Lecture4_5

Lecture4_5 - Lecture
4.5
...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Lecture
4.5
 Transla0ng
Theories
into
Research
 Hypotheses.
 Quick
Recap
 •  Theories are abstract knowledge claims constructed using symbolic representations of reality. •  According to the epistemological assumptions of the scientific world view, knowledge claims can not be assumed to be true a priori; they must be empirically confirmed as such. –  Empirical evaluation requires that measurements of reality be taken and the values of these measures be compared to those values suggested by the theory. •  In order to be empirically evaluated, the abstract theoretical statement must be translated into more concrete terms. •  The process of translating an abstract concept into a concrete and empirically measurable variable is called operationalization. Quick
Recap,
Con0nued.
 •  Hypotheses are the concrete knowledge claims we use to evaluate the validity of the abstract theoretical knowledge claims. –  Hypotheses: Concrete Statements About Relationships Between Variables. •  X is positively correlated with Y •  X is negatively correlated with Y •  Controlling for the affects of Z on Y, X is Negatively correlated with Y. Quick
Recap,
Con0nued.
 •  Variables come in at least three types: –  Dependent: the variable that changes in response to a change in the independent variable –  Independent: the variable responsible for inducing a change in the dependent variable. –  Control: variables that are known to be related to the dependent variable; these are included in the hypothesis in order to see if the independent variable significantly influences the dependent variable above and beyond the affects of the C.V. Quick
Recap,
Con0nued
 •  Ultimately, scientists are concerned with causality. –  Causality is nearly impossible to fully determine due to the constant problem of spurious relationships and the existence of unknown variables. •  Scientists talk about causality by “hedging their bets”: –  Speak of our results in terms of correlation –  Try to establish Time Order –  Try to specify the exact mechanism by which reality unfolds. –  Try to determine necessary and sufficient conditions. –  Empirically Test Null Hypotheses, not Research Hypotheses Null
vs.
Research
Hypothesis
 •  My theory tells me a change in the value of X causes a positive change in the value of Y. –  My Research Hypothesis is thus: an increase in the value of X co-occurs with an increase in the value of Y. •  My Null Hypothesis, though, is either: –  a change in X is unrelated to a change in Y. –  A change in X is negatively correlated with a change in Y. •  The goal is to reject or fail to reject the null hypothesis; you do not prove it or fail to prove it. –  By rejecting it, you are saying that your research hypothesis may be true (given further empirical evaluation). •  Why? Variables:
A
Closer
Look?
 •  Variable: an empirical property that can take on two or more values (called attributes). –  Range of Variation in the Values of Attributes. •  How low do you go; how high? Do you cover the full spectrum of possible answers? –  Variation Between Extremes •  How fine of a cut will you make between each value. –  Age: Yearly (1,2,3,4), Grouped ( 10-19, 20-29, etc), Categorical (Child, Teen, Adult)? •  Why does it matter? –  Number of car Accidents? Sexual History? Variables:
Yes,
A
Closer
Look.
 •  Values of Variables must be: –  Exhaustive: All possible responses have been included as potential values. •  Gender: Male and female? (What about Trans-Gender) •  Race: White, Black, Asian? (What about Hispanics, Indigenous Peoples, Middle Eastern, Mix? –  How does variation between extremes relate to this? –  Mutually Exclusive: Response Categories are Distinct and do not overlap. •  How old are you? A) 10-20; B) 20-30; C) 31-39; D) 40-49. •  What is your Profession? A) Law Enforcement, B) Fire Fighter, C) Highway Patrol Officer, D) Paramedic Variables:
S0ll
Lookin’
at
‘Em.
 •  Variables
can
be
measured
at
different
“levels”
 –  A
“level
of
measurement”
deals
with
two
things:
 •  The
“amount
of
informa0on”
that
can
be
garnered
 about
the
empirical
reality
being
measured.

 –  Are
things
simply
different,
or
are
they
hierarchically
ranked?
 Is
the
distance
between
ranks
constant
or
arbitrary?
Can
you
 mathema0cally
manipulate
the
variable
(i.e.
take
average,
 speak
of
“half
of
this”
or
“twice
as
much
of
that”

 •  The
type
of
sta0s0cal
techniques
that
can
be
used
to
 describe
the
rela0onship
between
the
variables.
 Nominal
Measure
 •  A
level
of
measurement
describing
a
variable
that
 has
aTributes
that
are
merely
different,
as
 dis0nguished
from
ordinal,
interval,
or
ra0o
 measures.

