Chapter 2 Flashcards
self-reporting questionnaire
most common way to measure personality
personality assessment
asks people to report their personalities, use adjectives or agreement with statements (Likert scale)
what is the largest negative about personality assessments
prone to socially desirable responding
how accurate are people at reporting own personalities
it depends on the person
response bias
responding in inaccurate ways, not always conscious, personalities can influence
social desirability bias
people self report in better light, more positives and less negatives
agreement/acquiescence bias
select positive responses all the time, reverse-score to prevent
reverse-scored
statements worded in the opposite direction
acquiescence
agree to everything
correlation (r)
measures statistical relationship between two variables, positive or negative
null correlation
two variables are not related (not statistically significant), correlation of zero
statistical significance
when probability of happen by random chance is less that 5%
what does reliability tell us?
if a scale is a reliable, gives consistent result each time or consistently measuring the same thing
internal reliability
when all items on scale report same concept, use Cronbach’s alpha statistic, 0.60 or more
Cronbach’s alpha statistic
average correlation between items
test-retest reliability
taking test at two different times produces similar results, at least 0.70, shows makes same results over time
intercoder reliability
when assessment involves writing samples/behaviors, indicates correlation between coders’ ratings
validity
measures what it is supposed to measure
face validity
items appears to measure what they are supposed to
predictive validity
measure is related to concrete behavior or outcome
convergent validity
scale correlates with similar scales
discriminant validity
scale doesn’t correlate with unrelated scales
Horoscope example with validity and reliability
sign is consistent (reliable) but not valid because does not correlate well, discriminant and convergent validity
Barnum’s effect
people’s tendency to believe vague, positive statements about themselves
using statistics to understand personality
descriptive statistics allow for comparison of scores (normal curve)
Marlowe-Crowne measures social desirability responding
responses show how people present in public
What are the alternatives to self-reporting questionnaires
informant reports, clinical interviews, measuring behavior, archival outcome data, projective tests, physiological measures
informant reports
when others report on someone’s personality (when person may lie, used for children)
clinical interview
assess abnormal levels of personality that make personality disorder (unstructured, semi-structured, fully structured - best format depends)
unstructured interview
interviewer can as any question in any order they want
semi-structured interview
interviewer must ask specific questions but can also ask follow ups
fully structured interview
interviewer can only ask specific questions in a specific order
projective tests (psychodynamic)
measures designed to elicit personality characteristics without directly asking, Rorschach inkblot, thematic apperception test (tell story by picture)
triangulation
using different research methods to answer same question, should correlate and show convergent validity (self report + other report + lab assessment = personality)
mi
mediating variable, where x is correlated to mi which is correlated to y, not x to y
correlational studies
relationship between characteristics, confounding variables, not causation
experiment
cause and effect, random assignments to experimental and control
longitudinal study
collects data from same people over time, help with direction of causality but still correlation
ethics for research
informed consent, truthful reporting, Institutional Review Boards have guidelines
getting accurate results
large enough sample (150-200), calculate number of people necessary with power analysis
meta-analysis
study of studies that analyzes all results together
many labs approach
findings are stronger when obtained in many different labs
open practices
higher transparency, should give roadmap for others to check their work (some academic badges)
open science
Center for Open Science, Open Science Framework, state goals before starting then gets linked when published, Harvard Data Repository is online database
