Research Methods: Experimental Method

aim

clear and concise intent of research

aim of an experiment

to investigate the cause-and-effect links between IV and DV

IV

a change (manipulation of groups) directly affecting DV

DV

measurement of the effect on IV (always a numerical value)

hypothesis

predicts outcome of research

directional (one-tailed) hypothesis

states the possible outcome of results based on previous research

non directional (two-tailed) hypothesis

only predicts that there will be an unspecified difference

null hypothesis

states there will be no significant difference

nominal measurement of DV

used to categorise data and requires mode to find average

ordinal measurement of DV

used to order/rank data in a numerical order and requires median to find average

interval measurement of DV

equal intervals or measurements and requires mean to find average

standard deviation

a measure of the variation of a set of numerical values where the intervals are consistent

operationalising variables

where experimenter needs to define the variable and state exactly how accurately and precisely they intend to measure or manipulate it

extraenuous variables

a variable that may affect the measurement of the DV and should be controlled by experimenter

confounding variables

when an extraenuous variable isnt properly controlled it distorts relationship between IV and DV

situational variables

variables that should be controlled in the setup of an experiment using a standardised procedure

participant variables

individual differences of participants that cannot be controlled

demand characteristics

participants that are unsure how to behave so try to work out what is required of them

researcher bias

where conciously or not the researcher may impose a bias to be favourable towarrds their hypothesis

single blind

a control for demand characteristics where participants dont know the aims of the study so dont alter behaviour

double blind

a control for demand characteristics and experimenter effects where participants and experimenter dont know the aims of the study so standardised procedures used

standardised procedures

a control for experimenter effects where it is ensured all participants have the same experience in the experiment

counterbalancing

randomise condition/group to control order/fatigue effects

randomisation

random allocation of participants to groups to conttrol participant variables

reliability

a measure of the consistency of results

test retest method to measure reliability

same participants complete the same experiment at different times and expect the same results

inter observer method to measure reliability

researchers independantly give same results at the same time and require 80% agreementt/0.8 conccordance for the results to be reliable

validity

measure of accuracy

internal validity

accuracy of measure of DV due to manipulation of IV

face validity

type of internal validity - the extent to which an experiment appears to measure what it intends to measure

external validity

whether data can be applied across different situations

ecological validity

type of external validity - the extent to which results can be applied to everyday settings

temporal validity

type of external validity - the extent to which the results can be applied across different times

population validity

type of external validity - the extent to which results can be applied to different people

2 ways to measure validiity

→ assessment of face validity

→ concurrent validity (resuults compared to valid results then correlation conducted)

purpose of controls

so variables dont affect data ensuring internal validity remains unaffected

4 types of experiments

→ lab

→ field

→ natural

→ quasi

laboratory experiments

→ takes place in an artificial environment (i.e lab or controlled setting)

→ strict procedures

→ establishes cause-effect relationship between IV and DV

laboratory experiments: strengths

→ replicable due to strict controls, so consistent results, so reliable

→ cause-effect relationship established as strict controls so high internal validity

→ objective (double blind increases level of objectivity)

laboratory experiments: weaknesses

→ low ecological validity due to unnatural settings and tasks - can be maintained with experimental realism

→ high demand characteristics - lowered with deception but raises ethical concerns

→ high researcher bias/effects - affect participants performance and validity

experimental realism

extent to which the setup of the laboratory feels real so participants behave naturally despite artificial setup

field experiments

→ takes place in a natural but controlled setting

→ mundane realism

mundane realism

refers to the extent in which the set up of the study reflectsthe everyday

field experiments: strengths

→ high ecological validity

→ high internal validity - can establish cause effect relationship

→ high objectivity so scientific

→ demand characteristics lower

field experiments: weaknesses

→ harder to control all variables

→ lower reliabiliity as difficult to repliicatte

→ experimenter effects due to lower

→ harder to get informed consent wo demand characteristics

natural experiments

→ researchers cannot directly manipulate IV (occurs and varies naturally)

→ research event or life experience

natural experiments: strengths

→ research allowed

→ allows study of real events/problems

→ + strengths whether used lab or field to collect data apply

natural experiments: weaknesses

→ research allowed

→ allows study of real events/problems

→ + weaknesses whether used lab or field to collect data apply

quasi experiments

→ researchers cannot directly manipulate IV (occurs and varies naturally)

→ research natural occurences we are born with

quasi experimnents: strengths

→ higher generalisability and population validity as IV happens in larger samples

→ lower demand characteristics

→ strengths whether used lab or field to collect data apply

quasi experiments: weaknesses

→ confounding and participaant variabbles affect results as IV not randomised

→ experimenter effects due to lower randomisation

→ + weaknesses whether used lab or field to collect data apply