MAIN RESEARCH CONCEPTS

1. THEORIES
A body of interrelated principles or generalized
statement that explain or predict phenomena
A logical system that organizes & explains that facts
Theories help answer the research questions and
achieve research objectives
Theory can be divided into 2 components which are
ABSTRACT CALCULUS & CONSTRUCT

2
Continued..

• Ex: Speech act theory, Communication

accommodation theory, Social penetration theory,
Uncertain reduction theory, Social cognitive theory,
Media systems dependency theory, Agenda setting
theory, Cultivation theory, Muted group theory and
hundreds more communication theories.

3
2. ABSTRACT CALCULUS

• Logical structure of relationship

• A logical method in which conclusions are drawn
from premises of a theory
• Ex: Major premise: if (P) name calling then (Q)
enemy/crisis
Minor premise : P
Therefore the conclusion : Q

4
 • CONCEPT
Term that express an abstract idea formed by generalizing
from particulars and summarizing related observations
Reasons of its importance:
• Simplify research process by combining particular
characteristics, objects, or people into general categories.
• Simplify communication among those who have shared
understanding of them.
An abstraction and need to be defined clearly under the
concept definition section of your proposal/research
Contains the detail characteristics or descriptions of
dimensions of constructs of the theories.

5
• CONSTRUCTS
A combination of concepts that is created to describe a specific situation.
Can’t see but exist ex: credibility, newspaper sensationalism, self
disclosure, power distance, job satisfaction, attitude, values etc.
The term used to describe a collection of variables that have been found
It has THREE distinct characteristics:
1. An abstract idea that usually broken down into dimensions represented
by lower-level concepts.
2. Because of abstraction, a construct cannot be observed directly.
3. A construct is usually design for a specific research purpose so that its
exact meaning related only to the context in which it is found.

6
 • DIMENSIONS
Smaller unit of construct of theory and construct
Construct can’t be observed directly and it has to be broken
into smaller unit called dimensions (smaller level concepts).
Eg: Authoritarian is a construct in Functional theory (group
communication theory) and the dimensions for this
construct are Conventionalism, submission, and cynicism.
‘Dimensions’ later on will be the most important aspect to be
considered during the development of items for your
research inventory/ questionnaire or in developing
interview protocol for in-depth or focus group interview

7
 VARIABLE
• A phenomenon or event that can be measured or
manipulated.
• Variable is empirical counterpart of a construct or concept.
• Variable is an important link between theory and empirical.
• Variable are the phenomenon and events that are measured
and manipulated in a research.
• Construct dimensions, situation (temperature, social
situation, event etc), personal background of the respondents
etc. can be the variables of a research.
Example:

8
 VARIABLES
• Independent variable: the variable that is systematically varied by
researchers and will be the input of hypothesis.
• Dependent variable: the variable that is observed and whose value is
presumed to depend on the effect of dependent variable (s) (Is what the
researcher wishes to explain). It’s will be the output of a hypothesis.
• Discrete variables: a finite set of values; it cannot be divided into sub parts.
E.g. Gender 1) Male 2) Female
• Continuous variables: can take on any value (including fractions) and can be
meaningfully broken into smaller subsections. E.g. Income, duration, height.
• In non-experimental research, where there is no active manipulation of
variables, different terms are sometimes substituted for independent and
dependent variables.

9
 VARIABLES
• Variable that is use for prediction or assumed to be causal is called
predictor variable or antecedents variable. (Independent variable)
• Variable that is predicted or assumed to be affected is called criterion
variables. (Dependent variable).
• The easy way to identify dependent (output) and independent (input)
variable is by looking at hypothesis
• Control variable are used to ensure that the results of the study are due to
the independent variables, not other sources. Sometimes, the researcher
used a control variable such as age, gender, or socioeconomic status to
divide subjects into specific, relevant categories.
For instance, look at the following hypothesis;

“People with high self-esteem read newspaper more often than people with
low self esteem”
The independent variable in this research hypothesis is self-esteem and
reading newspaper is dependent variable
10
 DEFINITION OF CONCEPTS OR VARIABLES
Theories provide relationships among concepts.
Concepts must be able to be defined clearly to enable them to be studied.
Definition can be categorized into 3 levels which are daily, poetic and scholarly
Scholarly definition of concept definition is the one used in research
Scholarly definition is certainly technical because it identifies elements to be
examined.
Ex: Motivation can be defined as “The process by which people initiate, direct and
sustain behaviors in order to fulfill psychological and physiological needs”
There are 5 other sub concepts related to processes involve in motivation definition
that need to be define clearly which are; 1) initiate, 2) direct, 3) sustain,
4)psychological & 5) physiological needs.

Scholarly definition for a concept can be found in many places such library or on the
internet (handbook and essay collections, textbooks, special dictionaries, or
journal articles)

11
I. CONCEPTUALIZATION
The process of specifying the vague mental imagery of
our concept sorting out the kinds of observations and
measurements that will be appropriate for our research.

II. OPERATIONALIZATION
• Operationalization means to add measurable indicators to complex
concepts to make them measurable variables.
• Necessary render the term observable and make the concept
empirical.
• Either experimental type or measured type (nominal, ordinal, interval
or ratio; quantitative or qualitative)

12
• Research depends on observation and observation
based on a statement.
• An operational definition is such a statement

Variable Operational definition

Gender Asking respondents to circle

M (Male) or F (Female)

Need for cognition Summated scores on a 4-point

Likert scale to eight cognition
scales.

13
 QUALITATIVE AND QUANTITATIVE RESEARCH

• The only difference between qualitative and

quantitative research is the style of questioning.
• Qualitative research uses flexible questioning;
quantitative uses standardized questions.
• Assuming that a large sample sizes are large enough
and that the samples are properly selected, the
results from both methods can be generalized to the
population from which the sample was drawn.

