Research Methods Flashcards
Most people think about human abilities in one of two ways:
-Fixed mindset
-Growth mindset
Fixed mindset
Human abilities and attributes,
such as intelligence or personality, are fixed
and cannot be changed
Growth Mindset
Abilities are malleable and can
be developed—through practice, feedback, and effort
Fixed Mindset
Our genes give us
different strengths and weaknesses
Growth Mindset
Our brains are highly
adaptable and can change
When people have a fixed mindset:
-Lower effort and willingness to approach
challenges
(believe if one is not smart and have to work hard, how smart can you really be (mindset))
-Negative responses to setbacks
(believe one can never improve)
-Worse academic performance
When people have a growth mindset:
-Higher effort and challenge seeking (believe one can grow and learn)
-Persistence in the face of obstacles and
failures (learning from mistakes)
-process errors more deeply at a
neurological level
-Better academic performance
True or false:
1.To change people’s behaviour toward members of other racial groups, we must first change their attitudes.
- The best way to ensure that a desired behaviour will persist is to reward the behaviour every time it occurs.
- Fortunately for babies, human beings have a strong maternal instinct.
- By giving a young baby lots of extra stimulation (e.g., mobiles and musical toys), we can markedly increase its
intelligence. - In love and friendship, more often than not, opposites attract one another
All false!
Hindsight Bias and Overconfidence
-The tendency to believe, after learning an outcome, that one would have foreseen it
- “I-knew-it-all-along” phenomenon
Examples:
-Monday morning quarterbacking
-Predicting tragedies
-Reviewing multiple choice tests
-We become overconfident in our power of deduction – intuition
-Looking at things scientifically can help separate reality from
illusion
The Scientific Method
-Observations
-Development of theories
-generate hypothesis
-design a research study
-collect evidence (data)
Theory
Set of principles that explain, organize, and predict events
− What is known about this behaviour already?
Theory of
Catharsis
-Negative feelings build up
and create pressure if not
vented
-If this tension continues to
build, people will ’explode’
-Releasing emotions
decreases the pressure or
tension
Hypothesis
A testable prediction that is based on (derived from) your theory
-What evidence would confirm my theory? Disprove it?
People who express their
anger will behave less
aggressively than those
who do not express their
anger is an example of a
hypothesis
Operational definitions
A clear statement of the procedures (operations) used to define research variables in a study
-Describes how we will measure our variables
Operational Definitions Example
Variable: Express anger
Operational definition: Punch a punching bag
(or other physical actions or increased volume)
Operational Definitions Example:
Which one is the variable and the operational definition?
-aggression
-intensity of noise blasts
Variable: aggression
Operational definition: intensity of noise blasts
(or aggressiveness, tone of voice)
Descriptive approaches (3 types)
(describes the characteristics of something)
-Surveys / Questionnaires (self-report)
-Naturalistic Observation
-Case Studies
These methods describe behaviour - they cannot explain it
Naturalistic observation
-Observe and record behavior as it naturally occurs
-No attempt to control the situation
-e.g. observing a couple fight instead of asking them how they fight
-Observation technique in which one person is studied in depth in the hope of revealing universal principles
-Issues?
Case studies
(an in depth exploration of one person to draw conclusions about them)
-issue is that things often cannot be generalized
Correlational research
-Measuring the degree to which two variables are related or predict one another (co-related)
⎻ Positive correlation: The variables tend to occur together
⎻ Negative correlation: The variables tend to occur at opposite times
e.g., Number of hours of sleep and number of arguments with romantic partner
is an example of
correlational research
Correlational research
The variables tend to occur together
Positive correlation (high levels of one correlates with high levels of the other)
Correlational research
The variables tend to occur at opposite times
Negative correlation (high levels of one are correlated with low levels of the other)
Correlational research
Correlation Coefficient
is a statistical measure of the relationship
between two variables
Correlation Coefficient
r = + 0.37
r = + 0.37
Correlation Coefficient:
+ indicates direction of relationship (positive or negative correlation)
0.37 (magnitude) indicates strength of relationship
0 = no relationship
+/-1 = strong
-1 = stronger negative correlation
close to 1 = high correlation where 1.0 is a perfect correlation
Correlation Example:
Sleep Deprivation and Arguing
What does a r= 1.0, 0,0, or -1.0 graph look like? (the x axis is the number of arguments and the y axis is the hours of sleep)
1.0: straight diagonal line pointing to the right direction
0.0: points are scattered
-1.0: straight diagonal line pointing to the left direction
The problem with correlation:
Should ice cream be blamed for murder?
