Unit 2 Research Methods
The need for Psychological Science
Critical
Thinking
“Smart
thinking”
Four
elements
Examines
assumptions
Discerns
hidden
values
Evaluates
evidence
Assesses
conclusions
Unit 2 Research Methods
Psychology is Empirical. It is based on
research.
Before we delve into how to do research,
you should be aware of common hurdles
that tend to skew our logic.
1. Hindsight Bias
Hindsight bias is the tendency to think that past events
were more predictable than they actually were. Also
known as the “I knew it along”. Did you know I was going
to play this?
http://www.youtube.com/watch?v=-bayV3wez50
Knowing about hindsight bias is useful in two ways:
Firstly, it gives you something to say when your parents accuse you of failing to
predict an event. You can accuse him of having hindsight bias.(Well, it's better
than saying "Yeah, we know that now!")
Imagine that you receive a letter from a publisher that states that the publisher is going to
publish your short story. You tell a friend that you knew that they would publish it.
2. Overconfidence
We like to think we know more than we do
For Example:
Some quizzes, people rate their answers as
"99% certain" but are wrong 40% of the
time.
A person who thinks his sense of direction is
much better than it actually is.
A person who thinks he is much smarter
than he actually is.
Confirmation Bias: (Barnum effect)A tendency
for people to accept information that confirms
what they want to believe
Psych335 - Confirmation Bias - Team 16
http://www.youtube.com/watch?v=fc0tiNGi8jw&feature=related
Hawthorne Effect: Experimenter Bias: Just
knowing you are participating in a study can
change the outcome. Pg. 43
http://www.youtube.com/watch?v=IxZoxN5IjFE
http://www.propagandaposters.us/poster11.html
4. Hawthorne Effect
http://www.youtube.com/watch?v=W7RHjwmVGhs
Hawthorne Effect: Experimenter Bias: Just
knowing you are participating in a study can
change the outcome. Pg. 43
Just the fact that you know you are in an experiment can cause change.
http://www.propagandaposters.us/poster11.html
Why do we have Ethical Guidelines
The next couple of slides are horrible
experiments that explain why it is
necessary.
During WWII the Nazi’s conducted some
very unethical studies. Many of their
subjects died during theses experiments.
What you need to know is:
1. These people were denied the
principles in the Belmont Report
including being asked to participate.
Milgram Study (1963): The Milgram
study involved instructing subjects to
administer electric shocks to a study
confederate in response to poor
performance. The subject believed that
he/she was involved in a study about
learning and memory with each shock
intended to affect the learning process.
The confederate pretended to be hurt by
the shock - in some cases, to the point of
losing consciousness; however, he/she did
not really feel any shock. The study
objective was to assess obedience to
authority. This study resulted in
significant psychological stress for some
subjects including sweating, trembling,
stuttering and serious seizures in three
subjects. The question of whether this
study was ethical remains open to debate
among scholars today.
http://www.youtube.com/watch?v=BcvSNg
0HZwk
Stanford Experiment
A group of men
volunteered for a study
and were given the roles
of prisoners or guards. In
a short time the guards
took it upon themselves to
start trouble with the
prisoners and the
experiment got out of
hand. Lesson in “the
Lucifer” Effect. How
good guys turn bad.
https://www.youtube.com/watch
?v=RKW_MzREPp4
Tuskegee Syphilis Study: In
1932, the Public Health Service enrolled
several hundred syphilitic black males to
document the effects of the untreated
disease over time. Tuskegee was chosen
because approximately 40% of the male
population of the town was infected with
the disease. Treatment was withheld from
study subjects when penicillin was
accepted as the treatment for syphilis in
1943. This study was stopped in 1973 but
not before many subjects became seriously
ill, transmitted their disease to others or
died. This study exemplifies unfair subject
selection practices denial of informed
consent and excessive risk in relation to
study benefits.
Ask me to show you the video located in Unit 2
Videos! I am telling you this because I know I
won’t remember where I put it and will spend
half an hour looking for it. Thank you, you are
very nice people.
http://www.youtube.com/watch?v=j6bmZ8cVB4o
Can I just start any old research project I want to?
No, I need permission from the IRB first.
Institutional Review Board. They make sure
you are following all ethical guidelines. Why do
we have to have this? There have been many
unethical studies done.
