Origins of Research Questions
and Process
What do research projects look like?
Types of social research
1. DESCRIPTIVE RESEARCH: (who, where, when
and what)
2. EXPLORATORY RESEARCH: (Identifying
patterns and relationships)
3. EXPLANATORY RESEARCH: (model
development and testing)
4. EVALUATION RESEARCH: ( seeks to
determine cause and effect of policy or
programs)
Research Process
1.
2.
3.
4.
5.
6.
Identify Research Question.
Review literature on that topic.
Form theory and hypotheses if explanatory
study.
Construct Research Design.
Test.
Conclude: Review and Implications.
Origin of Research Questions
 Real-World stimuli
 Curiosity
 Imagination
 Reformulation of research questions in
a literature.
 Evaluation of Public Policies
 Even though we may be thinking theoretically by
beginning to relate concepts to each other, we still
need to make a professional jump to a scientific
research program.
 This is a necessary step before we begin to more
formally think about a problem and know we can
make a contribution.
How do we know we can make a contribution and
how?
Answer: Undertake a review of the literature.
Purpose of Literature Reviews
 Lit reviews reveal the scientific community
researching a given topic. In doing so
establishes certain research questions as
important (albeit sometimes subject to fads).
 Research questions should also be
compelling to real world, naturally.
 To devise a research project without
reviewing lit may result in redundant and
insignificant research of interest to no one.
Goals of A Lit Review
1.
2.
3.
4.
5.
Again, identify a community of scholars.
Establishes your work in this community.
In doing so, survey variety of theories,
methods, etc. in other work.
Reformulate your question relative to other
work.
Form your work to be commensurate with
others in regard to paradigms, conventions,
theory, and data.
Theory and Testing
• Explanation relies on theory.
What is a theory?
A theory is a postulation based on logical
suppositions that explain generalizations.
Theories answer “Why” questions.
With Theory we Generalize and Test
• Using systematic methodologies and
theories help us learn about the world.
• We begin by describing our problem and
then move on to explaining our phenomenon
of interest.
• We do this by constructing theories and
testing them.
Important traits of theories
No theory can be completely correct.
2. Theories are simplifications of reality.
3. The goal is to seek general patterns.
4. Theories specify cause and effect
relationships between variables. X causes
Y
Theories and Concepts
 Theories order concepts as variables
specifying directional relationships.


For example, education (X) increases income
(Y).
Income (X) increases chance and individual
votes Republican (Y).
We test our theories by testing hypotheses.
Hypotheses are general propositions
(assertions).
For example,
H1: Dictatorships get into more wars than
democracies.
The Hypothesis
Testable Propositions linking theory to
knowledge
Hypotheses
 A hypothesis is a testable proposition that
formalizes the expectations of a theory.
 It does this by stating the relationships of
variables in a theory.
 One theory may produce several
hypotheses, each tested singularly.
Hypotheses continued
 Hypotheses are usually stated as direct or inverse
relationships, especially if they are your own, and
not the null.
 Hypotheses state how one variable affects another
variable, in other words how the Dependent Variable
is affected by an Independent Variable.
 We seek to discredit the Null Hypothesis, which
states that there is no relationship between the two
given variables.
For Example:
H1: State economic development increases
democracy. Direct relationship (inverse
relationship would be dev decreases dem)
H0: There is no relationship between Economic
development and democracy. (Null hypothesis,
note not stated in negative or inverse)
Which are the independent and dependent
variables?
Relationships between variables
Y
Direct or positive relationship
No relationship
Inverse or negative relationship
0
X
Hypotheses continued
 The variable being affected is the dependent
variable, and the affecter is the independent
variable.
 Other names for dependent variable is the
outcome variable or Y.
 Names for independent variables include
explanatory, covariates, X.
Cause & effect
SINGLE CAUSE
X
SIMULTANEOUS CAUSE
Y
Y
X
MULTIPLE CAUSE
INDIRECT CAUSE
X1
X2
X3
Y
X
Z
Y
Issue of Causality
 It is common to say something causes something
else, although we tend to use this word loosely. A
proper synonym is AFFECT.
 Causality is a deep issue that cannot really be true
in its most literal meaning, which would be
deterministic and not suitable for social science.
For example, Wealth causes Republican Party
membership. (wealth=Republican) or wealthy
individuals must be Republican!!
Or
Because Iraq has weapons of mass destruction, they
will use them against the USA.
Issue of Causality
 By cause, we usually mean that some variable
AFFECTS another.
 It is best to use language and methods related to
probability theory. How much does an increase in X
affect Y, and how strong is this effect?
The issue of Causality and lack of absolutism of
science are similar.
In the same way that we can not prove theories
to be absolutely true, we can never be
absolutely sure one variable is affecting another.
For this reason, it is more appropriate to say
that a X variable increases or decreases the
likelihood of a change in the Y variable with
some probability, even if very high or low.
Correlative Hypotheses
 Hypotheses may also speculate about whether
two variables are related but not venture a guess
to which affects which.
 This is a question of association where
correlation is useful, but is descriptive and not
relational or explanatory.



Education and wealth are highly related.
Poverty and crime are related.
Democracy and National Wealth are related.
Research Design
 The next few weeks discuss matters of
research design and later the evaluation and
interpretation of results.