Research designs/methods

advertisement
Operationalize the variables
In other words: Define your variable. How exactly will you measure
what you are studying?
Does watching violent television make children act
violently?
Today:
Research Methods
Continued
Turn in extra credit
“Life's most urgent
question is: What are
you doing for others?”
Martin Luther King, Jr.
Steps in the research process
1.
2.
3.
4.
5.
6.
7.
Select a topic
Literature Review
Research design/ Methods
Collect data
Code data
Results and discussion
Inform Others
Each steps has different importance
and significance
Research-Literature Review
Read 10 or more studies from academic sources, explain what has
already been said about your topic, and what you will add to this
body of work
Variables
Does watching violent television make
independent
children act violently?
dependent
Independent variable= cause
Dependent variable = effect
Unethical Studies
View Hidden Camera Show Experiment
Unethical Studies
1931: Dr. Cornelius Rhoads conducted a cancer experiment in Puerto Rico
purposely infecting subjects with cancer cells, 13 subjects died
1946-48: U.S. recently (Oct 2010) apologized for deadly experiments on
hundred of Guatemalans injected with STD’s
U.S. public health service doctors injected Guatemalan patients with
syphilis and gonorrhea without their knowledge, to study the effect of
penicillin as a treatment
1968: Sterilization study, Puerto Rican women were convinced to have
tubal ligation (“tubes tied”)
1/3 of the women in the study were not told the operation was permanent
1932-1972: Tuskegee, Alabama. 399 African-American men were recruited
for clinical study of syphilis Never told they had syphilis, never
treated Given free medical exams and free meals
IRB Approval
Institutional Review Board =
committee at colleges,
hospitals, research institutes
required by law
to ensure research involving
humans is conducted in a
responsible, ethical manner.
Ways of collecting data
• Experimental research
• Surveys
• Existing statistics
• Content Analysis
• Field Research /Participant Observation
Reliability
Reliability in research can be
increased by avoiding
common mistakes
Problems with reliability
Obstacles to reliability include behaviors that can mislead
the researcher:
1.
2.
3.
4.
5.
6.
Spurious correlations
Lies
Evasions
Misinformation
Fronts
Reactivity
1. Spurious correlation
When there incorrectly seems to be a connection
between variables.
A different factor is the actual cause.
Spurious correlations
Correlation is not causation
Just because two
things are occurring
at the same time
does not mean one
necessarily causes the
other
Gallup poll 2011
Correlation is not causation
Students who sit in the front of the class are
more likely to earn an A.
What are potential problems with this
correlation?
Lies
Intended to mislead, give a false
view, or protect identity.
Examples:
A gang member may give false
information about crimes
committed
A club may give inflated
membership numbers to appear
more successful
Misinformation
an unintended falsehood caused by uncertainty and
complexity of life.
Evasions: intentional acts of not revealing
information
This includes:
 not answering questions
 answering a different question
than asked
 switching topics
 answering in a vague manner
Fronts: shared and learned lies and deceptions
Example: a bar that has a true purpose of
providing a place to make illegal bets
Reactivity
Tendency for people being observed to act differently due
to being studied
Questions for surveys and interviews
5 Types of Questions
1. Open-ended
2. Close-ended
3. Double-barreled questions
4. Contingency questions
5. Sleeper questions
Questions for surveys and interviews: Rate your knowledge
1. I recognize this civil rights leader and know his or her name
2. I recognize the leader but can’t remember his or her name
3. I don’t recognize this leader
A.
D.
B.
E.
C.
F.
Raise your hand if: now that you see the leader’s name you
recognize him or her, remember learning about his or her work in school. Or
if: you already knew about the leader even before seeing the name.
Martin Luther King Jr.
