
test ............................................................................................................ 40
Chi
Square
Test ........................................................................................................................ 41
Simple
Linear
Regression .................................................................................................... 41
General
Linear
Regression.................................................................................................. 42
Multiple
Linear
Regression................................................................................................. 42
Multiple
General
Linear
Regression ............................................................................... 43
Analysis
of
Variance
(ANOVA) .......................................................................................... 44
Analysis
of
Covariance
(ANCOVA)................................................................................... 44
Multivariate
Analysis
of
Variance
(MANOVA) ........................................................... 45
Multivariate
Analysis
of
Covariance
(ANCOVA)........................................................ 45
Repeated
Measures
ANOVA................................................................................................ 46
Repeated
Measures
ANCOVA ............................................................................................. 46
Logistic
Regression ................................................................................................................ 47
Exploratory
Factor
Analysis .............................................................................................. 47
Reliability
Analysis ................................................................................................................. 48
CHAPTER
4:
RESULTS ............................................................................................... 48
FINDINGS .......................................................................................................................................48
CHAPTER
5:
CONCLUSIONS
AND
RECOMMENDATIONS................................. 49
CONCLUSIONS ................................................................................................................................50
IMPLICATIONS ...............................................................................................................................51
RECOMMENDATIONS ....................................................................................................................51
SUMMARY ......................................................................................................................................52
REFERENCES..................................................................................................................................52
2 | P a g e
Introduction
MCG
is
a
premiere
statistical
consulting
firm
that
specializes
in
dissertation
assistance.
Of
the
many
dissertation
services
we
offer,
a
few
include
the
development
of
research
concepts
and
ideas,
editing
and
formatting
documents
into
the
required
writing
style,
providing
suggestions
for
the
most
effective
methodologies,
and
comprehensive
analysis
of
collected
data.
Since
2006,
we
have
assisted
hundreds
of
doctoral
candidates
produce
top
quality
materials
to
be
submitted
to
their
respective
committee
review
boards.
MCG
is
the
dissertation
assistance
firm
of
choice
for
doctoral
candidates.
We
can
provide
you
with:
Expertise:
We
fully
understand
and
have
perfected
the
dissertation
processes
so
that
even
the
most
complicated
steps
are
simplified
into
direct
and
concise
procedures.
Our
team
of
editors
and
statisticians
includes
experts
familiar
with
a
variety
of
academic
areas.
Credentials:
We
are
highly
recommended
by
the
Association
for
Support
of
Graduate
Students,
and
we
are
members
in
good
standing
of
the
Statistical
Consulting
section
of
the
American
Statistical
Association.
Dedication:
We
are
highly
dedicated
to
our
task
of
ensuring
that
doctoral
candidates
end
up
with
the
best
possible
product
from
the
dissertation
process.
We
work
with
clients
on
a
continuing
basis,
assisting
with
any
revisions
required
until
the
job
is
done
and
the
3 | P a g e
4 | P a g e
function
as
a
step
by
step
map
to
guide
you
through
your
unique
dissertation
experience.
Writing
a
Dissertation
Dissertations
require
a
specific
chronological
process.
First,
a
student
is
required
to
write
and
submit
a
prospectus,
the
length
of
which
may
vary
by
school,
but
which
typically
ranges
from
between
5-‐15
pages.
The
prospectus
describes
the
overall
research
interest,
and
will
phrase
several
critical
questions
pertaining
to
the
research
along
with
ideas
of
how
the
student
will
seek
to
answer
them.
The
prospectus
is
then
submitted
to
the
dissertation
committee,
which
reviews
and
makes
comments
upon
it,
suggesting
areas
of
improvement
both
in
the
formulation
of
the
questions
and
the
means
by
which
they
may
be
answered.
It
is
during
the
finalization
of
the
prospectus
that
perhaps
the
most
critical
elements
of
the
work
are
formulated,
and
a
well
thought-‐out
prospectus
can
serve
to
make
the
remainder
of
the
process
much
more
efficient.
After
review,
potential
committee
members
will
commit
to
the
project
and
approve
the
prospectus.
Once
the
prospectus
is
approved,
the
student
can
begin
to
write
the
first
three
chapters
of
his
or
her
dissertation,
which
form
the
formal
dissertation
proposal.
Chapter
1
serves
as
an
introduction
to
the
study.
It
describes
the
nature
and
purpose
of
the
study,
presents
the
research
question(s),
defines
key
terms,
identifies
limitations
and
delimitations,
and
explains
the
significance
and
justification
for
the
study.
The
overarching
purpose
of
Chapter
1
is
to
5 | P a g e
provide
the
reader
with
a
basic
understanding
of
the
research
topic,
the
study’s
implications
and
importance,
and
where
it
will
eventually
lead.
It
is
also
here
that
preliminary
background
and
historical
context
is
provided.
In
addition,
the
student
examines
and
discusses
the
motivating
factors
that
led
him
or
her
to
arrive
at
the
question
of
interest.
Further,
the
overall
theoretical
approach
that
will
be
pursued
is
illustrated,
along
with
the
justifications
for
that
choice.
This
chapter
should
be
as
comprehensive
as
possible.
Next
to
the
abstract,
it
is
this
chapter
more
than
any
other
that
introduces
your
work
to
the
reader.
While
technical
in
nature,
it
is
important
to
remember
that
there
is
a
substantial
advantage
to
making
the
content
in
this
section
highly
engaging:
a
well-‐written
introduction
will
set
a
positive
tone
for
the
remainder
of
the
dissertation.
Chapter
2
contains
the
review
of
recent
relevant
literature
related
to
the
field
of
study.
The
main
purpose
of
the
literature
review
is
to
provide
a
solid
foundation
for
the
dissertation
based
on
past
literature
in
the
field
of
study.
This
should
provide
further
evidence
of
the
significance
and
justification
for
the
current
study
by
showing
how
it
will
contribute
to
the
field
in
a
meaningful
way.
It
should
also
explain
the
relevance
of
the
methods
used
by
relating
the
dissertation
to
other
studies
from
the
past.
At
the
end
of
the
second
chapter,
the
reader
should
have
a
clear
understanding
of
the
relevant
literature
from
government,
academic,
professional,
or
other
spheres
related
6 | P a g e
to
the
dissertation,
and
should
understand
clearly
how
the
study
will
fit
into
and
contribute
uniquely
to
the
existing
research
in
the
field
or
fields.
The
content
and
depth
of
a
literature
review
does
not
vary
substantially
from
school
to
school.
A
good
template
to
consider
as
an
example,
albeit
much
shorter,
is
the
background
section
in
a
peer
reviewed
paper.
Although
technically
these
appear
in
the
Introduction
section
of
such
a
paper,
they
often
have
less
in
common
with
the
introductory
segment
of
a
dissertation
than
with
the
literature
review.
This
is
due
mostly
to
the
limited
space
available
to
an
author
in
an
academic
journal.
Rather
than
apportion
space
to
a
comprehensive
background
section,
one
is
forced
instead
to
provide
a
dense
run-‐through
of
recent
and
relevant
literature,
with
appropriate
citations,
and
leave
it
to
the
reader
to
pursue
those
sources
and
obtain
a
broader
historical
and
theoretical
view,
if
so
desired.
This
is
perfectly
acceptable
in
such
a
context,
mainly
because
it
is
assumed
that
the
readers
will
almost
certainly
be
experts
in
the
field,
and
therefore
already
command
a
substantial
knowledge
regarding
the
essential
underpinnings.
