Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
Research Design
and
Measurement
design
Descriptive
and
scales
Construction
of
RESEARCH DESIGN
Meaning
A
research
project
conducted
research
from
the
problem
This
framework
of
conducting
Definition
According
to
Kerlinger,
Research
is
methods
the
specification
of
and
procedures
for
overall
framework
operational
of
the
pattern
project
or
that
Exploratory Descriptiv
Research
e Research
Design
Design
Experimen
tal/Causal
Research
Design
means
to
explore
the
research
is
type
of
and interviews
For example:
It
is one thing
to describe the
crime
countries.
rates
in
different
It
is
quite
different
thing
to
Why
increasing? Or
Why
Exploratory
research
provides
research is a type of
Techniques of Exploratory
Research Design
Literature Research
The
individuals
executives,
can
sales
be
top
managers,
Ex:
Henry
Mintzberg
interviewed
Case study
It
involves
study
of
the
one
comprehensive
or
few
specific
example:
management
the
of
effective
distributor
Focus
group:
carefully
selected
if
company
manufacturing
studies
are
designed
to
describe something.
Ex:
the
composition
groupings
Number
Descriptive
undertaken
learn
studies
in
about
characteristics
organizations
and
of
age
Educational
Job
status
level
to
describe
a
group
are
of
Example:
a
full-time/part-time
HYPOTHESIS TESTING
Studies
nature
of
certain
relationships, or
establish
groups or
the
Example:
A
Here,
if
advertising
is
establishes
the
difference
in
regard
to
whistle-blowing behaviour.
their
CAUSAL STUDY
Causal
X causes variable Y.
So
research
time,
therefore
the
Longitudinal
1.
2.
3.
What is an Experiment?
The
statistical
hypothesis,
relating
to
some
experiment
experiment
: Age
1. Dependent variable:
If one variable depends upon
or is a consequence of the
other variable, it is termed as
a dependent variable.
2. Independent variable:
If the variable is antecedent to
the dependent variable it is
termed
as
an
independent
For ex:
Height
Height
depends on gender
variables
independent
other
than
variables
the
which
4.
Experimental
and
control
groups:
When
novel
condition,
it
or
is
experimental group.
special
termed
as
Groups
Treatment
Experimental
Group I
Experimental
Group 2
Experimental
Group 3
Control group ( no
treatment)
Treatment
effect (%
increase in
production
over prepiece rate
system)
10
15
20
5. Treatments:
The different conditions under
which experimental and control
groups are put are usually termed
to as treatments
Ex: selling cookies with free gift and
without free gift
Basic principles of
Experimental design
1. Principle of Replication
2.
Principle of Randomization
3.
1. Principle of Replication
(Reproduction):
According
replication,
to
the
the
principle
of
experiment
doing
so
the
statistical
and
grow
and
We
of
two
parts
and
draw
parts,
Grow
parts and
the
parts.
We
the
varieties
and
draw
reliable
in
comparison
to
the
2. Principle of
Randomization
Principle of randomization
provides
protection.
It
For ex:
if
3. Principle of Local
Control
And
the
field
into
several
QUASI
EXPERIMENTAL
DESIGN
a.
Pretest
and
posttest
with
Experimental Group
b.
Posttest
Experimental
group
only
and
with
Control
2.
TRUE
EXPERIMENTAL
DESIGN
a.
Pretest
and
Experimental
Posttest
and
Group
b.
Blind studies
c.
with
Control
3. STATISTICAL DESIGN
i.
Completely
Randomized
iii.
Latin
Square
Design)
iv.
Factorial Designs
Design
(L.S.
1.
QUASI EXPERIMENTAL
DESIGNS
. It
is so because there is no
a.
. An
the
effects
of
the
Group
Pretest
Score
Experiment
al Group
Treatment Posttest
introduce Score
d
O1
O2
O Observation or Measurement
X Exposure of a group to an experimental treatment
effects
of
the
treatment
are
Group
Experimental
Group
Treatment
introduced
Control Group
Outcome
O1
O2
designs,
which
groups
and
record
to
as
the
true
treatment
is
experimental
a.
Pretest and
Experimental
Group
. Two
Posttest with
and
Control
to
the
pretest
and
Group
Pretest
Treatme
nt
introduc
ed
Posttest
O2
Experiment
al Group
O1
Control
Group
O3
Treatment Effect =
O4
b. Blind Studies:
In
case
of
pharmaceutical
developed
drugs
in
the
the
subjects
in
the
For ex:
Training
programs
might
have
been
3. STATISTICAL DESIGN
a. Completely Randomized Design
(C.R. Design)
b. Randomized Block Design (R.B.
Design)
c.
Latin
Square
Design)
d. Factorial Designs
Design
(L.S.
The
essential characteristic of
assigned
experimental treatments.
to
In
two
designs.
principles
of
experimental
equal
to
treatments
the
number
of
FERTILITY LEVEL
Seed
Differ
ence
II
III
IV
X1
X2
X3
X4
X5
d. Factorial Design:
The
used
to
test
two
or
more
types:
i.
ii.
