Psychology of Personality

1
Psychology of Personality Research Methods
Katherine Aumer-Ryan Lecture 2
2
Personality Psychology as a unique science
--Every assumption should be and often is
questioned
--Methodology and Statistical use of primary
importance
--The primary purpose of science is to understand
what is known and to discover/explore what is yet
to be known. NOT a technical or trade school
lesson.
3
The Data
--3 important qualities of data
--Reliability
--Validity
--Generalizability
--Can you guess what they mean?
How can someones argument be valid?
What does it mean to like people, in general?
How is a car reliable?
4
To be concerned about these qualities of data, we
need to have collected data in order to answer a
research question.
--First, we have to want to measure something.
For example How friendly are the people in this
class? Who is the most conscientious? How
emotionally stable are the people in the front
row?
--What we want to measure is a trait. How is
this different from a state?
--What circumstances are relevant when we want to
measure a state versus a trait?
5
--State v. Trait
--Lets say we are trying to measure general
happiness amongst UH students.
--Research question How happy are students at
UH?
--What kind of data do we use use the S in the
BLIS
--When should we ask students?
--Right after a football game where UH won?
--Only in the mornings?
--After the Ka Leo has announced increases in UH
tuition?
--At the end of semester, before or after finals?
6
--State v. Trait
--States are temporary and passing (e.g., bad
mood, hyper)
--Traits are stable and do not change much over
time (e.g., conscientious, open, or extraverted)
--importantly traits, unlike states, are what
make us unique. We each differ on various
traits, which allows us to be different from one
another.
--Once we have the research question, trait
defined, and know what kind of data we want to
collect we have to ensure the data are reliable,
valid, and generalizable.
7
What is reliable data?
--Scientifically reliable data is free of (or
minimally impacted by) measurement error.
--What is measurement error?
8
Things that can make our measurement of students
happiness unreliable
What we as researchers can do to prevent this
--Being especially careful. Double/triple check
--Low precision
--State of the participant
--Not too much you can do. Try to sample from
random times.
--State of the experimenter
--Make sure each experimenter gets the same
training
--Environment
--Standardize the environment as much as you can.
Same room, same temperature
9
Aggregation
--Probably the best thing one can do to get a
reliable result.
--Requires sampling same person, over and over
again.
--Relies on the principle that Random effects
sum to zero.
--Can require taking the same test over and over
again, or having similar questions on the same
test be asked over and over again.
10
--Validity
--the degree to which a measure actually reflects
what one thinks or hope it does.
--Construct Validity the degree to which a
measure (e.g., a psychological test) actually
reflects the components of a psychological
construct (e.g., intelligence, sociability,
conservatism, passionate love, racism)
--How would our measure of UH Students Happiness
show construct validity?
11
Generalizability
--How do these data apply, pertain, be indicative
of, other populations, people, time, place
--Encompasses BOTH reliability and validity.
--Current issues of generalizability with
psychology studies
College students
Gender
Participation
Cohort
Race and Culture
12
Results of Your Graphology tests
--First look at only your results. Think about
how well they pertain to you.
--Fill out the survey again.
--Answer an additional question How
accurate is my personality profile? 1 2 3 4 5 6 7
8 9 Not at all Completely
13
Graphology
Questions or Comments?
--What do you think of Graphology?
--What do you think of your results?
Forer (Barnum) Effect
--Personal Validation Effect a very general,
vague statement about ones personality that
actually applies to everyone.
--Practiced by astrology, palm reading, fortune
telling, tarot card reading
--Used by pseudoscience.
14
How Scientific are these disciplines?
--Your pretest results
--Do you think your posttest scores will differ?
15
Break time
16
--Research Design
--In order to gather data, one needs to have an
approach, specifically a design.
--3 Main types of Research Design
--Case Study single, isolated example.
--Experimental manipulation of variable.
--Correlational assesses relationships, no
manipulation.
17
--Case Studies or Method
--Involves very close study of a single incident,
person, or phenomenon.
--For example
You (introspection)
Any person
Natural Disaster
Rare cases, like with PDs
Rare people
18
Pros
Cons
Case Studies
-Lots of information
-No controlvariables going everywhere!
-Generate theory, ideas, prepare for future
-Can you really generalize?
-Cant get the information any other way
19
Experimental
--Random assignment of participants, variable is
manipulated, and outcome is measured.
20
Pros
Cons
Case Studies
-Lots of information
-No controlvariables going everywhere!
-Generate theory, ideas, prepare for future
-Can you really generalize?
-Cant get the information any other way
-May create artificial groups.
Experimental
-Infer causality
-Can make theoretical and pragmatic claims
-Unethical (e.g., deception) or impossible.
21
Correlational
--Peoples emotions, behaviors, and/or cognitions
are surveyed and a relationship is inferred. No
manipulation of variables.
22
Pros
Cons
Case Studies
-Lots of information
-No controlvariables going everywhere!
-Generate theory, ideas, prepare for future
-Can you really generalize?
-Cant get the information any other way
-May create artificial groups.
Experimental
-Infer causality
-Can make theoretical and pragmatic claims
-Unethical (e.g., deception) or impossible.
-Relatively easy to implement
Correlational
-Cannot infer causality
-Third variable problem
-Can be more realistic--sometimes
23
--How do we know we have a significant result?
--Significance testing what is the probability
that my data was due to chance? (plt.05)
--Correlations how strong is the relationship
between two variables? ( or - )
24
--In any research design there is the probability
for error
--Type I
Deciding there is an effect when none exists.
--Type II
Deciding there is NOT an effect when one DOES
exist.
--Ethics
--Generalizability is your study applicable to
many people? If only to a few, shouldnt you
just study those few or have a representative
sample.
--Truthfulness dont make up data.
--Deception is it needed? Knowingly telling a
lie?
25
--Break and video
26
--Class Overview
--What are your predictions for how scientific
is the field
--Get on facebook and join the class group
--Chapter 3 First part
--What are the three qualities of data?
--Difference between State and Trait?
--Aggregation can increase
--Chapter 3 Second part
--3 main types of research design?
--Type I and II errors?
--Ethical problems? (name 3)