Although
the
dissertation
is
certainly
going
to
be
read
by
experts
in
its
given
field,
it
is
important
to
remember
that
it
is
also
a
document
meant
for
the
consumption
by
a
wider
range
of
audience.
Your
university
or
college
will
make
it
available
to
the
public
in
its
archives,
with
the
expectation
that
it
will
be
viewed
by
an
inquiring
public.
It
is
therefore
best
to
think
of
this
chapter,
as
well
as
the
first,
as
the
7 | P a g e
opportunity
to
write
at
length
about
those
elements
of
your
work
that
would
not
be
possible
elsewhere.
Be
sure
to
provide
the
most
detailed
information
you
can,
and
demonstrate
above
all
else
that
you
possess
a
thorough
command
of
your
chosen
field.
Chapter
3
explains
the
methodology
used
to
carry
out
the
study.
It
should
provide
the
design,
procedures,
and
analysis
plan
needed
to
conduct
the
research.
At
the
end
of
Chapter
3,
the
reader
should
fully
understand
the
purpose
of
the
study
and
exactly
how
it
was
planned
and
eventually
carried
out.
Additionally,
through
the
literature
review
and
the
methodology
sections,
the
reader
should
completely
understand
the
relevance
and
the
reasoning
for
the
method
of
analysis.
The
most
common
subsections
employed
in
Chapter
3
are
an
introduction,
a
description
of
the
research
design,
the
population
and
sampling
procedures,
data
collection
instrumentation
and
procedures,
data
analysis,
and
data
interpretation.
More
than
any
other,
this
chapter
should
bear
the
most
in
common
with
the
content
of
an
academic
journal
article.
It
is
crucial
to
demonstrate
a
thorough
understanding
of
all
the
theoretical,
experimental,
and
analytical
methods
used
in
your
dissertation.
Researchers
will
often
utilize
certain
analytical
tools
without
fully
understanding
why;
they
will
simply
utilize
certain
statistical
tests,
for
example,
because
they
appear
frequently
in
the
literature,
and
give
the
impression
of
being
the
standard.
While
this
may
in
fact
be
the
case,
it
is
important
to
lay
out
the
explanations
as
to
why
this
is
the
8 | P a g e
case.
At
all
times
in
the
dissertation,
the
author
should
appear
in
sufficient
command
of
knowledge
such
that
he
or
she
would
be
able
to
respond
competently
to
any
question
put
to
them
about
it.
A
rule
of
thumb
that
will
be
emphasized
throughout
this
guide
is
that
the
author
should
have
full
and
complete
knowledge
about
any
piece
of
information
appearing
within
the
dissertation.
Traditionally,
once
the
first
three
chapters
are
completed,
they
must
be
submitted
to
the
dissertation
committee
for
approval.
Naturally,
the
specifics
of
this
requirement
will
vary
by
school,
but
in
almost
all
cases,
a
preliminary
approval
of
some
early
section
of
the
dissertation
is
required.
Another
approval
that
usually
must
be
obtained
is
through
the
Institutional
Review
Board
(IRB),
which
ensures
that
no
ethical
improprieties
will
be
involved
in
carrying
out
the
research.
Although
sometimes
routine,
IRB
approval
may
require
substantial
reworking
of
the
research
plan.
Typical
hurdles
that
might
cause
the
IRB
to
require
changes
include
disclosure
of
important
information
to
research
subjects,
conflicts
of
interest,
issues
of
monetary
remuneration
of
research
subjects,
bio-‐ethics,
and
animal
cruelty,
among
potential
others.
Once
approval
has
been
received
and
all
potential
issues
resolved,
the
author
then
moves
onto
Chapters
4
and
5.
Chapter
4
contains
the
results
of
the
data
collection
and
analysis
described
in
the
methodology
section.
The
presentation
of
the
data
and
the
data
analysis
must
be
entirely
factual
and
be
presented
in
APA
format
and
9 | P a g e
style.
The
results
should
be
presented
in
the
same
order
that
the
research
questions
were
asked
in
Chapter
1.
This
creates
a
consistent
and
cohesive
feel
for
your
study
and
helps
avoid
confusion
among
your
readers.
Each
research
question
should
have
its
own
subsection
in
this
chapter.
In
each
of
these
sections,
the
research
question
should
be
restated.
Any
charts,
equations,
or
graphs
associated
with
that
question
should
be
shown
here.
The
importance
of
well-‐placed,
well-‐designed
graphs
and
charts
cannot
be
overemphasized.
A
logically,
efficiently
constructed
graph
or
chart
has
the
ability
to
convey
enormous
amounts
of
information
in
a
highly
compact,
effective
way.
In
keeping
with
the
requirements
of
the
APA
format,
any
graphic
should
have
associated
with
it
an
explanatory
caption
that
delivers
the
full
message
of
what
is
being
conveyed
in
that
graphic.
A
worthwhile
rule
of
thumb
for
an
author
is
to
construct
any
chart
or
graph
such
that
the
entire
experiment,
from
start
to
finish,
including
the
method
and
rationality
behind
the
data
collection,
can
be
inferred
from
reading
the
entire
graphic,
associated
text
included.
The
act
of
condensing
a
large
amount
of
information
into
a
small
space
can
be
challenging.
Once
again,
a
good
example
of
an
efficient
graphic
can
be
obtained
by
looking
at
a
published
article
in
a
journal
with
rigidly
enforced
length
requirements.
The
figures
below
provide
concrete
examples
of
efficient
graphics,
and
are
modeled
after
those
that
appear
in
the
international
journal
Science,
which
publishes
work
of
great
sophistication
and
depth
while
requiring
authors
to
adhere
to
very
strict
space
limits.
More
than
just
efficiently
conveying
information,
constructing
graphics
is
10 | P a g e
another
excellent
opportunity
for
the
author
to
demonstrate
his
or
her
knowledge
of
the
subject
matter.
Without
a
coherent
understanding
of
the
material,
putting
together
an
excellent
graphic
is
extremely
difficult.
Indeed,
even
for
experienced
authors
with
many
years
of
experience
in
a
field
it
can
still
be
a
complicated
task.
The
integration
of
large
amounts
of
information
into
a
small
space
is
never
simple;
and
yet
it
is
one
of
the
most
important
elements
of
any
successful
research
write-‐up.
It
is
in
this
area
especially
that
help
from
expert
consultants
with
familiarity
in
analysis
can
be
indispensible.
Their
comfort
in
working
with
publishable-‐
quality
graphics
can
dramatically
improve
the
overall
structure
of
a
dissertation.
Chapter
5
is
intended
to
explain
the
findings
of
the
study
and
their
significance
and
conclude
the
dissertation
by
suggesting
future
research.
The
chapter
should
begin
by
summarizing
the
information
from
Chapters
1-‐4
to
clarify
the
purpose
of
the
study
again.
Then
it
must
discuss
what
can
be
concluded
from
the
study’s
findings
and
how
these
conclusions
align
with
the
research
question(s)
and
hypotheses.
Finally,
the
chapter
should
discuss
the
implications
of
the
findings
and
the
limitations
of
the
study,
followed
by
recommendations
for
future
research.
11 | P a g e
section
will
explore
each
aspect
in
detail
and
will
provide
you
with
some
essential
insight
into
how
you
can
convey
your
research
objectively
and
professionally.
Let’s
get
to
work
with
some
basics!
Chapter
1
is
the
Introduction
chapter
that
establishes
the
framework
for
the
entire
study.
