Q2: Research Methods Flashcards
Hindsight Bias
“i knew it all along” “hindsight is 20/20” ex looking at an answer to a question and thinking you actually knew
Confirmation Bias
looking for info to confirm an already held belief /not actively looking for data to refute your hypothesis. (ex numbers doubling pattern example from class)
Overconfidence
overestimating what you know.
operational definitions
define concepts in terms of procedures used to measure or create them (ex when judging how many people smile in photos… what is considered a smile?)
Naturalistic Obseravtion
Observing what would happen even if we weren’t there.
-No bias of the person recorded.
-Hard to tell things sometimes
Case Study
Studying a situation in which it would be unethical for us to create our own version.
-Can’t make generalizations.
Survey
Asking lots of people questions.
-Volunteer bias
-Representative sample
Experiments
Involve the manipulation of variables to determine their effects on measured variables.
-Establish cause and effect.
-Careful control is needed when designing experiments to not compromise results
-IV + DV
Cause and Effect
Established by experiments
Independent Variable (IV)
controlled by the experimenter; the “cause” variable.
Dependent Variable (DV)
the outcome measured by the experimenter; the “effect” variable.
Experimental Group
exposed to IV (ex students given cookies before tests)
Control Group
not exposed to IV (ex students without cookies before tests)
Within Group Design
subjects serve as their own control (all are exposed to the IV)
Between Group Design
subjects are split into experimental and control groups
Confounding Variables (extraneous variables)
-Environment (keep consistent)
-Expectations (blind procedure)
-Individual difference (randomly assign subjects to groups so diffs have same avg. impact)
-Double Blinding (neither researcher or subject know what they’re getting, like placebo or medicine)
Random Sampling
to select participants from population.
allows you to generalize results (ex people will do better after having cookies before a test)
Random Assignment
to divide participants into groups.
controls individual differences/confounding variables.
effect size
measure of how large an impact the IV has on the DV (is this just a coincidence or is it factual?)
statistically significant
results are unlikely due to chance / results are likely due to the change in IV.
p ≤ 0.05
p = likelihood it is chance
Correlational Research
involves the use of statistical methods to reveal and describe the relationship between two variables. (correlation coefficient)
-variables are not manipulated by researcher
-does not establish cause + effect (may suggest it though)
Correlation Coefficient
the degree to which variables are statistically associated is expressed as a correlation coefficient. demonstrates how powerful a variable is as the other changes. (R value)
R value
(correlation coefficient) varies from -1.00 to +1.00 (it is a calculation error if >1 or <-1) as the value gets closer to 1 or -1 the stronger the relationship between variables is. if the value is near zero there is little to no relationship.
positive correlation
r > 0
as one variable increases so does the other
negative/inverse correlation
r < 0
as one variable increases the other decreases