When doing a Research study the first step is to get permission
from the IRB:
Protection of Participants Privacy Consent Withdrawal,
Confidentiality Deception Debriefing
Write a brief description of each:
1. Protection from Harm:
2. Debriefing:
3. Privacy:
4. Informed Consent:
5. Deception:
6. Right to Withdrawal:
Animal Research: YES: Otherwise important issues could not be investigated. Relativity little
animal research involves pain or harm.
NO: Animals are entitled to the same rights as humans. Animal studies are
often trivial or may not apply to humans.
Ethics worksheet
Speed theory
http://www.youtu
be.com/watch?v=i
ynzHWwJXaA
Theory:
http://www.youtube.com/watch?v=2lXh2n0aPyw
explanation of some aspect of the natural world that is acquired through
the scientific method, and repeatedly confirmed through
observation and experimentation …Aims to explain
Hypothesis: is a testable prediction that lets
us accept, reject or revise a theory.
If families do not stress gender differences then there will be fewer sex differences in
siblings.
Example: (not the one you
can use). Music influences
concentration in study halls.
The null hypothesis states that there is no
relationship or difference between two
sets of data. When conducting a
psychology experiment, you can either not
reject the null hypothesis (suggesting that
there is no relationship between the
variables) or reject the null hypothesis
(suggesting that there is a relationship
between the variables).
For example, let's suppose that you are
conducting an experiment on the effect of
sleep deprivation on math scores. Your
hypothesis is that students who receive
less than five hours of sleep the night
before a mathematics exam will do worse
than students who sleep for more than
five hours. After performing your
experiment, you find that there is a
statistically significant relationship
between sleep deprivation and math
scores, which means that you can reject
the null hypothesis.
Identify the:
1. Control Group
2. Independent Variable
3. Dependent Variable
4. What should Smithers'
conclusion be?
Smithers thinks that a special
juice will increase the
productivity of workers. He
creates two groups of 50
workers each and assigns
each group the same task (in
this case, they're supposed to
staple a set of papers). Group
A is given the special juice to
drink while they work. Group
B is not given the special
juice. After an hour, Smithers
counts how many stacks of
papers each group has made.
Group A made 1,587 stacks,
Group B made 2,113 stacks.
Identify the7. Control Group
8. Independent Variable
9. Dependent Variable
10. What should Homer's
conclusion be?
Homer notices that his
shower is covered in a
strange green slime. His
friend Barney tells him that
coconut juice will get rid of
the green slime. Homer
decides to check this this out
by spraying half of the
shower with coconut juice.
He sprays the other half of
the shower with water. After
3 days of "treatment" there is
no change in the appearance
of the green slime on either
side of the shower.
Experimental Group: Receives Treatment
Control Group: receives no treatment
Beware of
Confounding Variables
If I wanted to prove that
smoking causes heart
issues, what are some
confounding variables?
The object of an
experiment is to prove
that A causes B.
A confounding variable is
anything that could cause
change in B, that is not A.
Lifestyle and family history may also
effect the heart.
Validity and Reliability
Valid: it is accurate.
The extent to which a test or experiment measures or predicts
what it is supposed to.
Reliable: It can be replicated. Repeating the essence of a
research study.
Get a partner.
Get lego’s, sheet of paper, markers, pencils
You will have 10 minutes
Double Blind vs. Single Blind
When conducting research, it is almost
always impossible to study the entire
population that you are interested in. As a
result, researchers use samples as a way to
gather data. A sample is a subset of the
population being studied. It represents the
larger population and is used to draw
inferences about that population.
six-sigma-material.com
http://mips.stanford.edu/courses/stats_data_analsys/lesson_1/234_0_a.html
Random Sampling
From a population if each
member has an equal
chance of inclusion into a
sample, we call that a
random sample
(unbiased). If the survey
sample is biased, its
results are questionable.
The fastest way to know about the
marble color ratio is to blindly
transfer a few into a smaller jar and
count them.
BAD
Finding the average height of men or women
by using basketball players for your sample.
Give me a few more.
Random Sample: equal chance of being
picked.
Random Assignment: Equal chance of being
assigned to a group.
Placebo: “sugar pill” that is given and the
patient believes to be the real thing.
Operational definition: is one makes it clear how the
Researcher should go about measuring the process,
activity, or thing. EG. Hunger for example might be defined
as “hours without eating” When you word statements
carefully with an operational definition you make it possible
for others to replicate your study.
Let’s say your hypothesis is that chocolate causes violent
behavior.
• What do you mean by chocolate?