An
effective
first
chapter
should
start
off
with
a
brief
introduction
to
the
project
that
will
capture
the
reader’s
attention
and
interest
by
presenting
a
problem
that
illustrates
the
importance
of
the
current
research.
This
is
called
the
“problem
statement.”
An
opening
paragraph
should
tell
a
brief
“story,”
supported
by
references,
about
why
the
research
is
relevant
and
needed
in
a
contemporary
context,
yet
it
should
leave
the
reader
wanting
to
know
more.
After
the
introduction,
the
reader
should
be
left
hanging,
excited
to
see
where
the
study
eventually
goes.
It
is
helpful
to
bring
up
current
and
past
issues
related
to
the
research
subject.
Presenting
current
and
past
issues
will
help
the
reader
to
identify
with
the
researcher’s
personal
passions
and
reasons
for
conducting
the
research.
It
is
not
necessary
to
fully
explain
certain
concepts
or
events
here,
because
they
will
be
discussed
in
detail
in
the
Chapter
2
Literature
Review.
In
fact,
it
is
best
to
leave
the
reader
wanting
more,
so
that
they
are
excited
to
read
on.
Remember
that
it
is
important
to
capture
the
reader
in
the
opening
paragraphs,
so
some
finesse
can
be
used
to
grasp
their
attention.
However,
it
12 | P a g e
is
important
to
note
that
the
writing
style
for
dissertations
should
be
absolutely
objective
in
nature.
Personal
opinions,
biases,
and
persuasions
should
be
completely
absent
from
the
content.
By
inserting
these
elements
into
the
writing
style,
the
dissertation’s
strength
and
credibility
can
be
greatly
diminished.
The
opening
paragraphs
should
also
include
a
brief
description
of
what
Chapter
1
contains.
Over
the
course
of
a
few
sentences,
state
the
different
sections
that
will
be
included
in
Chapter
1.
The
section
titles
are
self-‐explanatory,
so
it
is
not
necessary
to
go
into
great
detail
about
why
the
sections
are
included
or
what
information
will
be
included
within.
The
last
sentence
of
this
opening
section
summarizes
the
overall
purpose
of
chapter
one,
which
is
to
provide
a
framework
for
the
entire
dissertation.
These
opening
remarks
should
be
kept
to
less
than
one
page.
13 | P a g e
14 | P a g e
conducted.
The
section
begins
with
the
purpose
statement,
which
in
one
or
two
succinct
sentences
declares
the
current
study’s
research
methods,
problem
to
be
examined,
audience
to
which
the
problem
is
significant,
and
setting.
These
sentences
should
be
declarative
and
provide
the
basis
for
which
the
entire
study
will
emanate.
After
these
statements,
it
might
be
necessary
to
provide
clarification
about
the
four
elements.
To
provide
additional
information
about
why
a
given
research
method
is
used,
establish
the
type
of
data
and
information
to
be
collected.
State
the
most
effective
approach
in
collecting
the
data
and
give
reasons
why.
For
example,
conducting
an
interview
or
focus
group
might
be
the
best
way
to
collect
needed
data,
hence
a
qualitative
research
method
should
be
used.
Alternatively,
a
quantitative
study
is
best
conducted
through
surveys
and
harvesting
existing
information.
However,
details
about
the
research
methodology
should
be
kept
to
the
“Nature
of
the
Study”
section,
detailed
below.
Additional
information
about
the
problem
to
be
examined
should
explain
the
nuanced
characteristics
and
unfamiliar
concepts
utilized
in
the
current
study.
It
is
helpful
to
describe
the
implications
that
these
items
may
have
on
the
study
as
well
as
any
interesting
and/or
novel
approaches
to
the
problem
that
the
study
uses.
Identify
the
unique
characteristics
about
the
study’s
pertinent
audience.
15 | P a g e
The
study’s
pertinent
audience
can
be
defined
as
the
people,
places,
or
things
that
the
research
examines
within
the
context
of
the
problem.
For
example,
when
examining
a
sample
of
psychiatric
patients,
it
is
helpful
to
explain
the
differentiating
characteristics
that
make
them
appropriate
for
investigation
in
the
current
study.
Last,
provide
more
detailed
information
about
the
setting
of
the
study.
Give
reasons
why
this
setting
was
chosen
and
link
this
to
the
research
problem.
For
example,
explain
the
implications
that
additional
information
regarding
geographical
location,
industry,
and
culture
may
have
on
the
current
study.
Clarifying
and
providing
additional
information
about
the
four
required
elements
of
a
declarative
purpose
statement
will
provide
the
reader
with
the
important
specifics
about
the
study
for
the
first
time
in
the
dissertation.
The
reader
should
now
understand
the
reason
why
the
study
is
being
conducted
within
the
context
of
the
specific
research
field.
In
general,
the
purpose
statement
should
identify
what
is
being
studied,
who
is
included,
where
the
study
takes
place,
and
how
it
is
being
conducted.
16 | P a g e
that
could
be
recognized
on
a
global
scale.
It
is
important
to
emphasize
the
far-‐reaching
influence
that
the
study
may
have,
both
within
and
outside
the
research
field.
Explain
why
the
study
is
unique.
Identify
who
will
benefit
from
the
outcomes.
This
could
range
from
a
small
group
of
other
researchers
to
the
entire
human
population.
Start
by
noting
the
significance
of
the
study
on
smaller
scales,
and
then
zoom
out
to
more
general
applications.
Clearly
identify
the
contribution
the
study
makes
to
the
research
field,
and
then
explain
how
developments
in
this
field
can
benefit
the
rest
of
the
world.
From
the
significance
statement,
the
reader
should
understand
the
potential
implications
of
the
study’s
outcomes
and
why
it
is
important.
17 | P a g e
Research
Questions
This
section
officially
states
the
study’s
research
questions
and
should
pique
the
reader’s
interest
to
learn
more
about
the
answers
to
those
questions.
Start
by
restating
the
purpose
and
goals
of
the
study
to
clearly
link
these
items
to
the
research
questions.
Know
that
the
eventual
answer
to
your
research
questions
must
align
with
the
purpose
and
goals
of
the
study
in
order
to
be
successful.
Clearly
list
the
relevant
research
questions
to
direct
the
study
in
the
proper
direction.
There
are
not
any
rules
that
stipulate
how
many
research
questions
a
dissertation
can
have.
Having
only
one
research
question
can
be
acceptable,
but
make
sure
it
is
not
too
vague
or
general.
Alternatively,
having
too
many
research
questions
can
reduce
the
clarity
of
your
final
product
and
therefore
the
significance
of
your
findings.
Depending
on
the
study,
there
should
typically
be
around
2-‐5
evocative
research
questions.
18 | P a g e
A
well-‐formulated
research
question
will
draw
your
readers
in
and
convince
them
that
your
study
is
important
and
applicable.
They
should
be
left
wondering
what
sort
of
findings
you
may
have,
and
how
they
may
change
the
landscape
of
your
specific
area
of
study.
The
theoretical
framework
of
your
dissertation
will
put
your
study
in
perspective
relative
to
other
studies.
It
will
relate
your
study
to
the
many
competing
concepts
and
theories
within
your
discipline.
Assumptions
Make
sure
to
understand
completely,
and
be
aware,
of
all
the
assumptions
underlying
the
fabric
of
your
research
questions.
Shaky
assumptions
can
completely
undercut
your
entire
study.
Provide
a
short
list
of
all
of
your
assumptions
and
the
arguments
that
can
be
brought
against
them.
