reaearch methods

what happens in an experiment

you change the it to measure it’s effect on the dv

why do we carry out experiments?

to determine and see cause and effect

key features of an experiment?

• replicability

• control of variables

what is meant by operationalisation

researcher defining how a variable is measured or changed

extraneous variable

any variable that could affect the dv excluding the iv

confounding variable

a variable that hasn’t been controlled that affected the dv

demand characteristics

cues given that cause a p to guess the aim and change behavior accordingly

investigator effects

cues given by researcher that may influence participant behavior

• lab experiment

controlled environment that allows high control over variables

+ high control > easy to establish cause and effect

- low ecological validity > don’t reflect behaviours shown in everyday life

field experiment

real world setting with ig manipulation

+ higher ecological validity > more realistic

- loss of control > higher chance of confounding variables > precise replication not possible

natural experiment

iv naturally occurs

+high ecological validity > behaviour studied in irl settings

-no control over iv > cause and effect is difficult to establish

naturalistic observation

behaviour observed in a natural setting

+high ecological validity

-low control > low reliability

controlled observation

observation done in a controlled setting

+high control > easy to establish cause and effect

-low ecological validity > less realistic > limited application irl

covert

ps don’t know they’re being observed

-ethical issues (not consent)

+low chance of demand characteristics

overt

ps know they’re being observed

+ethical

-demand characteristics may be shown

participant

researcher joins group

+more insight > intricate details may be seen

-researcher bias > interpretations may be subjective

random sampling

everyone has an equal chance of being selected

+no researcher bias

-time consuming > researchers need a list and need to contact ppl

opportunity sampling

selecting ppl that are available at the time

+convinient > using people that are already present

-may not be representative > low generalisability

volunteer sampling

participants choose to take part

+participants are less likely to leave

-people who volunteer may be similar so reduce change of generalisation

systematic sampling

selecting every nth person

+usually representative of population > increases generalisability

-may still be bias if the list has patterns

independent groups design

different groups participate in different conditions\

+removes order effects

-no control over p variables (diff abilities of ps may influence dv)

fix: random allocation

repeated measures

same group doing different conditions

+fewer ps > less time consuming to find and contact them

-order effects may affect dv

fix: counter balancing > half of ps do conditions in a diff order

matched pairs design

ps are paired based on variables that may affect dv

+no order effects > diff p doing diff condition

-time consuming > matching ps may be expensive and take time

time sampling

recording behaviour within set time intervals

+less time consuming > less observations are to be made

-behaviours may occur our]geode of these intervals > not all behaviours may be observed

events sampling

recording events every time they are observed

+good for infrequent behaviours

-important behaviours may be overlooked

fix: operationalised behavioural categories

type 1 error

when researcher rejects null hypothesis when it was supposed to be accepted