week 4 research approaches Flashcards
falsifiability
the ability to disprove hypotheses as criterion for scientific research
refuting a general principle
causality
distinguishing cause and effect relationships from mere correlation
validity
ensuring internal (study-specific) and external (generalisable) conclusions
what did alhazen do
advocated observation and systematic inquiry
what did Francis bacon introduce
inductive method and systematic experiments
what did Karl popper emphasize
falsifiability and critical rationalism
what do you associate popper and bacon with
observation
what is another way to say raven paradox
the paradox of indoor ornithology
explain the raven paradox
work on the logic of confirmation and is based on the principle of inductive reasoning. It can be summarized as follows:
- A universal statement, such as “All ravens are black,” is logically equivalent to its contrapositive: “If something is not black, then it is not a raven.”
- Observing a black raven confirms the statement “All ravens are black.”
- By the same logic, observing a green apple (a non-black, non-raven) should also confirm the statement, because it supports the contrapositive.
what is inductive reasoning
drawing general conclusions or forming hypotheses based on specific observations or evidence.
what did Karl popper refute
inductive reasoning - can’t make a general statement from a few examples
4 steps of hypothetico-deductive model
- research question
- hypothesis
- predictions through deductive inference
- test - falsify
what is deductive inference
logical reasoning in which a conclusion is derived from one or more general premises that are assumed to be true.
what is the emphasis of hypothetico-deductive model
avoid logical error like affirming the consequence
operationalisation
translate general hypothesis into specific prediction of measurements - defining variables
generalisation
translate specific results of a test into a general statement that contributes to a theory
confounder
condition/factor whose variation systematically affects the DV, but is not part of the IV
noise
unsystematic random variation of measurements producing uncertainty
how to control confounds
keep all conditions constant even those not part of the experiment
how to control noise
can’t fully eliminate noise but you can minimize it
measurements should be as precise as possible - little variation
weakness of observational research
no test of causality as no manipulation or control
clutter of data
ptsp can act differently when they know they’re being watched
ethical problems - consent
what does cooccurrence not allow for
establishing cause and effect
what do confounds produce
spurious correlations/relationships between IV and DV
spurious correlation
A spurious correlation is a misleading statistical association between two variables that occurs because of the influence of one or more confounding variables, not because of any actual causal relationship between the two variables.
quasi experiment
IV is partially but not completely independent - conditions of IV are measured but not controlled or manipulated
weakness of quasi experiment
systematic differences between groups can be con founds that may account for observed effect
sampling error - not representative
uncontrolled variation of conditions can produce systematic or random variation of DV
4 major criteria of empirical evidence
- reliability
- accuracy
- internal validity
- external validity
reliability
reproducibility of results
test reliability
how reliable the measurements are
statistical reliability
how high is random noise in the data
experimental reliability
how stable results are across experiments
repeatability
results can be reproduced by same researched in same lab
reproducibility
other researchers in another lab reproduce results
replicability
other researcher in another lab reproduce same results by replicating conditions
accuracy
how correct the data is
internal validity
validity of results - does it test hypothesis
2 types of internal validity
- confounds - validity of experimental control
- construct and content validity - whether a measure actually measures the idea eg are IQ tests a measure of intelligence
external validity
the extent to which the results of a study can be generalized so that conclusions apply beyond the context of the study
2 types of external validity
- ecological validity - whether results apply to different settings
- population validity - whether the result of sample applies to whole population
pro and con of lab exp
pro = full control and max internal validity
con = low external validity - not real life conditions
field exp
real environment including all factors
pro and con of field studies
pro = high external validity
con = low internal validity and ethical issues
how to organize research report
- intro - assumptions, research question, hypothesis, prediction
- method = exp control of confounds
- results - data and analysis
- discussion - internal and external validity
- conclusion
the impossibility of verification is a logical result of
the generality of proposition