Then
put
together
a
collection
of
counterarguments
that
support
your
topic
of
study
and
research
approach.
Assumptions
are
required
in
all
studies,
and
in
order
to
control
researcher
bias,
it
is
important
that
the
assumptions
of
the
study
are
posited
for
reasons
of
transparency.
19 | P a g e
20 | P a g e
data
sets
and
data
collection
techniques.
Validity
and
reliability
go
to
the
very
heart
of
the
research
study.
Furthermore,
validity
itself
can
be
subdivided
into
separate
categories
depending
on
the
source
of
information
and/or
nature
of
a
given
research
question.
The
concept
of
internal
validity
is
dealt
with
in
the
following
section
on
Limitations,
and
external
validity
under
the
section
entitled
Delimitations.
Limitations
Limitations
are
related
to
the
concept
of
internal
validity.
This
refers
to
the
believability
of
your
findings
and
results.
Naturally,
therefore,
the
robustness
of
a
study’s
internal
validity
is
directly
related
to
the
logical
strength
of
its
questions.
Questions
and
research
formats
that
leave
too
much
room
for
subjectivity,
vagueness,
or
leading
elements
can
all
negatively
impact
the
internal
validity
of
your
study.
There
are
a
number
of
hard
and
fast
rules
that
are
often
employed
to
address
the
limitations
and
reinforce
the
internal
validity
of
your
study.
In
other
words,
what
can
you
do
to
avoid
the
questions,
“How
do
you
know
things
happened
for
the
reasons
you
gave?”
or
“Could
some
of
my
questions
have
led
to
the
subject
to
a
particular
answer,
without
realizing
it?”
As
addressed
in
the
following
section,
it
is
an
extremely
difficult,
if
not
impossible,
task
to
eliminate
all
“contaminating”
factors
from
your
questions’
content.
There
will
always
be
some
subjective
element
in
their
phraseology,
or
in
the
mindset
of
your
responding
subject,
that
will
introduce
error
and
21 | P a g e
contaminants
into
your
results.
The
means
by
which
these
are
dealt
with
are
statistical
in
nature.
Specifically,
you
should
seek
to
randomize
as
wide
a
variety
of
these
contaminating
factors
or
variables.
If
there
is
a
question
that
might
be
answered
differently
by
respondents
living
in
different
geographic
areas,
for
example,
then
it
makes
sense
that
you
would
attempt
to
sample
the
largest
geographical
area
possible
in
order
to
remove
the
bias
induced
by
location.
Delimitations
Delimitations
are
related
to
the
concept
of
external
validity.
This
relates
to
the
aspect
of
your
study
that
may
be
affected
by
issues
of
with
or
to
whom
you
are
administering
a
test,
where
the
test
is
run,
and
when
it
is
run.
All
these
factors
can
alter
the
answers
received
from
subjects,
and
thus
the
uncertainty
introduced
by
them
must
be
taken
into
account
in
the
analysis.
Essentially,
this
deals
with
the
practical
impossibility
of
constructing
a
test
that
is
universally
generalizable.
In
discussing
the
threats
to
external
validity,
you
should
be
sure
to
include
all
possible
examples,
and
especially
those
that
offer
the
greatest
potential
of
damage
to
your
conclusions.
For
example,
if
you
were
studying
the
perception
of
access
to
higher
education,
and
were
questioning
students
from
a
wide
range
of
socioeconomic
backgrounds,
it
would
stand
to
reason
that
you
would
receive
strongly
different
replies.
Often,
isolating
all
the
delimiting
factors
means
confronting
sensitive
issues:
economic
status,
race,
sexual
orientation,
religious
and
political
beliefs,
etc.
These
are
the
most
22 | P a g e
difficult
to
deal
with
because
they
usually
influence
the
backgrounds,
thoughts,
and
ideas
of
the
question
subjects.
Documentation
In
Chapter
2,
you
document
the
body
of
work
concerning
your
research.
This
includes
giving
a
thorough
review
of
your
predecessors
and
contemporaries’
works,
the
most
current
analysis
and
professional
opinion,
and
providing
accurate
sourcing.
It
is
also
the
place
to
make
the
reader
familiar
with
the
types
of
sources
is
typically
used
in
your
field.
This
means
understanding
the
assumptions,
statistical
methods,
and
research
constructions
used
in
the
larger
body
of
literature.
An
essential
part
of
the
dissertation
is
writing
comprehensively,
and
with
great
understanding,
about
the
research
field
and
showing
how
your
work
directly
ties
into
it.
This
also
helps
to
show
why
your
work
is
significant.
By
placing
it
in
a
broader
historical
context,
you
can
show
your
study
to
be
an
essential
step
in
the
process
towards
answering
a
question,
and
not
just
a
“report”
on
what
has
been,
and
is
being
done.
The
following
section
describes
the
process
of
the
Literature
Review,
in
which
this
detail
is
presented
in
the
dissertation.
23 | P a g e
Literature
Review
The
literature
review
is
intended
to
provide
a
complete
background
of
all
the
pertinent
publications
regarding
the
research
subject.
This
will
provide
historical
context
and
a
completely
updated
status
of
the
field
as
it
currently
stands.
Generally,
the
review
of
the
literature
should
be
presented
in
chronological
fashion.
In
making
the
literature
review
completely
comprehensive,
a
good
amount
of
detail
needs
to
be
provided
about
specific
research
studies.
The
specific
amount
of
detail
to
include
is
not
fixed,
but
as
a
good
rule
of
thumb,
consider
writing
a
half-‐
to
page-‐long
mini-‐review
of
each
article
included
in
the
review.
For
longer
articles,
a
longer
mini-‐review
will
probably
be
needed.
In
these
minis,
make
sure
you
highlight
all
the
critical
assumptions,
hypotheses,
analytical
methods,
and
conclusions
drawn.
You
will
want
to
include
all
this
information
in
the
review
section.
In
other
words,
it
should
be
possible
for
a
reader
without
any
prior
knowledge
of
the
specific
research
included
in
the
overall
review,
to
come
away
from
this
section
with
a
full
understanding.
In
some
sense,
this
section
should
read
as
an
historical
story.
Explain
how
your
work
fits
into
this
story.
Why
is
it
significant?
What
questions
does
it
ask
that
haven’t
been
addressed
before?
What
are
the
possible
problems
in
carrying
out
research
in
this
area?
How
have
I
overcome
these?
As
you
construct
the
review
of
the
literature,
it
is
important
to
demonstrate
your
24 | P a g e
understanding
of
the
topics
by
developing
the
text
further
than
a
simple
review
of
existing
studies.
The
relevant
concepts
should
work
their
way
through
your
review
of
past
and
current
research,
so
that
the
reader
finishes
the
chapter
with
a
comprehensive
understanding
of
the
context
and
need
for
the
study.
If
you
have
published
any
work
of
your
own,
here
is
the
place
to
include
that
also.
Or
if
you
have
appeared
on
a
published
research
article
with
a
colleague,
or
advisor,
you
would
also
include
that
in
this
section.
Any
arguments
that
you
have
made
or
conclusions
that
you
have
reached
can
be
considered
just
as
much
a
part
of
the
review.
Close
this
section
with
a
review
summary,
and
lead
into
the
following
with
mention
of
how
you
propose
to
address
your
research
question.
25 | P a g e
Research
Design
Although
the
previous
chapters
will
have
addressed
some
of
the
statistical
elements
used
in
conducting
your
research,
this
section
will
provide
the
first
in-‐depth
discussion
of
them.
