Transcript Chapter 2 - The Independent School

Psychological Methods
Chapter 2
Section I – Conducting Research
Psychology is an experimental science.
Psychologists (surprise!) use the scientific
method, usually some version of the
following:
Forming a research question
Forming a hypothesis
Testing the hypothesis
Analyzing the results
Drawing conclusions
Replication in Experimentation
As with all experiments, outliers must be
taken into account.
Thus any research needs to be repeated
and to produce the same results.
Different results ? - bad
Different circumstances ? – good
Different participants? – good
Most experiments lead to new questions.
Section 2 – And the survey says…
The Survey Method
Surveys require people to answer
questions about a particular subject.
Can be written questionnaires or person to
person interviews.
The findings of surveys are not necessarily
completely accurate – fears about
confidentiality, desire to “do well” or desire
to deceive are all reasons people may lie.
The Toothpaste Example
 People were given a
survey on toothbrushing habits.
 If their responses
were accurate, three
times as much
toothpaste would
have been sold,
compared with what
was sold.
 Why?
Populations and Samples
Target population: The whole group you
want to study or describe.
Sample: Only a part of the target
population.
“Among women ages 18-26”…
Methods of selecting accurate and
representative sample groups are of prime
importance to psychologists.
Selecting Samples I
 Method I – Random Sample
Individuals are selected by chance from the target
population.
Each member has an equal chance of being selected.
If the random sample is of sufficient size, it can be said
to be representative of the whole population.
Random sample of 1,000-1,500 people is typically
representative of the general American population.
USA Today and many news stations conduct “nonscientific” polls. Do these polls have meaning?
Selecting Samples II
 Method 2 – Stratified
Sample
Subgroups of the
population are
represented in the
sample in a proportional
manner.
US Children 0-18 by
ethnic group in 2000.
Generalizing Results
If a sample is not representative of an
entire population, it can be dangerous to
generalize the results.
If a sample only contains one gender,
nationality or age level, the results of the
survey may not apply to the population.
This class – representative sample?
Example - School presidential elections.
Volunteer Bias
 Volunteer bias: The idea that those who respond
to surveys are different from those who choose
not to respond.
 Volunteers are more likely to disclose personal
information, be more interested in research or
have more spare time compared to others.
 Any of these factors could skew the results of
any given survey.
 Example – mail in surveys.
 What are your survey taking experiences?
Section 3 – Methods of Observation
Method 1 – The Testing Method
Intelligence tests
Measure general learning ability.
Aptitude and skill tests
Vocational tests
Personality Tests
 Myers-Briggs – I-E, S-N, T-F, J-P
How are tests different from surveys?
Method 2 – The Case-Study Method
 An in-depth investigation of an individual or a small
group.
 Freud developed his theories largely on the basis of
case studies – his individual clients.
 Some case studies focus on rare circumstances – Genie
http://www.youtube.com/watch?v=dEnkY2iaKis&list=PL2B8A2ABF83246F22
 Jim Twins
 Case studies are difficult to make generalizations about,
partially because they are difficult to replicate.
 Also can fall into survey pitfalls – incomplete memories,
inaccuracies, wish to do well, lying.
Methods 3/4 – Longitudinal / Cross Sectional
 Select a group of participants and observe those
participants over a period of time, often years or
decades.
 Usually conducted at intervals.
 These studies are extremely time consuming – may take
15-20 years for conclusive results.
 Risky due to continuing participant problems and
financial concern.
 Famous example – The “Up Series”
 John http://www.youtube.com/watch?v=BJj-fc6h0fo
 Sue www.youtube.com/watch?v=w09DaosxTQ&feature=related
 Cross-Sectional: Selecting a group of participants of
different ages, comparing behavior among similar age
groups. Less reliable results than longitudinal studies.
Method 5 – Naturalistic-Observation
 Watching any participants
in their given habitats.
 Used anywhere people
spend time.
 Psychologists do not
interfere with participants.
 Useful for gaining insight
into human and animal
behavior.
 Jane Goodall – Chimps.
Method 6 - Laboratory Observation
Observing behavior in a laboratory setting.
Many laboratories are quite informal – they
are not all men in white coats.
These are small artificial settings which
allow greater control of the environment
Studying how lighting affects eating, or
how noise affects learning.
Analyzing Observations
Once observations have been made,
psychologists look for correlation.
Positive correlation – as one factor goes
up, so does the studied variable.
Negative correlation – as one factor goes
up, the variable decreases.
Correlation does not equal causation.
Section 4 – The Experimental Method
Why Use Experiments?
 Experiments are a method for answering
questions about cause and effect.
 Participants receive some kind of treatment, and
researchers are then more easily able to
determine if the treatment influenced the
participants.
 Limits:
Conditions may not accurately reflect reality.
Often simplify the conditions in order to isolate
cause/effect.
Sometimes difficult to determine validity.
Independent and Dependent Variables
Independent variable = the variable that
researchers manipulate
Dependent variable = the variable that is
expected to change due to the i.v.
Examples:
Mice are injected with different levels of a new drug
to see if it makes them more resistant to cancer.
Humidity is lowered and raised in a room to see how
it changes the concentration level of people inside.
Experimental and Control Groups
 Ideal experiments have both experimental and
control groups.
 Experimental group members receive the
treatment (aka the independent variable).
 Control group members do not receive the
treatment, but are measured anyhow.
 Aside from the presence of the independent
variable, every effort is made to ensure that
conditions are constant for both groups.
