Methods after the midterm: 7 Small n research Flashcards
what type of designs does small n reseach use (within, between, mixed)?
within-subject design
what are the seperate experiments in small n research
the subjects
how is reliabilty assesed in small n research
replication
what type of situations is small n research useful?
assumption of minimal biological or psychological variability
- avalibility or convenience constraints
- interest in small area of population
- detailed comparison
- need to understand process in time
- no need for generalization
consistent ____ and ____ in small n research
level (magnitude of treatment effects) and trend (one-way/ unidirectional changes)
spatio-temperal patterns in small n research mean what
search of swquential analysis of behaviour. Serial configuration of events and actions in time
what is type of series for the search of temporal patterns?
time series analyses
what famous ethologist used small n research and for what? What n was used?
BF skinner used small n of 4 for rats and assumed vary low biological variability and generalized rats to humans very quickly
explain what a consistent trend is in small n research (we want a consistent trend)
expected every subject to react the same way
explain what a stability is in small n research
we need to know what the subject looks like before the independent variabel is applied (baseline established)
explain what a temperal patterns is in small n research
detailed analysis over time
tracking something in space and time e.g. time series
explain what sequences of spatio-temperol patterns is in small n research
sequences analysis of behaviour in time and space
e.g. movement analysis
what is pooling falicy (small n research)
cannot compensate for power by gathering more data per participant - doesnt make it more genralizable and does not increase power
what is the prinicple of experimental research
replication
who said it is important to replicate and observe change in a subject e.g. learning
bernard
how to know if you are able to generalize a small n study
replicate to determine if effect was fluke
should you replicate every small n stufy to see if it is generalizable?
no - if there is no effect when you first start off you don’t need to continue
what is beneficial about small n studies (extraneous varibales)?
easier to determine/ pinpoint because being more precise about data you are collecting for each subject
problems with small n research?
- carry-over effects (like within subject)
- weak effect if IV doesnt work
- if baseline is off, phenoma is unobservable (DV)
What are the 3 types of small-n study designs?
Baseline- goof for small n, make sure you do this before applying IV
dynamic - moment to moment changes
discrete- individual subject performance
what is important in basline exams
replication needs to reproduce the same or very similar results - baseline only met when a stability criterion is reached
why can’t repliable baselines be created with large n research
differences in baseline in populations are different - average out dont aply to majority, not reliable baseline
what lets you know your IV is soley responsible for the effect seen?
a table baseline
intra-subject replication?
subjects are their own control , wait till they reach baselin until performing next treatment
what is reversal strategy in small n designs
looking at the recovery from a treatment back to subjects baseline
what is a stable baseline?
differs by study - establish an operatinoally defined stability criterion - acceptable time to reach criterion
issues with baselines (uncontrolled variation)
slowly drifting baseline over time - might not notice it - it’ll influence results causing issue
issues with baselines (irreversible)
participant is unable to return to baseline
What are the two contrasting approaches to baselines (explain and name them)
group approach - statistical methods should be used to control for failure to reduce uncontrolled variation
single-subject approach -
should try to identitfy extraneous variables not under control and bring them under experimental control
what is systematic replication
replication beyond single-subject procedures - you do a smaller one and then if it turns out well replicate study with more subjects
what is a drifting baseline
basline that slowly systemically changes
what is an unrecoverable baseline
baseline that won’t have reversal due to carryover effects
what are inappropriate baseline levels
low baseline good UNLESS data has a floor effect.
high baseline good UNLESS data has ceiling effect
- martha and john example with different starting - seeing if effect is same -
explain single factor single-subject baseline designs
1 independent varaible (AB, ABA, or ABAB)
multifactor designs - single-subject baseline designs
two or more independent variables e.g. melatonin + bright light
multiple-baselines designs - single-subject baseline designs
several dependent variables
e.g polygraph
dynamic designs
processes and behaviours through time ; often called time series
discrete trials design (example)
subject receive each treatment or condition many times and each tiral (session) is a data point;
e.g dog session, go for lunch, come back new session. 1-5 session per day
what is controlled for in discrete traisl designs
extraneous variables
how are treatments presented to subjects
randomized or counterbalanced (change up order for no order effects; not all subjects see same order)
what is another name of sequential analysis
sequential hypothesis testing
what does inter-subject variation mean within discrete trials
they look at them and compare between subjects
what is special about discrete trials compared to other small n studies
they can be large designs
what is essential in small n designs (final exam material)
every single instance matters
what is psychophysics an example of (sub sections: signal detection theory, and detection of stimuli)
discrete trial design examples
what type of designs are decision-making, judgment and diagnosis studies?
discrete trial design examples
learning, cognition (memory), concepts: discrimmination/ categorizetion are what type of designs
discrete trial designs
in neuroscience, what kinds of designs are most often used
discrete trials designs
how long are sequential analysies
as long as they need to be until significant results are observed or the quota “N” is obtained –> the stopping rule
advantages to sequential analysies
- could be economical
- conclusion might be reached ealier than with common hypothesis testing principles
- comprimise between idiographic and nomenthetic research
when are sequential analysis used
when subject availability is questionable
often used in clinical trials
explain general procedure of sequential analysis
start with 2 trials or so,,,, compare two groups check if alternative hypothesis is accepted. If yes, its terminated. if not add more n and continue. Continue until alternative is acceptive or ‘n’ quota is reached