This
is
because
the
collection
of
these
analytical
tools
form
the
theoretical
backbone
for
the
assumptions
made
in
creating
and
executing
your
research
design.
Questions
such
as
why
you
intended
to
seek
specific
data
sets,
as
well
as
your
manner
of
data
collection,
sorting,
and
analysis,
will
be
discussed
at
length.
Research
Questions
The
research
questions
are
the
essence
of
your
study,
and
should
essentially
sum
up
the
valid
intent
of
the
study.
They
have
already
been
posited
in
Chapter
1,
but
they
should
be
repeated
in
Chapter
3.
Use
them
as
a
tool
to
help
focus
the
content
of
the
dissertation
as
it
is
written.
Everything
included
in
the
dissertation
should
be
relevant,
in
some
way,
to
answering
the
research
questions.
It
is
best
to
state
the
research
questions
early
in
Chapter
3
in
order
to
motivate
and
focus
the
remainder
of
the
chapter,
which
is
an
explanation
of
how
you
will
arrive
at
valid
and
reliable
answers
to
your
research
questions.
Population
Correct
choice
of
a
sample
population
in
collecting
data
points
is
possibly
one
of
the
most
critical
elements
of
your
research
design,
and
therefore
a
large
amount
of
detail
concerning
it
will
have
to
be
included.
Choosing
an
incorrect
population
can
undercut
or
completely
invalidate
the
entirety
of
your
research.
When
finished
with
this
section,
a
reader
should
know
26 | P a g e
precisely
why
you
chose
the
sample
population
you
did.
They
should
also
know
why
you
specifically
rejected
other
populations
that
may,
at
least
at
first
glance,
appear
to
be
strong
candidates
for
your
purposes.
These
means
you
should
make
mention
of
rival
populations,
provided
in-‐depth
explanations
of
what
assumptions
underlie
their
use,
and
then
illustrate
specifically
why
they
were
rejected
in
favor
of
the
chosen
population.
Informed
Consent
The
question
of
informed
consent
is
critical,
and
you
will
have
dealt
with
it
previously
during
your
back-‐and-‐forth
with
the
IRB
and
their
particular
regulations.
Nonetheless,
it
is
important
to
explain
clearly
how
informed
consent
was
obtained
from
the
research
participants.
Sampling
Frame
The
sampling
frame
represents
the
range
over
which
you
use
a
given
population.
Sampling
frames
are
intended
to
provide
for
the
most
representative
population
possible
for
a
given
study.
This
may
include
questioning
respondents
over
a
wide
range
of
geographical
locations,
economic
backgrounds,
races,
education
levels,
and
so
on.
The
sampling
frame
will
be
dependent
on
the
particular
research
techniques
that
you
are
employing
in
your
study.
Confidentiality
According
to
IRB
regulations
and
general
scientific
practice,
the
confidentiality
of
research
participants
is
something
that
must
be
ensured
and
promised
to
the
respondents
ahead
of
time.
These
regulations
are
in
place
to
27 | P a g e
guarantee
the
privacy
of
a
person
who
choose
to
participate
in
research
studies,
and
provides
an
umbrella
of
safety
that
makes
respondents
far
more
likely
to
provide
the
most
accurate,
honest
information
possible.
(It
is
very
similar
to
the
concept
of
a
journalist
protecting
his/her
confidential
sources.)
Confidentiality
also
extends
to
the
person
conducting
the
research
study
and
analyzing
the
data.
He/she
does
not
want
to
be
inadvertently
swayed
by
knowing
which
respondent
filled
out
which
research
survey.
This
is
because
the
author
of
the
study
can
just
as
easily,
although
maybe
not
consciously,
insert
biased
opinions
based
on
any
number
of
subjective
factors,
such
as
race,
apparent
education
level,
socioeconomic
status,
etc.
Ideally,
therefore,
the
people
administering
the
questionnaires
would
be
separate
from
the
author
of
the
study;
but
many
times
that
is
not
practically
possible,
and
in
the
more
frequent
case
of
the
author
administering
the
questions
his/herself,
the
general
practice
is
simply
to
keep
the
questionnaires
anonymous.
Further,
the
potential
for
contamination
via
the
administration
process
on
behalf
of
the
author
is
probably
not
too
great,
considering
he/she
will
already
have
formulated
a
picture
of
the
overall
sample
population
anyway.
Geographic
Location
Geographic
location
can
play
a
significant
factor
in
a
research
study.
First
off,
in
choosing
a
diverse
background
for
your
study,
geography
may
very
likely
become
a
factor.
However,
this
brings
along
with
it
certain
difficulties,
particularly
with
regard
to
the
actual
administration
of
the
test
questions
and
in
finding
participants
from
different
locations.
28 | P a g e
Data
Collection
Data
collection
can
range
over
a
number
of
different
examples.
Probably
the
most
straightforward
is
the
administration
of
interview
questions
to
subject
respondents,
as
written
about
at
length
in
the
following
section.
Other
means
of
data
collection
may
involve
retrieval
of
previously
compiled
databases
for
subsequent
integration
with
your
own.
This
may
include
using
data
that
does
not
fall
directly
under
the
heading
of
your
particular
field,
but
which
may
still
be
necessary
for
inclusion
in
your
study.
For
example,
if
you
are
a
social
sciences
student
studying
the
quality
of
life
of
seniors
living
in
assisted
care,
medical
data
on
a
sample
set
of
the
cross
section
of
the
population
that
typically
goes
into
these
communities
may
be
necessary.
This
means
drawing
upon
data
sets
that
have
already
been
compiled
in
previous
studies.
There
are
a
number
of
ways
to
go
about
obtaining
such
data
sets.
The
most
direct
is
to
contact
the
author
of
the
study
directly.
Usually,
this
involves
simply
sending
an
email
to
the
address
provided
for
in
the
author’s
contact
information
section
of
their
published
research.
There
are
also
a
number
of
online
repositories
of
statistical
research
data
ranging
from
medical
to
physical
sciences.
You
can
also
examine
the
dissertation
archive
at
your
college/university.
These
dissertations,
along
with
their
underlying
data
sets,
are
a
part
of
the
public
record
of
research,
and
their
data
can
be
drawn
upon
just
as
can
any
published
academic
journal
article.
29 | P a g e
Interview
Questions
When
constructing
well
thought-‐out,
thorough
questions,
make
sure
to
think
about
all
possible
objections
that
could
be
raised
against
the
conclusions
of
your
hypotheses.
In
seeking
to
achieve
this,
there
are
a
number
of
strong
rules
that
ought
to
be
followed
in
order
to
prevent
the
questions
from
falling
prey
to
unseen
weaknesses.
The
following
section
provides
a
summation
of
what
factors
to
keep
in
mind/avoid
when
putting
together
your
questions.
First,
do
your
best
to
make
the
elements
of
your
question
as
straightforward
as
possible.
Avoid
any
terminology
that
sounds
vague
or
ambiguous.
It
is
also
critical
to
use
terminology
that
you
are
sure
your
research
subjects
will
completely
understand.
The
overall
goal
should
be
to
formulate
a
question
and
use
phraseology
such
that
the
subject
(and
the
reader)
will
know
precisely
what
you
mean,
and
therefore
provide
the
most
accurate
response
possible,
increasing
the
validity
and
reliability
of
your
overall
study.
Next,
keep
the
survey
items
simple
by
including
only
one
concept
per
question.