• What do you mean by violent behavior?
Methods of
Research
There are 3 categories:
Descriptive, Correlational and Experimental
We like Survey’s because:
1. They are cheap
2. you can get a large amount of information
quickly.
We don’t like Survey’s because:
1. Accuracy depends on the ability and
willingness of the participants.
2. Sampling Bias can skew results
3. Bad questions can corrupt data
Survey
False Consensus Effect
A tendency to overestimate the extent to
which others share our beliefs and
behaviors.
http://www.youtube.com/watch?v=7dJLzXFOC_I
Naturalistic Observation
Good:
Let’s you
observe in a participants
natural setting. There is
Hawthorne effect.
bad
The
is that
we can never really
show cause and
effect.
•Observer can alter
behavior
•Observational Bias
•Cannot be
generalized
•More
accurate than
reports after
the fact
•Behavior is
more natural
http://www.youtube.com/watch?v=fSiWXkOfHBY&feature=related
Research project that investigates the degree to which two
variables are related to each other.
Does NOT say that one variable causes another.
There is a positive correlation
between ice cream and murder rates.
Does that mean that ice cream causes
murder?
Correlation does NOT
mean Causation!!
Good: Determines relationship between 2 variables. Predicts future behavior.
Bad: Will uncover a relationship but that does not mean it is the cause
https://www.youtube.com/watch?v=8B271L3NtAw
Correlations also vary in the strength of the
association
Zero correlation: no relationship between the 2
variables
Strong correlation: knowing the value of one
variable permits one to accurately estimate the
value of the other variable
Strong correlation can be positive or negative
Correlations can be seen in scatter plots
© 2004 John Wiley & Sons,
Correlational Research
The correlation coefficient a
statistical measure that indicates
the degree of association between
2 variables
Correlations vary in direction:
Positive association: increases in the value
of variable 1 are associated with increases
in the value of variable 2
Negative association: increases in the
value of variable 1 are associated with
decreases in the value of variable 2
No relation: values of variable 1 are not
related to variable 2 values
© 2004 John Wiley & Sons,
How to Read a Correlation Coefficient
Redelmeier and Tversky (1996) assessed 18 arthritis
patients over 15 months, while also taking
comprehensive meteorological data. Virtually all of the
patients were certain that their condition was
correlated with the weather.
In fact the actual correlation was close to zero.
Ice cream sales and the number of
shark attacks on swimmers are
correlated.
• Skirt lengths and stock prices are highly
correlated (as stock prices go up, skirt
lengths get shorter).
• The number of cavities in elementary
school children and vocabulary size are
strongly correlated.
Cross-Sectional Studies:
Data is collected from groups of individuals of
different ages and compared.
Advantages
Disadvantages
data on many
variables
increased
chances of error
data from a large increased cost
number of
with more
subjects
subjects
http://www.youtube.com/watch?v=6EjJsPylEOY
Done in a lab.
Good: You have control over your environment, can determine the
cause and effect of an experiment.
Bad: You don’t see behaviors in their natural surroundings.
Case Studies
An intense study of a person or group. Diaries, Tests, and
interviews.
Pros: Rich description of
subject, easy to control
Cons: Observer bias, difficult
to summarize subject's
experience
Data is taken from a group over a period of time.
http://www.google.com/imgres?imgurl=http://www.cfr.nichd.nih.gov/images/children_linedup2.jpg&imgrefurl=http://www.cfr.nichd. nih.gov/longitudinal.html&usg=__w_lHT2-TFLy8l_gLnyOF8W7ctE=&h=364&w=589&sz=68&hl=en&start=0&zoom=1&tbnid=J01pJhEZ4AmeM:&tbnh=99&tbnw=160&prev=/images%3Fq%3Dlongitudinal%2Bstudies%26um%3D1%26hl%3Den%26sa%3DN%26rlz%3D1T4ADRA_enUS376US377%26biw %3D1899%26bih%3D922%26tbs%3Disch:1&um=1&it
bs=1&iact=rc&dur=334&ei=OheETPijDIn4swPDutH2Bw&oei=OheETPijDIn4swPDutH2Bw&esq=1&page=1&ndsp=37&ved=1t:429,r:10,s:0&tx=61&ty=59
Advantages
Disadvantages
data easy to collect
data collection method may change over time
easy to present in graphs
difficult to show more than one variable at a time
easy to interpret
needs qualitative research to explain fluctuations
can forecast short term trends
assumes present trends will continue unchanged
How can math and psychology be related?