14
MEASUREMENT
A researcher assigns numerals to objects, events, or
properties according to certain rules.
Level of measurements:
Nominal
Ordinal
Interval
Scale

15
Nominal: the level of measurement at which arbitrary
numerals or other symbols are used to classify people,
objects, or characteristics.
• Weakest form of measurement.
• Numerals or other symbols are used to classify people
object or characteristics.
E.g.:
Gender of respondent
1. Male
2. Female

16
Ordinal: the level of measurement at which items are
ranked along a continuum that is for example from
smallest to the largest.
• Possesses the property of equivalence and property
of order.
E.g.: Instruction: Rank the following brands of shampoo
based on your preference.

BRAND RANK ORDER

1. Sunsilk 2

2. Head & Shoulder 3

3. Rejoice 1

4. Himalaya 5

5. Pantene 6

6. Clear 4
17
Interval: a measurement systems in which an intervals
between adjacent points on a scale are equal (e.g.
thermometer).
• Incorporate the formal property of equal difference;
that is numbers are assigned to the positions of
objects on an interval scales in such a way that one
may carry out arithmetic operations on the difference
between them.
• Lacks a true zero point.
E.g.: Select your income category:
Below RM1000
RM1001 – RM1999
RM2000 – RM3999
Above RM4000
18
Scale: a level of measurement that has all the properties
of an interval level scale and also has a true zero point
(e.g. income)

19
 METHODS OF SCALING OR MEASUREMENT

Scale: a form of measurement

Specialized rating scales:
i. Psychometric Scales (Likert-type, multiple choice,
semantic differential and dichotomous scales)
Purpose is to place respondents along a dimension/scale.
Eg: want to measure respondents communication
competent or power distance or IQ etc. which are in
form of psychometric scaling.

20
Continued….
• Subject are asked to a series of items all measuring the same
dimension using any type of suitable scale:
Examples:
# Likert-type scale (strongly agree, agree, neutral, disagree, strongly
disagree),
#Dichotomous (true/false, agree/disagree, yes/no etc)
#Multiple choice
(Satisfied 1 2 3 4 5 6 7 Dissatisfied)
#Semantic Differential Scales
Used to measure the meaning an item has for an individual.
Name or concept is placed at the top of a series of 7-point scales anchored
by bipolar attitudes. E.g.:
COM566 lectures are:
Boring __:__:__:__:__:__:__ Interesting
Useful __:__:__:__:__:__:__ Useless
Respondents are asked to place an “x” on the scales. 21
ii. Thurston scaling/Consensual location scaling

Purpose is to determine how people perceive differentiate a

set of objects on some defined dimension or scale;
Ex: how voters rate political candidates on a BN/Bebas
dimension
Respondents or judges are given a set of object or dimension
then they are asked to rank order the object, or rate the
object, or sort the object along the dimension and
continuum

22
Continued…

The average of rating/ judging/ ranking/ sorting of all

respondents determines the location of each object on the
scales (points) relative to one another

During the data collection, respondents will check the

statements which they agree. The scale is scored by adding
the points for all items for which the respondents indicate
agreement

23
iii. Guttman scale
The purpose is to place respondents along attitudinal
dimension
Items are written and rank ordered in intensity of expressed
attitude within each question
Respondents are instructed to indicate which items they agree
with and at which point along the attitudinal dimension they
no longer agree. The item where they stop agreement
indicates the attitudinal position for each item
Difficult to develop and not popular

24
SUMMARY OF SCALING
Likert Scale • commonly used scale in mass media research
• statements are developed with respect to a topic and
respondents
• in terms of ordinal –level categories – strongly agree,
agree, disagree, strongly disagree
Guttman • requires a great deal of time and energy to develop
Scaling • scale used in political science, sociology, public opinion
research and anthropology
• difficult to develop and not popular
• measured through questionnaire, vote or observed
characteristics
• In terms of response to indicate which items they agree
at which point
Thurstone • is an interval measurement scale
Scales • time consuming and labor intensive
• commonly used in psychology and education research
• In terms of ranking the order of the subject. Eg: How
voters rate political candidates on a PAS/DAP dimension 25
 RELIABILITY
• Reliability: the property of a measure that consistently
gives the same answer at different times . Replicability –
could produce similar result
Ways to check reliability of a measurement or instrument
▪ Test-retest – over time reliability, coefficient of consistency
▪ For summated scales- internal consistency, coefficient of
homogeneity (split halves, odd-even, item-analysis,
Cronbach's coefficient alpha)
▪ For alternative forma- coefficient of equivalency
▪ If using raters/coders- inter rater/ inter coder, reliability,
coefficient of agreement, Scott’s pi

26
VALIDITY
• Validity: the degree to which a test actually measures what it
intends to measure
• Accuracy – freedom from systematic error; freedom from bias;
correctness
• Instrument validity – measure what it was intended to measure
• Ways to check the validity of an instrument
1. Face/content/ expert/ judgmental validity: examining the
measurement device to see what it appear to measure. A logical
validity check: does the range of content cover the domain or
construct of the concept being measured; does expert agree
that the content is valid; does it look like it measures what is
intended?

27
2. Predictive validity: To test whether the measurement instrument
can predict something. Eg; for measures that support to predict
future behavior, such as entrance exam, job skills etc.) how well
does the measure correlate the actual future performance?
3. Concurrent validity: the measuring instrument is checked
against some present criterion; how well does this measure
correlate with similar measures of related concept?
4. Construct validity: A logical & empirical validity check which
relates a measuring instrument to some overall theoretical
framework to ensure that the measurement is logically related
to other concepts in the framework

28
THANK YOU

29