Once
again,
this
increases
both
the
reliability
of
your
question,
because
the
subject
is
less
likely
to
insert
his/her
own
preferences
30 | P a g e
into
the
response,
or
become
confused
about
what
is
being
specifically
asked.
Avoid
the
use
of
leading
questions.
This
is
extremely
important,
and
at
the
same
time,
a
very
difficult
thing
to
completely
avoid.
All
questioners
insert
some
element
of
bias
into
their
research
items,
and
this
bias
is
then
compounded
by
the
subject
biases
of
the
research
subjects
themselves.
As
a
means
of
attempting
to
avoid
this
practice,
authors
should
do
their
best
to
eliminate
any
kind
of
loaded
questions,
those
with
implies
assumptions,
or
anticipated
outcomes.
These
have
the
ability
of
subtly
suggesting
preferred
responses,
and
oftentimes
can
lead
to
a
respondent
giving
the
answer
that
he/she
either
thinks
you
want
to
hear,
or
is
induced
by
some
other
kind
of
pressure.
If
asking
questions
that
deal
with
sensitive
moral
issues
or
social
norms,
then
this
can
become
a
particularly
difficult
area
to
navigate,
(e.g.,
illegal
drug
use,
spousal
abuse,
etc).
Therefore
you
want
the
questions
to
be
devoid
of
potential
judgment
and
social
pressure,
too.
Relating
to
the
directly
above,
if
at
all
possible
avoid
questions
that
require
sensitive
information
from
the
subjects.
This
includes
staying
clear
of
information
that
asks
about
illegal
behavior,
exact
age,
and
sexual
orientation/preferences/practices.
Questions
of
income
and
personal
relationships
should
be
avoided
also,
if
possible.
However,
depending
on
the
nature
of
your
study,
these
questions
might
well
be
integral
to
the
overall
research
project.
In
such
cases
the
questions
should
be
phrased
as
clinically
and
dethatched
a
manner
as
possible.
And
be
aware
that
no
matter
how
31 | P a g e
apparently
willing
the
participant
and
indifferent
the
question,
when
sensitive
issues
are
broached
the
reliability
of
the
responses
you
will
receive
are
much
more
likely
to
be
less
reliable
and
forthright.
Such
factors
will
have
to
be
taken
into
account
during
the
analytical
phase,
and
dealt
with
especially
in
your
qualitative
analysis.
Do
not
“blind
side”
the
respondent.
In
other
words
you
should
not
ask
any
questions
that
he/she
does
not
already
know,
or
is
not
expecting.
You
want
the
information
provided
to
be
the
most
accurate
and
reliable
possible,
and
providing
questions
that
are
outside
the
agreed
upon
bounds
of
the
original
study
may
well
cause
the
respondent
to
become
less
or
uncooperative.
Furthermore,
you
want
to
be
sure
you
ask
only
questions
that
the
respondents
are
able
to
provide
accurate
answers
for.
This
relates
to
the
appropriate
choice
of
sample
population.
By
asking
questions
beyond
the
reasonable
scope
of
the
respondents,
you
negatively
impact
the
reliability
and
validity
of
your
study.
You
should
also
be
sure
to
make
the
questions
accessible
to
the
respondent.
This
means
using
common
words,
as
opposed
to
complicated
or
technical
terms
that
could
be
perceived
as
intimidating.
Often
a
respondent
will
not
ask
the
meaning
of
a
word
he/she
doesn’t
understand
out
of
simple
embarrassment,
and
as
a
result
may
either
provide
an
incorrect
response
or
skip
the
question
altogether.
Further,
an
overly
complicated
vocabulary
is
more
likely
to
cause
the
respondent
to
feel
intimidated,
and
may
negatively
32 | P a g e
33 | P a g e
and
more
efficient.
Quantitative
analyses
can
be
carried
out
with
much
greater
ease,
and
to
some
extent,
greater
accuracy.
Furthermore,
subjects
tend
to
find
these
times
of
questions
easier
to
respond
to.
This
enhances
their
cooperativity
and
willingness
to
participate
in
the
study.
There
are
associated
disadvantages,
too.
These
questions
may
have
the
unintended
consequence
of
constraining
the
respond
too
much.
They
may
feel
that
the
provided-‐for
answers
do
not
accurately
represent
the
answer
the
have
to
share.
In
addition
to
having
the
potential
effect
of
reducing
cooperativity
among
the
subjects,
but
also
clouding
their
judgment
and
forcing
them
to
choose
and
thus
reducing
the
overall
accuracy
of
the
response.
34 | P a g e
very
words
of
your
subjects,
and
these
elements
can
be
included
in
your
analysis
to
provide
a
more
comprehensive
overview.
However,
there
are
quite
a
few
disadvantages
associated
with
these
types
of
questions.
For
one,
they
are
much
more
open
to
misinterpretation.
Different
subjects
are
more
likely
to
“read
in”
to
questions
differently,
and
a
potential
result
may
be
that
responses
are
so
different
that
they
are
of
limited
use.
There
is
also
a
large
risk
for
the
inclusion
of
irrelevant
information.
If
the
respondent
thinks
you
mean
something
than
you
do,
he/she
may
include
an
irrelevant
answer.
He/she
may
also
include
reasoning
and
explanations
that
are
highly
subjective
and
further
work
to
undermine
the
validity
of
your
study.
There
are
also
strong
practical
problems
with
these.
Just
as
forced
response
questions
are
easier
to
code
and
analyze,
this
are
much
more
difficult
to
deal
with.
A
general
coding
scheme
cannot
be
used
without
essentially
losing
the
richness
of
the
questions
that
might
have
been
the
goal
in
the
first
place.
And
so
a
more
manual
method
of
data
entry
is
required,
and
in
addition
to
requiring
more
time,
this
also
reduces
the
effectiveness
and
accuracy.
Yet
these
types
of
questions
are
not
mutually
exclusive.
It
is
entirely
possible,
and
common,
to
layer
elements
of
forced-‐choice
structure
into
free
response
questions.
So
it
is
more
appropriate
to
view
them
as
lying
along
a
35 | P a g e
continuum
that
can
be
modified
to
include
elements
of
one
another,
depending
on
the
needs
of
the
researcher
and/or
research
project.
It
is
useful
then
to
look
at
an
example
of
how
structure
can
be
built
into
an
otherwise
free
response
question.
Consider
the
following
query:
“What
is
your
opinion
on
the
home
schooling
of
children?”
Overall,
this
question
has
an
unstructured
form.
The
respondent
will
have
the
freedom
to
insert
a
wide
array
of
opinions
that
might
not
be
relevant
to
the
study,
or
may
in
fact
negatively
impact
the
results
by
the
insertion
of
superfluous
information.
This
is
because
there
is
no
structure
or
directionality
built
into
the
question.
Furthermore,
because
it
is
question
that
could
be
read
by
some
as
controversial,
there
is
even
more
of
a
threat
of
overall
contamination.
However,
if
the
question
were
reworded
to
“What
aspects
of
home
schooling
do
you
think
are
positive
for
a
child?
What
aspects
do
you
think
are
detrimental?”
then
the
respondent
is
immediately
forced
into
a
much
narrower
range
of
possible
answers.
You
compel
him/her
to
offer
both
a
positive
and
negative
perspective,
and
instead
force
the
respondent
to
focus
on
particular
aspects
of
home
schooling.
Even
more
structure
could
be
built
into
the
question.
For
example,
you
might
ask
what
particular
subjects
the
respondent
believes
are
more
difficult
to
teach
in
a
home
school
setting:
“Do
you
think
biology
is
difficult
to
teach
in
a
home
setting,
without
a
functioning
science
lab?”
