Document 7868817

Download Report

Transcript Document 7868817

研究设计:基本知识
School of Public Administration & Policy
Dr. Kaifeng Yang

When I think about a scientist, I think of ____

If I were going to be a scientist, I should like to be the kind of scientist who ___

If I were going to be a scientist, I should not like to be the kind of scientist who ___

When I think of a social scientist, I think of ___

Is Public Administration a science?
Figure 1. A Disciplinary Continuum: From Science to Art.
Physics Chemistry Biology Economics Psychology Political Science Sociology Anthropology Humanities Fine Arts
Science
Source: Adapted from Dempster, Beth, "Toward a Post-Normal Science: New (?) Approaches to Research," May 1998.
http://bethd.ca/webs/pnsresearch/index.html, date accessed September 4, 2004.
“Scientific” Inquiry: Characteristics







Empirical verification
Non-normative
Transmissible
General
Explanatory
Probabilistic
Provisional







Thou shall not kill
Al Gore received more popular votes in the 2000
presidential election than G.W.Bush
The death penalty deters crime
All people are created equal
There will always be some people living in
poverty no matter how hard government tries to
eliminate it
Catholics are more likely to vote than Protestants
People in the Middle East would be far better off
if they lived under democratic governments
Scientific Reasoning

Induction
–
–
–
–

Bottom up
Probabilistic and uncertainty
Degree of match
Validity and generalizability
Deduction
– Top down
– Valid or invalid
– Alternative theories or biased evidence
研究过程
Three main concerns as the researcher narrows in on
researchable questions:
1. What entities (individual, people, groups, formal
organizations, nations) are to be studied;
2. What aspects or characteristics of these entities are of
interest;
3. What kinds of relationships among the characteristics
are anticipated.
Units of Analysis

The entities (objects or events) under study
 Include individual people, social groupings, social artifacts

The research is to compare the units of analysis; The
data or the variable is about the units of analysis

It is important to identify accurately the unit of analysis
since confusion over units may result in false conclusions
about research findings
Unit of Analysis

The entity about whom or which the researcher
gathers information.
– Not:
The entity from which the researcher gathers
–
information.
– Rather: the information about what?

It is the analysis you do in your study that
determines what the unit is.
– The unit is simply what or who to be described or
analyzed
Units of Analysis Examples

If you are comparing the children in two
classrooms on achievement test scores, the unit is
the individual child because you have a score for
each child.

On the other hand, if you are comparing the two
classes on classroom climate, your unit of analysis
is the classroom, because you only have a score
for the class as a whole and not for each individual
student.
Related issues

The use of aggregate data: combining information about
individuals to describe the social unit to which they
belong.
 U.S. News rankings, individual students- college as a whole.

Ecological fallacy
This occurs when relationships between properties of
groups or geographic areas are used to make inference
about the individual behaviors of the people within those
groups or areas.

Misleading to draw conclusions about individual-level
processes from aggregate or group-level data.
Conceptualization
The process … we specify what we mean when we
use particular terms. It involves describing the
essential features of the terms. Sometimes, it involves
indicators we’ll be using to measure our concept and
the different aspects of the concept (dimensions).

Indicator: a sign of the presence or absence of the concept
 Dimension: a specifiable aspect of a concept
How to Conceptualize Religiosity?
A religious person:
..…….
A non-religious person
….. …..
A Template:
The concept of _____is defined as the extent to which
____exhibit the characteristic of _________________
The concept of religiosity is defined as the extent to which
individuals exhibit the characteristic of attending religious
services.
Variables
There are two types of explanatory variables: dependent and
independent
1) Dependent variable (DV)
• the one the researcher is interested in explaining and
predicting
• variation in the dependent variable is thought to depend
on or to be influenced by certain other variables
2) Independent variable (IV)
• independent variable refers to variables that influence or
explain the dependent variable
• if we think in terms of cause and effect, the IV is the
presumed cause and the DV the presumed effect.
IV & DV