-some things that occur together are not always a causal effect
-correlation does not always mean it is causing the other
-does not mean ice cream causes murder just because more ice cream sales just happen at a time where there are higher rates of homicide
-reasons could be that both events happen in the summer (more heat makes people more likely to cause conflict or be violent, but also people buy ice cream in hot weather)
Correlation is not Causation
of
Any combination of 3 things could be happening….
1. Sleep deprivation could cause arguing (X causes Y)
2. Arguing could cause sleep deprivation (Y causes X)
3. Third variable (e.g. stress) could cause both (Z causes X and Y)
Requirements for a causal relationship
-relationship
-causal variable happens first
-should not be a 3rd variable
Correlation is not Causation
More Examples:
-The more electrical appliances people have in their homes, the more likely they to use birth control
-Couples who decorate the Christmas tree together are more satisfied with their relationship
-Children who are read to as newborns do better in elementary school
-Adults who regularly do crossword puzzles have better memories
Experimental Research
-A means of determining the cause of events
-If we think one variable causes changes in another, we manipulate that variable to its effects
-Expose people to either high or low levels of Variable X
-Measure change in Variable Y
there is a dependant variable and a dependant variable
Independent Variable (IV):
Factor that we think causes
something else to happen. We
control (manipulate) the level of
this variable.
e.g., Sleep deprivation
Dependent Variable (DV):
Outcome that we think depends
on the level of the IV. We measure
the level of this variable.
e.g., Conflict
What problem does manipulation of the independent variable do and the random assignment to conditions?
-Manipulation of IV solves directionality problem (which came first?)
-Random assignment to conditions solves 3rd variable problem
Experimental Research
-Independent variable (random assignment to conditions) or dependent variable?:
-High/low sleep deprivation
-Average # of arguments
-Independent variable (random assignment to conditions): High sleep deprivation and Low sleep deprivation
-Dependent variable:
Average # of arguments
Choosing a Method
Why not always use an experiment?
-not ethical to test some variables, such as binge eating or sleep deprivation
-not always possible to randomly assign
-people will know a variable is manipulated
-time intensive
-experiments are costly
-hard to collect large sample
-not interested in causation (depends on research goals)
What Makes ‘Good’ Research? (5 requirements)
-objective measurement
-generalizable
-unbiased
-made public
-replicated
What Makes ‘Good’ Research? (5 requirements)
Objective measurement
measures are reliable (consistent) and valid
What Makes ‘Good’ Research? (5 requirements)
Generalizable
-Participants are representative of the population
-Procedure is relevant outside the laboratory
What Makes ‘Good’ Research? (5 requirements)
Unbiased
-Researchers do not influence the results
-Participants do not intentionally influence the results
What Makes ‘Good’ Research? (5 requirements)
Made Public
-Methods and results are disseminated
-Methods and results are peer reviewed
What Makes ‘Good’ Research? (5 requirements)
Replicated
-Studies are repeated (directly or conceptually)
-Similar results are obtained
Reliability
-A measure’s ability to produce consistent results
⎻ Can we count on it to behave consistently?
-Test-retest reliability
-Interrater reliability
Reliability
Test-retest reliability
Should be getting same results/scores always
Reliability
Interrator reliability
Person doing the measurement should not affect the results
-e.g. grading essays are subjective and not reliable
Construct Validity
The extent to which a test/assessment actually measures what it says it does
-e.g. Graduate Record Exam – determines aptitude and expected success in graduate school
Explanation:
-the GRE is not a good way to predict how one does in grad school
-same score does not mean accurate, consistency does not mean accuracy
Generalizability (External Validity)
-A population is everyone/everything we are interested in
-A sample is a limited number of people from the population
-Does our sample accurately represent the population of interest?
⎻ WEIRD samples
WEIRD” is an acronym coined by Henrich, Heine, and Norenzayan: samples that are drawn from populations that are White, Educated, Industrialized, Rich, and Democratic
Generalizability (External Validity)
are the experiment’s procedures relevant to the outside world?
Examples:
-memory
-remembering a list of words that start with a
-remember a series of grocery lists
-remember a stream of faces