36 | P a g e
Obviously,
this
further
constrains
the
range
of
answers
that
can
be
provided.
The
respondent
is
now
forced
to
consider
a
very
specific
subject
within
the
context
of
home
schooling
(an
element
of
forced-‐choice),
while
having
to
offer
an
opinion
as
to
why
it
is
either
practical
or
not
to
teach
it
outside
a
traditional
school
setting
(an
element
of
free
response).
In
a
question
like
this,
there
would
have
to
be
a
clear
definition
as
to
what
is
meant
by
home
schooling,
what
age
of
students
are
being
considered,
along
with
the
education
level
of
the
parents
and
their
fitness
to
teach
in
particular
subjects.
These
also
have
the
effect
of
corralling
the
respondent
further
into
a
constrained
position.
He
or
she
will
have
to
work
with
a
prefabricated
definition
of
home
schooling
that
may
or
may
not
complete
sync
with
their
own.
Overall,
the
point
is
that
there
is
a
wide
range
over
which
the
researcher
can
work
in
order
to
build
elements
of
structure
into
questions.
By
using
this
sort
of
synthesis
many
of
the
attractive
aspects
of
both
forced-‐choice
and
free
response
can
be
incorporated
into
single
questions,
while
controlling
for
many
of
the
undesirable
elements.
Data
Analysis
In
analyzing
your
data,
it
is
important
to
have
a
strong
understanding
of
the
statistical
and
analytical
tests
or
valid
qualitative
methods
most
frequently
utilized
in
dissertations.
To
that
end,
the
following
provides
a
comprehensive
37 | P a g e
list
of
some
of
the
methods,
coefficients,
and
tests
that
can
be
of
particular
use.
The
methods
for
data
analysis
are
split
into
two
sections:
qualitative
and
quantitative
methods.
The
name
of
each
is
given,
along
with
a
brief
description,
which
includes
the
appropriate
place
of
usage:
Qualitative
Methods
Phenomenology
Phenomenology
is
a
common
method
for
analyzing
the
verbal
or
written
data
from
open-‐response
questions.
Standardized
as
a
phenomenological
research
methodology
by
Moustakas
(1994),
the
modified,
seven-‐step
van
Kaam
(1959)
method
involves
understanding
the
essence,
meaning,
and
structure
of
an
individual's
lived
experiences.
This
methodology
is
used
to
look
for
patterns
and
trends
by
identifying
shared
beliefs
that
have
yet
to
be
addressed
by
existing
literature.
Through
the
seven
steps
of
the
modified
van
Kaam
method,
raw
data
can
be
organized,
reduced,
and
transformed
into
the
essences
of
the
experiences
of
the
participants
as
they
related
to
the
research
goals
of
the
study.
38 | P a g e
Quantitative
Methods
Pearson’s
Correlation
Coefficients
Pearson’s
correlation
coefficients
are
used
to
determine
if
there
is
a
positive
or
negative
relationship
between
two
variables
that
are
continuous
in
nature.
The
coefficient
ranges
from
a
–
1
to
+
1,
with
a
score
of
around
–
1
indicating
that
there
is
a
strong
negative
relationship
between
the
variables
and
a
score
of
around
+
1
indicating
the
there
is
a
strong
positive
relationship
between
the
variables.
If
there
is
a
strong
negative
correlation
then
this
means
that
when
one
variable
increased
the
other
variable
decreased
while
a
strong
positive
correlation
then
this
means
that
when
one
variable
increased
the
other
variable
increased
as
well.
39 | P a g e
Mann-Whitney
U
test
Like
the
Spearman’s
rank
correlation
coefficient,
the
Mann
Whitney
U
test
is
a
nonparametric
version
of
the
independent
samples
t-‐test.
Instead
of
comparing
the
average
values
of
the
dependent
variables
with
the
independent
populations,
the
average
ranks
of
the
dependent
variables
are
40 | P a g e
compared
with
one
another.
This
has
an
advantage
over
the
independent
samples
t-‐test
in
that
one
does
not
have
to
assume
normality
in
the
data.
By
using
the
Mann-‐Whitney
U
test
one
is
able
to
determine
whether
one
of
the
independent
populations
scored
significantly
higher
than
the
other
population
with
respect
to
the
average
rank
of
the
dependent
variable.
If
there
is
a
significant
positive
test
statistic
then
this
would
indicate
that
one
population
scored
significantly
higher
than
the
other
population,
while
if
there
was
a
significant
negative
statistic
then
this
would
indicate
that
one
population
scored
significantly
lower
than
the
other
population.
41 | P a g e
model
which
is
equal
to
the
value
of
the
dependent
variable
when
the
independent
variable
is
equal
to
zero,
B
is
the
coefficient
for
the
independent
variable
and
indicates
how
many
units
change
there
is
in
the
dependent
variable
for
every
one
unit
increase
in
the
independent
variable,
X
is
the
value
of
the
independent
variable
that
is
observed
in
the
data
and
e
is
the
random
error
term
that
is
normally
distributed
with
a
mean
of
zero
and
a
constant
variance.
42 | P a g e
Y=A+B1X1
+B2X2
+BpXp
+e
where
Y
is
the
dependent
variable,
A
is
the
intercept
of
the
model
which
is
equal
to
the
value
of
the
dependent
variable
when
the
independent
variable
is
equal
to
zero,
B1,
B2,
...
Bp
are
the
coefficients
for
the
independent
variables
and
indicates
how
many
units
change
there
is
in
the
dependent
variable
for
every
one
unit
increase
in
the
independent
variable
when
controlling
for
the
other
independent
variables
in
the
model,
X1,
X2,
...,
Xp
are
the
values
of
the
independent
variables
that
are
observed
in
the
data
and
e
is
the
random
error
term
that
is
normally
distributed
with
a
mean
of
zero
and
a
constant
variance.
43 | P a g e
44 | P a g e
45 | P a g e
significantly
differed
from
one
another
with
respect
to
the
average
scores
of
the
dependent
variable
observed
for
each
category.
As
for
the
continuous
covariate
(or
control
variable)
in
the
model,
a
significant
relationship
would
indicate
that
the
dependent
variable
significantly
varies
with
the
independent
variable(s),
which
also
indicates
that
the
covariates
are
able
to
significantly
explain
the
variation
in
the
dependent
variables.
46 | P a g e
dependent
variable
across
time
in
terms
of
whether
it
increased
or
decreased
at
the
different
time
points
in
which
the
dependent
variable
was
measured.
This
means
that
measurements
for
the
dependent
variable
are
taken
from
the
same
participant
at
several
different
time
points
(i.e.,
minutes,
hours,
days,
weeks,
etc.).
The
repeated
measures
ANCOVA
can
then
be
used
to
assess
whether
there
was
a
change
in
the
dependent
variable
for
each
of
the
independent
variables
as
well
as
whether
there
was
a
difference
between
the
independent
variables
as
would
be
conducted
in
a
normal
ANCOVA.
Logistic
Regression
Logistic
regression
is
used
in
order
to
determine
whether
a
single
or
several
independent
variables
significantly
predict
the
dependent
variable.
This
is
similar
to
the
other
regression
analyses
except
that
the
dependent
variable
is
dichotomous.
This
means
that
the
dependent
variable
is
binary
or
is
comprised
of
two
categories.
By
using
the
logistic
regression
model
one
is
able
to
indicate
whether
the
independent
variable
significantly
predicts
the
probability
or
odds
of
the
dependent
variable
occurring.