Presumed cause
Stimulus
Predicted from…
Antecedent
Manipulated
Predictor






Presumed effect
Response
Predicted to…
Consequence
Measured outcome
Criterion
Variables (4)
3) Antecedent variable
• the one that occurs prior in time to both independent and dependent
variable
Antecedent
Independent
Dependent
4) Intervening variable
• the one that is an effect of the IV and a cause of the DV
Independent
Intervening
Dependent
Variables
5) Control variable
• the one that is held constant, or prevented from varying,
during the course of observation or analysis to limit the focus
of the research or to test hypotheses pertaining to specific
subgroups
• a means of ruling out variables that are not of immediate
interest but that might otherwise explain part of the
phenomenon that the investigator wishes to understand
6) Quantitative vs. Qualitative variable
• quantitative- if values or categories consist of numbers
• qualitative- has discrete categories, usually designed by
words or labels
Relationships
Features of relationships in social science:
1) they always involve two or more entities.
2) the pairs or combinations of things usually occur
and change together.
• the manner in which the variables change or
vary together will depend on whether the
variables are qualitative or quantitative.
3) four properties of relationships: strength,
directionality, linearity, and statistical significance.
Relationships
Relationships among quantitative variables
•
we can measure not only the strength of the relationship but also two
other aspects of relationship: direction and linearity
• positive (direct) vs. negative (inverse) relationship
• linear vs. non-linear relationship
Variable Y
Linear Relationships
Variable X
Variable Y
Curvilinear Relationships
Variable X
Relationships
Measure of strength and direction of relationship
correlation coefficient (r)
• a common statistical measure of the strength and
direction of linear relationships between two
quantitative variables
• ranging from -1.00 to +1.00
• the signs indicate the direction of the relationship
• the magnitude of the coefficient, ignoring the sign,
indicates the strength of association
Relationships
Causal relationships in social science
1) association
• variables must be associated in a statistical sense
• meaningful causal relationship, not necessarily perfect association
2) direction of influence
• a cause must precede its effect
• at least direction of influence should be from cause to effect
3) non-spuriousness (elimination of rival hypotheses)
• spurious relationship occurs when a correlation has been produced by
a third factor and neither of the variables involved in the correlation
has influenced the other.
Stating Problems and Hypotheses
Hypotheses
• Tentative answers to research questions.
• More formally, expected but unconfirmed relationship
between two or more variables.
• While stated in a variety of ways, all hypotheses should
speculate about the nature and form of relationship.
• Include two aspects:
1) which variable causes, explains, or predicts the other
2) how changes in one variable are related to changes in
the other
Stating Problems and Hypotheses
Forms of expressing testable hypotheses
1) If-then (conditional) statements
2) Mathematical statements
3) Continuous statements
4) Difference statements
It is important to know which variable is presumed to cause the other
How a hypothesis is expressed in a given study will depend on several
factors
• Researcher’s discretion, the present state of knowledge about
research problem, and whether qualitative or quantitative variables
are involved.
• Regardless the form of expression, it should indicate at least the form
of the relationship between variables and, ideally, should specify the
causal linkage between variables.
Hypothesis Templates

In comparing [units of analysis], those having [one value
on the independent variable] will be more likely to have
[one value on the dependent variable] than will those
having [a different value on the independent variable].
E.g.: In comparing [individuals], those who [are women]
will be more likely to [favor a handgun ban] than those
[who are men].

Another Template: the more/less…the more/less
E.g.: The more intelligent a person is, the more likely he or
she is to support civil liberties.
Pitfalls

Affirmative vs. question
– Do democratic nations provoke wars?
– Democracies are less likely to provoke wars with other
nations than dictatorships

Evaluative statement ×
– Democracy is the best form of government

Comparative value statement ×
– Swedes are less moral than Americans

Prescriptive Value Statement ×
– The campaign finance laws should be changed
Pitfalls

Compound Hypotheses ×
– The poor and alienated are unlikely to vote

Tautology ×
– Liberal members of the House of Representatives
receive high ratings from liberal interest groups

Narrow Hypotheses ×
– Northern Ireland has experienced conflict because of
religious differences
How about These?





In comparing individuals, some people are more likely to
donate money to political candidates than are other people.
In comparing individuals, gender and abortion attitudes are
related.
Because of important cultural changes that began in the
1960s, many current political conflicts are based on
generational differences.
Should people support gun-control?
Are there more democrats than republicans in Florida?
…must transform questions into hypotheses

Do people’s attitudes toward the environment vary
by income?
– People with higher incomes have stronger pro-
environment attitudes than people with lower incomes
– The lower one’s income, the more one cares about the
environment
– Peoples’ environment attitudes are not related to their
income
Group Practice:
Identifying variables and UOA

Tone of campaign advertising
 closeness of election
 Voter turnout

Interest in politics
 Perceived difference b/t major parties
 Turning out to vote
Group Practice:
Evaluating the articles





For the chosen article
Why do you like it most?
Identify IV, DV
Identify unit of analysis and evaluate whether it is
appropriate
Identify hypotheses and evaluate whether they are
well-written and make sense
Groups’ Design Assignment
主题领域







责任政府
绩效管理
电子政府
行政伦理
反腐败
社区治理
公民参与
战略管理
 财政与预算