Just describes sets
of data.
You might create a
frequency distribution.
Frequency polygons or
histograms.
A branch of math that summarizes and makes meaningful
inferences from the data.
http://www.blinkx.com/watch-video/amazing-statistics/ERt_t-mY1EwAypRxeXdO_Q
Frequency is how often something occurs.
Frequency Distributions: list of scores from
highest to lowest. What type of graph would
you use to show a frequency distribution?
Typically, a bar graph.
Let’s make one:
Scores from voc. Quiz.
30, 54, 27, 46, 38, 42, 55, 51, 55, 42, 38, 42,31, 55
Put the numbers in order,
then added up:
•how often 1 occurs (2 times),
•how often 2 occurs (5 times),
•etc,
and wrote them down as a
Frequency Distribution table
What can you see?
 Mode
 the most frequently occurring score in a
distribution
 Mean
 the arithmetic average of a distribution
 obtained by adding the scores and then dividing by
the number of scores
 Median
 the middle score in a distribution
 half the scores are above it and half are below it
Which is best (mean, median, or mode)?
It would not be a good idea to use the mean to report the central tendency for
housing costs in a community because most communities have a few very
valuable homes. When you calculate a mean, these few expensive homes will
affect the mean much more than each of the moderately priced homes will. As a
result, housing will appear to be more expensive than it really is.
Another time when we usually prefer the
median over the mean (or mode) is when our
data is skewed (i.e., the frequency
distribution for our data is skewed).
as the data becomes skewed the mean loses its ability to provide the best
central location for the data because the skewed data is dragging it away from
the typical value
Measures of Central Tendency
http://www.youtube.com/watch?v=VuRyosm3t-s
• Mean is generally
used, unless extreme
values (outliers) exist
— then median is
often used, since the
median is not
sensitive to extreme
values.
Team I has range 6 inches, Team II has range 17 inches.
Disadvantages of the Range
•Ignores the way in which data are distributed
•Only uses two entries from the data set
•Sensitive to outliers
.
http://rchsbowman.wordpress.com/2010/09/01/statistics-notes-%E2%80%94-measures-of-variation/
Skewed Distributions
A skewed distribution represents a set of scores or numbers
that is not equal on both sides.
ttp://www.sophia.org/identifying-positive-skew/identifying-positive-skew-tutorial?topic=measures-of-shape
• Outliers skew
distributions.
• If group has one high
score, the curve has a
positive skew (contains
more low scores)
• If a group has a low
outlier, the curve has a
negative skew
(contains more high
scores)
Statistical Terms
 Range
 the difference between the highest and lowest
scores in a distribution
 Standard Deviation
 a computed measure of how much scores vary
around the mean
 Statistical Significance
 a statistical statement of how likely it is that an
obtained result occurred by chance
Its symbol is σ
The formula is easy: it is the square root of the Variance.
So now you ask, "What is the Variance?"
Variance
The Variance is defined as:
The average of the squared differences from the Mean.
To calculate the variance follow these steps:
Work out the Mean (the simple average of the numbers)
Then for each number: subtract the Mean and square the
result (the squared difference). Then work out the
average of those squared differences. (Why Square?)
http://www.youtube.com/watch?v=pFGcMIL2NVo
Normal Distribution
http://www.youtube.com/watch?v=xgQhefFOXrM&feature=related
Z score
A unit that measures the distance of one score
from the mean.
•A positive z score means a
number above the mean.
•A negative z score
means a number below
the mean.
“Inferential Statistics”
http://www.youtube.com/watch?v=oHGr0M3TIcA
Inferential Statistics
p value = likelihood that
results are a fluke or
coincidental
Which should you trust
more, results with a
low or high p value?
How low?
If p < 0.05, then the results
are “statistically
significant”.
Statistically significant – not
likely due to random chance
Infer your data get it infer
hahahahaha
Probability and Significance:
Probability, or p, is expressed as a number between 0
and 1. 0 means an event will not happen, 1 means
that an event will definitely happen. The P value will
always be found to be between 0 and 1 due to the
way in which it is calculated. To calculate the
probability that a particular outcome will occur, it has
to be divided by the number of possible outcomes.
One way to work out the probability of something
occurring is to use this formula:
Probability = number of particular outcomes
number of possible outcomes
http://www.youtube.com/watch?v=BE54mDs6St4