For
the
purpose
of
the
logistic
regression
model
the
independent
variables
can
be
either
continuous
or
categorical.
47 | P a g e
factor
analysis
finds
the
questions
or
variables
that
explain
the
largest
amount
of
variation
in
the
questions
or
variables.
As
a
result
of
this,
these
questions
or
variables
could
be
combined
to
provide
a
single
measure
for
a
latent
variable.
Reliability
Analysis
A
reliability
analysis
is
used
to
determine
how
correlated
a
set
of
questions
or
variables
are
with
one
another
when
it
comes
to
a
latent
variable.
This
is
often
used
in
conjunction
with
the
factor
analysis
to
illustrate
that
the
questions
or
variables
provide
an
adequate
measure
of
the
underlying
variable.
In
general,
Cronbach’s
alpha
coefficients
are
used
to
provide
information
with
respect
to
the
internal
consistency/reliability
of
the
items,
with
a
Cronbach’s
alpha
of
around
.70
indicating
that
the
questions
or
variables
provide
an
adequate
measurement
for
the
latent
variable
or
a
Cronbach’s
alpha
of
around
.80
indicating
that
the
questions
or
variables
provide
a
good
measurement
for
the
latent
variable.
Findings
This
section
should
be
thought
of
as
an
analytical
summary
of
the
entire
work
thus
far.
It
is
here
that
you
will
pull
the
entire
body
of
the
work
together
to
provide
the
basis
for
your
conclusions.
Here,
you
will
also
make
the
case
for
48 | P a g e
the
statistical
significance
of
what
you’ve
found.
The
findings
need
to
be
placed
within
the
appropriate
context,
not
just
“listed”
in
the
manner
of
“I
did
this,
and
then
this
happened.”
Instead,
each
of
the
questions
you
were
seeking
to
answer
with
each
piece
of
research
should
be
restated
in
the
order
in
which
they
were
presented
in
the
broad
body
of
the
dissertation.
Do
not
omit
mention
of
any
data,
even
if
it
conflicts
with
your
original
hypothesis.
If
there
is
contradictory
data
that
you
do
not
believe
undercuts
your
overall
hypothesis,
explain
precisely
why
that
is.
(Avoiding
mention
of
possibly
contradictory
findings/data
makes
it
seem
as
if
you
are
not
entirely
sure
of
the
position
of
your
own
work.)
There
should
not
be
nearly
as
much
explanation
of
the
statistical
or
qualitative
methods
of
the
study
in
this
section,
as
much
of
this
will
have
been
dealt
with
in
Chapter
3,
but
it
is
still
necessary
to
mention
the
significance
of
certain
elements
on
the
analysis,
and
how
they
were
engaged
in
the
analysis
of
the
data
and
the
generation
of
the
results.
49 | P a g e
Conclusions
This
section
differs
slightly
from
the
summation
of
the
findings.
Here
you
will
discuss
the
conclusions
drawn
from
your
study
and
analyses,
and
how
they
compare
to
the
hypotheses
you
had
originally
formulated.
By
this
time,
the
author
will
have
had
a
great
deal
of
feedback
from
his
or
her
committee,
and
will
have
a
firm
idea
of
how
the
conclusions
have
ended
up
supporting
or
arguing
against
the
hypothesis.
It
is
entirely
possible
that
your
data
may
have
led
to
the
invalidation
of
your
original
hypothesis.
This
is
entirely
satisfactory,
as
long
as
your
committee
accepts
that
the
null
findings
still
contributed
to
the
overall
understanding
within
the
field.
Essentially,
this
section
intends
to
answer,
in
a
succinct
yet
thorough
way,
the
research
questions
of
the
study.
The
results
supporting
the
conclusions
should
be
touched
upon,
but
should
not
be
addressed
in
excessive
detail.
Many
doctoral
students
choose
to
structure
this
section
based
around
the
research
questions
that
focused
the
study.
In
this
section,
you
will
also
make
inferences
based
on
the
findings.
Being
able
to
generalize
your
results
to
broader
cases
beyond
the
specific
question(s)
in
your
dissertation
is
an
important
element
in
demonstrating
the
overarching
importance
of
your
research.
This
involves
being
able
to
phrase
inferences
in
a
well-‐informed
and
logical
manner.
Be
sure
to
draw
upon
the
examples
from
the
Literature
Review
section,
and
show
how
your
results
50 | P a g e
apply
directly
to
previous
work
and
where
they
fall
in
the
context
of
existing
research.
Implications
This
section
presents
general
implications
that
have
arisen
from
the
conclusions
of
your
research.
Implications
can
apply
to
further
research,
policy-‐making,
and
a
number
of
other
areas.
The
implications
of
the
study
can
be
considered
the
reasons
for
why
the
study
was
significant,
and
what
about
the
results
(in
particular)
was
significant
to
whom.
What
are
the
possible
real-‐world
applications
and
effects
of
the
results
and
conclusions
of
the
study?
Typically,
the
most
important,
relevant,
and
salient
implications
should
be
presented
here,
and
students
are
advised
to
remain
focused
and
avoid
falling
into
abstract
discussion.
Limitations
The
limitations
of
the
study
have
been
posited
in
Chapter
3,
but
they
should
be
reviewed
in
the
conclusions
section
in
light
of
the
results
of
the
completed
study.
It
is
appropriate
to
list
the
limitations
and
then
explain
how
they
have
affected
the
results
and
conclusions
of
the
study.
Special
attention
should
be
paid
to
how
the
limitations
of
the
study
affected
the
generalizability
of
the
results
to
other
populations
or
conditions.
Recommendations
It
is
important
to
suggest
future
directions
for
the
work
you
have
produced.
In
addition
to
the
possibility
that
you
may
well
end
up
carrying
out
further
research
on
the
topic
of
your
choice,
it
is
also
possible
that
another
51 | P a g e
researcher
may
wish
to
take
the
work
a
few
steps
further
along.
Moreover,
suggesting
recommendations
for
how
to
continue
your
line
of
work
underscores
your
thorough
understanding
of
what
you
have
done.
By
being
able
to
“see”
several
steps
ahead,
and
know
what
sorts
of
research
questions
and/or
methodologies
should
be
applied
in
order
to
begin
approaching
those
issues.
While
you
can
speculate
on
what
you
think
the
answers
to
these
questions
may
be,
based
upon
the
results
you
came
to,
as
well
as
those
results
which
have
already
been
published,
but
do
not
attempt
to
conclusively
answer
them.
Summary
This
is
the
final
section
of
the
chapter
and
the
dissertation.
It
is
a
succinct
summary
of
the
conclusions,
implications,
and
recommendations
covered
in
chapter
5,
and
should
contextualize
those
elements
in
the
original
problem
and
purpose
of
the
study.
A
new
reader
should
be
able
read
the
summary
section
at
the
end
of
chapter
5
and
understand
the
study
and
its
results
in
its
entirety.
References
All
sources
should
be
properly
referenced,
according
to
the
APA
format,
or
if
your
college/university
has
different
requirements,
then
in
line
with
their
requested
guidelines
or
style.
Failure
to
properly
reference
a
source
can
lead
to
investigations
into
investigative/scientific
misconduct,
and
call
into
question
the
validity
and
believability
of
the
entire
body
of
your
work.
52 | P a g e
Accurate
referencing
is
key
for
readers
to
be
able
to
do
further
research
on
your
topic
of
study
after
they
have
completed
reading
your
dissertation.
53 | P a g e