PSYC3017 FINAL Flashcards
scientific ideals
confirmation of findings, self-correction, scientific explanations are tentative, scientific explanations are rigorously evaluated, replication
Jacob Cohen
first power analysis of psychological research, results - very low power to detect small effects, modest power to detect medium effects, good power to detect large effects
Stapel - the secret life of emotions
do we have feelings we are unaware of that nevertheless affect us? primed fear/disgust using photos and measured word fragment completion. Stapel said he collected data but he actually made it up himself
Bem
precognition - participants correctly identified the future position of erotic pictures more frequently than the 50% hit rate expected by chance
fall out from Bem
Wagenmakers et al. (exploration vs confirmation findings, bayesian test), focus on faulty processes (peer review, self-correcting science), focus on systematic aspects (status and achievement measures, rewards)
is our research replicable? - Many Labs Study
first investigation of variation in replicability, 36 labs replicated 13 studies (10 effects replicated consistently)
reproducibility project
replicating 100 different studies and effects - 36% of replication had significant results, 47% of original effect size were in the 95% confident interval, 39% of effects were rated to have replicated the original result
questionable research practices (QRPs)
false-positive psychology, type I error rate, outlier strategies (what is considered an outlier?), picking at data (stop data collection before sample size issue), stopping data collection, file drawer (not reporting nonsignificant findings), selective reporting
replication is important but also have to focus on
internal validity, external validity, construct validity, consequentiality, communicativeness
Comment on “Estimating the reproducibility of psychological science” - Gilbert et al.
argued the methodology used was finding a crisis when one was not there, said the population of some replication attempts didn’t make sense for the research question
Response to Comment on “Estimating the reproducibility of psychological science” - Anderson et al.
stated Gilbert et al. treated all studies not falling in the confidence interval range of the original publication as not replicating whereas this is not the appropriate approach
solutions and recommendations for replication crisis
p-values, open science, sample non WEIRD populations, focus on boundary conditions, adversarial collaborations, use overarching theory, badges
solutions and recommendations for replication crisis - p-values
get rid of p-values and null hypothesis testing (approach hasn’t been adopted)
solutions and recommendations for replication crisis - open science
open data, source, access, methodology, peer review, educational resources
open science framework
publish replications to motivate researchers, pre-registration of what they’ll do with data, include all variables (rather than discarding later to get significant findings)
solutions and recommendations for replication crisis - focus on boundary conditions
when is the effect relevant? how would that effect emerge? not just whether an effect exists or not
solutions and recommendations for replication crisis - adversarial collaborations
work parallel with people who don’t necessarily agree on concept/approach to research - one person’s bias nullifies other person’s
solutions and recommendations for replication crisis - use overarching theory
use of a specific, metatheory - not a flimsy theoretical basis for findings
terror definition
the existential fear of our own mortality
terror management theory
two responses to mortality salience - proximal and distal
terror management theory proximal response
occurs immediately - deny vulnerability (deny importance of it) or distract yourself
terror management theory distal response
occurs after reflection - world-view defence (cultural background) or self-esteem
terror management theory - why does world-view/cultural perspective ameliorate concerns?
culture makes man seem important and more vital to the universe, not just me as an individuals but the ideal i’m part of
terror management theory - main elements in our worldviews
ingroup/outgroup perceptions, values, practices, gender-roles, social structure, symbols, reputation