 •  Gender
and
Race
are
common
examples
of
a
 nominal
measure.
 –  Male=
1;
Female=
2.
 –  All
this
says
is
males
are
different
than
females;
 although
the
female
aTribute
is
1
more
than
male
the
 male
aTribute,
it
doesn’t
not
mean
that
female
is
 “more
gendery”
than
male.
 Ordinal
Measure
 •  A
level
of
measurement
describing
a
variable
 with
aTributes
we
can
rank‐order
along
some
 dimension.

 •  An
example
is
socioeconomic
status
as
 composed
of
the
aTributes
high,
medium,
 low.
 –  The
aTribute
“high”
does
represent
more
 socioeconomic
status
than
the
aTributes
 “medium”
or
“low”.
 Interval
Measures
 •  A level of measurement describing a variable whose attributes are rank-ordered and have equal distances between adjacent attributes. •  Temperature is a good example of an interval measure. –  Assume your measure of temperature has 3 choices: 1) 70-75F, 2) 76-80F, 3) 81-85F. –  Unlike ordinal measures where you only know that the attributes are hierarchically ranked, interval measures also tell you that the range of measures between 70-75 (i.e. 71,72,73, 74,75) is the same the same as the range of measures between 76-80 (i.e. 77, 78, 79, 80). Ra0o
Measure
 •  A level of measurement describing a variable with attributes that have all the qualities of nominal, ordinal, and interval measures and in addition are based on a “true zero” point. –  Because of this “true zero” point, you can now specify “by how much” one attribute is different from another. •  Example: Age –  10 years old is twice as old as 5 years, and half as old as 20 years. –  You can also take averages, standard deviations, etc. Measurement
Quality
 •  How
Precise
and
Accurate
is
your
measure?
 •  When
we
refer
to
measurement
quality,
we
 talk
about:
 –  Reliability
 –  Validity
 Reliability
 •  Highly
reliable
measurement
methods
 produce
the
same
data
(i.e.
responses
to
 variables)
over
repeated
measures.
 •  The
ques0on
“Did
you
aTend
religious
 services
last
week?”
would
have
higher
 reliability
than
“About
how
many
0mes
have
 you
aTended
religious
services
in
your
life?”
 Tests
for
Checking
Reliability
 •  Test‐retest
method
‐
take
the
same
measurement
more
 than
once.
 –  Take
a
measure
once,
follow
up
again
in
the
future
and
take
the
 measure
a
second
0me.

If
reliable,
the
two
measures
should
be
 the
same.

 –  Say
you
have
10
indicators
of
prejudice(i.e.
10
ques0ons
on
a
 survey).

 –  Split
the
10
between
two
different
variables
and
compare
.
 –  If
the
two
measures
classify
people
differently,
measures
are
not
 reliable.
 •  Split‐half
method
‐
make
more
than
one
measurement
of
a
 social
concept
(prejudice).
 •  Use
established
measures.
 •  Check
reliability
of
research‐workers.
 –  Differences
between
interviewers.

 –  Differences
between
coders.

 Validity
 •  A
term
describing
a
measure
that
accurately
reflects
 the
concept
it
is
intended
to
measure.
 –  Example:
IQ
would
seem
a
more
valid
measure
of
 intelligence
than
the
number
of
hours
spent
in
the
 library.

 •  Rela0ve
validity
can
be
determined
on
the
basis
of
 face
validity,
criterion
validity,
content
validity,
 construct
validity,
internal
valida0on,
and
external
 valida0on.

 Face
Validity
 •  That
quality
of
an
indicator
that
makes
it
seem
 a
reasonable
measure
of
some
variable.
 –  That
the
frequency
of
aTendance
at
religious
 services
is
some
indica0on
of
a
person’s

religiosity
 seems
to
make
sense
without
a
lot
of
explana0on.
 –  Number
of
hours
a
person
spends
in
the
gym
 doesn’t
make
as
much
sense.
 Construct
and
Content
Validity
 •  Construct
Validity
 –  The
degree
to
which
a
measure
relates
to
other
 variables
as
expected
within
a
system
of
 theore0cal
rela0onships.
 •  Marital
Sa0sfac0on:
Desire
to
con0nue
being
married
 to
ones
spouse.
 •  Probability
of
Commifng
Adultery.
 •  Willingness
to
Go
to
Marriage
Counseling

 •  Content
Validity
 –  Refers
to
how
much
a
measure
covers
the
range
of
 meanings
included
within
a
concept.
 •  High‐Class:
Value
of
House?
Or
Value
of
House
–
Debt
 Owed?
Value
of
Car?
Value
of
Clothes?

 ...
View Full Document

Ask a homework question - tutors are online