research methods in dev psych Flashcards
what is research agreement
research conclusions aren’t always the same, new sources may provide bias - read critically
what is science
studying the world empirically, asking questions, recognizing and overcoming biases
order of the scientific process
observation of a phenomenon - formulate testable hypothesis - select methods/design study - submit research proposal to IRB - collect/analyze data - write paper - submit for review
what should a developed hypothesis include
prediction stated in a testable way, offers support or refutes explanation of phenomena
three types of research designs
scientific observation, experiment, survey
what is scientific obs
researcher systematically and objectively records behavior, requires careful and measurable definitions
measures in scientific observations
reliability, validity
what is reliability
produce consistent results, ex: internal consistency, test-retest reliability, IRR
what is validity
accurately measures what it’s supposed to
ex: construct validity - correlates with other measures of the construct
three types of obs. methods
naturalistic observation, ethnography, structured observation
what is naturalistic observation
observe participants in their natural environment
disadv: does not allow for control over factors of interest, not practical for low base rate behaviors
what is an ethnography
thick description, investigates cultural questions, researcher is a participant observer
disadv: possible researcher influence
what is structured obs
in a lab potentially, assigns participants to specific tasks
disadv: would behavior occur in a natural setting?
what is an experiment
manipulation of environ. or variables to establish a causal relationship between independent and dependent variables
how to know if IV is the cause
measure DV before the treatment (pretest) and after treatment (post test) and compare results – difference = IV cause
pros of experiments
internal validity, enables researchers to learn more clearly about relationship btwn IV and DV; random assignment and blinding increases
cons of experiments
external validity decreases as int. validity increases
what is a survey
collecting info from large number of people using interviews, questionnaires
pros of surveys
can control sampling and sample size, good at answering questions about specific behaviors
types of surveys
self-report, national (w/ nationally representative sample
cons of surveys
lacks ability to manipulate variables, self-report, validity of measures harder to control, less ideal for examining process behind behavior
what is a cross-sectional study, pros vs cons
compare different aged children at the same time point
pros: convenient, quick, inexpensive
cons: can’t answer q’s about development, differences may due to cohort not development
what is a longitudinal study, pros vs convs
follows same children over time and assesses at different ages
pros: can uncover changes over time, strong support for causal relationships
cons: takes longer, need more resources (time, $), participant burden/fatigue
what is a cross-sequential study, pros and cons
follows several groups of dif ages over time
pros: checks for historical context/cohort differences
cons: resource intensive
correlational studies
cannot prove causality, can provide information
natural or quasi-experiments
measures effects or changes that occur naturally in real world
- comparing groups. that were not set up by researcher, used when assignment to groups is not practical
cons: more difficult to determine causality (3rd variable problem)
case studies
used with individuals or small groups to come to more generalized conclusions
psychophysiological studies
- not subject to report bias or verbal fluency
ex: EEG, CAT, fMRI etc
statistical significance vs practical significance
results not found by change vs how large difference is between groups (is it realistically significant) - Cohen’s D (effect size statistics)
pros of meta analysis
small samples can be added together, widely accepted as standardized less biased way to weight evidence
cons of meta analysis
variations across studies make combining impossible sometimes, publication bias, unpublished study exclusion
experiment disusage
logically impossible - some situations uncontrollable (ex: who will present with symptoms of ASD)
ethically impossible
ethical guidelines for children by Society for Research in Child Dev. (SRCD)
- beneficence and non-maleficence - protect from hard
- informed consent - must be obtained before study (or parental consent)
- confidentiality - participant privacy must be maintained