Assigning Experimental Groups
 Groups members are assigned randomly to
avoid potential researcher bias.
 Once groups are assigned, researchers typically
study the two groups to make sure they are
approximately the same (i.e. one group is not all
men and the other all women in a study of
consumer preferences).
 When an experiment uses both types of groups
it is called (drumroll please) a controlled
experiment.
The Placebo Effect
“Imagine Yourself Well”
 A placebo is a substance or treatment which has
no effect apart from a person’s belief in it.
 To test new medications, a control group is
sometimes given a sugar pill as a placebo to test
the effectiveness of the medication.
 Sometimes people taking the placebo report
improvement with their symptoms despite the
fact there is no medicine in the pill.
 This can be due to the fact that the person
believes it will treat them.
For example…
 In 1946 a researcher named Jellinek tested a
new headache drug.
 199 headache sufferers took three different drug
combinations and one placebo at different times
for eight weeks.
 At the end of the study, 120 of the participants
responded favorably to the placebo.
 Jellinek concluded that these participants may
have been suffering from psychological
problems and focused his research on the other
79 participants.
…or try this one…
 Seema Assefi and Maryanne Garry, two
psychologists from New Zealand, studied the
effects of alcohol placebos.
 Tests showed that participants who were told
they were drinking vodka (but weren’t) were
more swayed by misleading information and
exhibited some physical signs of intoxication.
 Participants insisted that they had felt drunk at
the time and could not believe that all they had
to drink was tonic water.
Single-Blind Studies
 Single-blind studies help keep participants
unaware of the treatment they are receiving.
 Participants are not told whether they are in the
experimental group or the control group.
 This helps eliminate the influence of
expectations in the experiment.
 However, participants may not be the only ones
with expectations…
Double Blind Studies
 In a double blind study, neither the participants
or the researchers know who is in the
experimental group or the control group.
 This way the expectations of the researchers do
not influence their measurement of data, and the
participants cannot infer which group they may
be in.
 A double-blind study is required by the FDA
before new drugs can be allowed on the market.
Section 5 – Ethical Issues
What are “Psychological Ethics”?
 Ethics are standards for proper and responsible
behavior.
 Psychologists follow ethical standards to
promote individual dignity, foster human welfare,
and maintain scientific integrity.
 The American Psychological Association (APA)
has established several ethical guidelines to
help maintain professionalism and provide
guidance for researchers and experimental
psychologists.
Research With People
 Ethical standards limit the types of research that
psychologists may conduct.
i.e. A study to determine the impact of early
separation from a mother is conducted by
purposefully separating random infants from their
respective mothers.
 The APA provides guidelines on what
requirements must be met in order to make a
study ethical. Two of the most important
guidelines are confidentiality and informed
consent.
Confidentiality
 Due to a respect for
privacy and in an effort to
receive results which are
closer to the truth,
psychologists do not
disclose personal
information.
 On certain rare
circumstances, a
psychologist may break
confidentiality (such as
when threats are made).
Informed Consent
 The APA has restrictions against research that
could pose a serious threat to the physical or
psychological health of participants, or that may
have long-term, irreversible effects on them.
 Informed consent means that people agree to
participate in a study after they have been
briefed on the research.
 The provision of information and the opportunity
to choose whether or not to participate give
people some degree of control and make
participation less stressful.
Deception
 Some research
cannot be done
without some
deception (placebos)
 Three conditions:
Benefits outweigh harm
Individuals might have
participated if they
understood the benefits
A debriefing after the
study has occurred.
Deception in Action…
 In 1961 and 1962, Stanley Milgram conducted a
series of experiments to test the impact of
punishment on learning.
 Men were paired into “teachers” and learners” at
random.
 The teacher read pairs of words to the learner
once, and was then given one of the words and
asked to remember the word it was paired with.
 If the learner got the word wrong, they received
progressively worse electric shocks.
…to Test the Limits of Authority.
 In reality, the teachers were the ones being
experimented with, the learners were in on the
experiment, and no electric shock was given.
 The experiment was designed to test how
people obey authority (i.e. the men in lab coats
running the experiment).
 How many people continued with the experiment
until the highest shock was given, despite cries
of pain, complaints about his heart, and
ultimately silence from the learner?
 www.youtube.com/watch?v=y6GxIuljT3w
Research with Animals
 In the example of children and early separation
from their mothers, studies have been carried
out using animals.
 Most studies that use animals do not harm them
in any way (rats in a maze or a “Skinner box”)
 Sometimes animals are used in place of humans
when harm may be done. For example, studying
sections of a rat’s brain by removing pieces.
 Psychologists use animals only when there is no
alternative and the benefits outweigh the harm.
Examples for Discussion
“I currently have in treatment a psychiatrist
who is still in the midst of a six year affair
with a patient. He wishes to end the affair
but is afraid to face the consequences”.
This example and those that follow come
from a national survey of 1,913 APA
members.
“I am employed as a school psychologist
in a large school district.... I am asked to
provide a diagnosis which will qualify the
examinee for certain services, even if the
test results do not justify that diagnosis. I
have not done so, but the frequency with
which the request is made is troubling”.
One of my clients claimed she was raped;
the police did not believe her and refused
to follow up (because of her mental
history). Another of my clients described
how he raped a woman (the same
woman).
The 1971 Stanford Prison Experiment
http://www.dailymotion.com/video/x2ya1j_
stanford-prison-experiment_events
What ethical problems are evidenced in
this study? Or is it ethical?