First 676 words of the document:
Research methods sampling
VARIABLES DESIGN experiments random: All members of pop. have an
extraneous: Nuisance variables but INDEPENDENT: Less economical, no lab: IV is manipulated in controlled equal chance of selection. No
order effects, p variables not setting. High control, low realism, researcher bias, time consuming, may
randomly distributed. end up with biased sample.
controlled cause and effect, replication, demand
cONFOUNDING Vary systematically REPEATED:Order effects, demand characteristics. systematic:Selecting every nth
with the IV.
characteristics, no p variable field:IV is manipulated in a natural person from a list. No researcher
dEMAND CHARACTERISTICS:P's second problem, more economical setting. Lower control, higher realism, bias, fairly representative, may end up
guess the aims and alter their beh. with biased sample.
MATCHED PAIRS: No order effects, ethical issues.
INVESTIGATOR EFFECTS: Unconscious cannot be matched exactly, more natural: IV has been manipulated stratified: Sample reflects
influence of researcher on situation. time consuming. naturally, effect on DV is recorded. proportion of people within different
population strata. Cannot count for
RANDOMISATION:Use of chance to Low control, high realism, unique
all sub-groups, no researcher bias,
research, opportunities may be rare.
reduce researcher's influence.
pilot studies quasi:IV based on existing
STANDARDISATION: Ensuring all p's opportunity:Choosing whoever is
are subject to same experience.
pilot studies: Checking procedures difference. between people, effect on
available. Researcher bias,
and materials. Making modifications. DV is recorded. Low control, high
single blind:P's aren't made aware of realism.
volunteer:P's `self select'. Attracts
ETHICAL ISSUES research aims until end. certain profile of person, less time-
INFORMED CONSENT: Advising p's of double-blind: Neither p's nor the observations consuming.
researcher know aim beforehand.
what is involved. May reveal
research aims. Get permission. Control group/condition: Used as a naturalistic: Behaviour observed
Presumptive, prior general,
where it would normally occur. No
control over variables. High realism,
DECEPTION:Telling the truth. low control. structured: Researcher controls
controlled:Some control over what is recorded. May miss
PROTECTION FROM HARM:
correlations environment, including manipulation behaviours.
Minimising psychology and STRENGTHS:Useful preliminary tool. of variables to observe effects. Low unstructured:Researcher records
control, though some extraneous everything. More information, lots of
physical risk. Debrief. Quick and economic to carry out, variables may be controlled, high
using secondary data. qualitative data to analyse.
PRIVACY AND realism (especially when covert).
behavioural categories: Target
CONFIDENTIALITY:Protecting WEAKNESSES: Cannot demonstrate covert/overt: Observing p's with or
cause and effect. The third variable behaviours broken down into
personal data. Maintaining without their knowledge. Covert is observable component. Categories
problem. Misuse and ethically questionable but low p
anonymity. Use numbers not can be too vague and can overlap.
misinterpretation. reactivity. Beh. might be affected in
overt. event sampling: Count events as
participant/non-participant: To join they occur. Useful for infrequent
self-report techniques the group or remain as an outsider. behaviour, may miss complexity.
questionnaires: Pre-set list of
Participant has increased realism but time sampling: Count events at
risk of `going native'. Non-participant timed intervals. Less effort but may
written open/closed questions. Can has more objectivity but less insight. not represent whole behaviour.
distribute to many people, easy to
analyse, social desirability bias.
questions in a fixed order. Same as
questionnaire, but fewer
unstructured interviews:No set
formula, just a general topic.
Questions developed based on
responses. Analysis is more difficult,
more flexibility, social desirability bias
may be reduced by rapport.
questions with flexibility to ask
follow-ups. Advantages of both
structured and unstructured.