Validity in
Experimentation
Validity
The researchers must make sure
that
any
measuring
instrument
Reliability
Refers
Internal validity:
It refers to the
extent
does
the
research
permit
us
to
say
design
that
the
Internal
by
(independent
question.
the
treatments
variables)
in
External
validity
validity:
External
refers
to
the
the
experiment
can
beyond
the
situations.
result
be
of
an
generalized
experimental
Ex:
Test
Their
sales
improve
because
knowledge
rather
productivity
than
program.
and
the
of
might
their
experience
compensation
Testing
Testing
before-and-after study.
Instrumentation
A
Selection bias
Sample
differential
selection
of
respondents.
Mortality
Some
experiment
before
it
is
may
different
from
the
Environments of Conducting
Experiments
Laboratory
Environment - In a
laboratory
experiment,
the
researcher
conducts
the
experiment
environment
in
an
artificial
constructed
Field
experiment
actual
is
conducted
market
in
conditions.
real-life
environment.
nature
of
the
Variables in Research
Variable
A variable is anything that can
take
on
differing
or
varying
Types of Variables
Dependent
Variable
Independent
variable
Moderating
variable
Extraneous
variable
Intervening
variable
have
contributory
originally
relationship.
significant
effect
stated
IV
on
the
DV
4.Extraneous
variables
(EV):
These are the variables
other
than
independent
the
variables
phenomenon
but
The
term
measurement
means
a measurement
Scaling:
extension
Scaling
Scaling
of
measurements
located.
an
measurement.
involves
continuum
is
on
on
creating
which
objects
are
SCALE
ORDINAL
SCALE
INTERVAL
RATIO
SCALE
SCALE
NOMINAL SCALE:
In nominal scale, numbers are used to
identify
or
categorize
objects
or
events.
For example, the population of any
town may be classified according to
gender as males and females or
according to religion into Hindus,
Muslims, and Christians.
No. 2.
Do you have a car?
(a) Yes
(b) No
The
added,
subtracted,
multiplied
or divided.
The
Therefore,
frequency
ORDINAL SCALE:
This
is
the
next
higher
level
of
measurement.
The
Rank
The
Example:
Rank the following attributes while choosing a
restaurant for dinner. The most important attribute
may be ranked 1, the next important may be
assigned a rank of 2 and soInon.the ordinal scale,
the assigned ranks
cannot
be
added,
multiplied, subtracted
or divided. One can
compute
median,
and percentiles
of
the distribution. The
other major statistical
analysis which can be
carried out is the rank
order
correlation
coefficient,
sign
test.
INTERVAL SCALE:
The
scale.
In
addition,
the
units
of
measure
or
RATIO SCALE:
This
It
Example:
Measures of weight, height, length, etc
All
mathematical
and
statistical
SCALING
Scaling
may be considered as an
extension
of
measurement.
It
Scaling techniques or
classification of scales/Attitude
Scaling
scales
techniques
techniques
Rating Scales
Ranking scales
Graphic Rating
Scale
Itemized Rating
Scale
GuttmanScale/
Scalogram
Likert Scale
Semantic
Differential Scale
Thurstone Scale
Staples
Multi
Scaling
Scale
Dimensional
Method of Paired
comparison
Method of Rank
Order
Two
main
categories
of
Attitudinal Scale
RATING SCALES
Rating
response
scales
have
categories
several
and
are
RANKING SCALES
Ranking
scales
make
RATING SCALE
1.
. Respondents
placing
at
the
i.
Guttman Scales/Scalogram
Ex:
Customers expectation on
Reliance Fresh
Item
No.
(i)
(ii)
Expectation
Would you expect price discounts in
Reliance Fresh?
Do you need free door delivery
service?
(iii)
(iv)
Item Number
(ii)
1.
2.
3.
4.
5.
Responden
t Score
(iii)
The
scale
compare
is
widely
the
used
images
to
of
of
five-or
rating scales.
seven-point
This
of
statements
(items)
are
literature
experience,
and
survey,
personal
discussions
knowledgeable persons
with
On
the
basis
of
above
procedure,
the
The
statements
questionnaire.
are
embodied
in
v. Stapels Scale
Used
as an alternative to semantic
differential scale.
The
Stapel
Scale
consists
analytical
of
group
techniques
of
which
is
technique.
computer
based
RANKING SCALES
i.
Under
this
respondent
method,
can
express
the
his
scales.
For
on
attributes
such
as
Construction of
Instrument
The
quantitative
of
theoretical variable
an
abstract
scales: developed on ad
scales
Guttmans
Thurstone
scalogram analysis
Consensus
scale
scaling
Measurement Error
This occurs when the observed measurement on a
construct or concept deviates from its true values.
Reasons
Mood,
Variations
taken
A
Some
questionnaire
spreadsheet
to
the
reliability
reliability
Cronbachs
Alpha
Validity
validity
Concurrent
Predictive
validity
validity
Sensitivity
Sensitivity refers to an instruments
ability to accurately measure the
variability in a concept.