Malcom X
Dolores Huerta
Edward G. Tolliver
Cesar Chavez
Mohandas Gandhi
Questions for surveys and interviews
• Sleeper Questions: about non-existent people, events,
or places to determine if respondents are being honest
about knowledge
Edward G. Tolliver
is not a civil rights leader
Program Director at Florida University
Example: in a study to determine which Civil Rights leaders U.S.
adults recognize, the name of a fictitious person was added. 15% of
respondents said they “recognized” the leader.
Sleeper Questions
about non-existent people, events, or places to determine
if respondents are being honest about knowledge
View Sample
Types of questions to avoid
Avoid double-barreled questions
Ex: “Is that class easy and
interesting?”
Why is this questions problematic?
• consist of two or more questions joined together
• makes answer unclear
Types of questions to avoid
Avoid leading/loaded questions
• Wording leads respondent to choose one answer
• Don’t feel all responses are ok
• Makes them aware of answer researcher wants
Example: “You read the whole chapter like you were supposed to,
didn’t you?”
Types of questions to avoid
Loaded questions can be stated to get either positive or
negative answers.
“Should the mayor spend even more money trying to
keep the streets in top shape?”
“Should the mayor fix the potholed and dangerous streets
in our city?”
Avoid qualifying terms “dangerous” “serious”
Threatening and sensitive questions
• Often do not get honest
responses
• Respondents may be:
 Afraid
 Embarrassed
 Unable to confront their own
actions honestly
Threat of questions and
issues from high to low:
1. Sexual behavior
2. Mental health problems
3. Drug or alcohol use while
driving
4. Criminal behavior
5. Income
6. Political association
7. Education
8. Occupation
9. Social activities
10. Hobbies and sports
Survey Questions
Types of questions to re-word:
• Leading/loaded
• Double-barreled
• Threatening
• Sensitive
• Unclear
1.
2.
3.
4.
5.
Work with 1 partner or alone
Read the questions
Decide if they are strong or weak
Re-word the weak questions
Discuss why they are weak
Questions for surveys and interviews
Open-ended question: free to offer any answer
EX: What is one factor of the school environment
that you find frustrating as a teacher?
Questions for surveys and interviews
Closed-ended question: must choose from a fixed
set of answers.
EX: Lack of planning time is a factor of my school
environment that I find frustrating as a teacher.
a.Strongly agree
b.Agree
c.Disagree
d.Strongly Disagree
Questions for surveys and interviews
Contingency Questions: two- (or more) part
question.
Ex: 1. Do you play basketball? _____(If yes, answer
question 1a, if no skip to question 2)
1a. How many days a week do you play basketball?
_____________
2. Do you watch basketball games on TV? _______
Using a sample to collect data
• Instead of gathering data from 20 million people, a
researcher may draw a sample of 2,000
Types of samples
1. Random sample
2. Snowball sampling
3. Quota Sampling
4. Convenience
sampling
Males
Under 18
Ages 1824
Ages 25+
Females
Sampling Errors
Convenience sampling can produce
ineffective, highly unrepresentative
samples
The person-on-the-street interview conducted by television programs
is an example of a convenience sample.
Why is this sampling strategy problematic?
Sampling Errors
• Example: newspaper or magazine that asks readers to
clip a questionnaire from the newspaper
• The number who respond may seem large but the sample
cannot be used to generalize accurately to the population.
Types of Samples
•
•
•
•
•
Cross sectional
Time series
Panel
Cohort
Case study
Next class
Social Change Read Ch 15: p 376-386
Turn in extra credit
Steps in the research process
1. Select a topic, create a research question: specific aims, purpose,
betterment of society?
2. Literature review: useful because it helps you refine your
question, state what you will add to current research
3. Research designs/methods: carefully design research methods
steps to ensure high reliability and ethical methods (consider
sampling strategy and type of questions)
4. Collect data: ethical, organized manner, may need IRB approval
5. Code data: organize your data and look for themes, create
charts/graphs
6. Interpret: analyze, interpret, and discuss your results, include a
discussion of any shortcomings in your methods
7. Inform: share conclusions and recommendations
Download