Research Methods
Experimental Method
♡ aims- general statements that describe the purpose of the investigation
♡ hypothesis- statement made at the start of a study stating relationship between variables as stated by
the theory
♡ directional hypothesis- states the sort of difference anticipated between the conditions (scores in
condition A will be higher than in condition B), used when findings of previous research suggest particular
outcome
♡ non-directional hypothesis- states there will be a difference between conditions, but doesn’t specify the
nature of the difference (scores in condition A will be different to scores in condition B), used when there is
no previous research or earlier studies are contradictory
♡ researcher manipulates the independent variable and measures the effect of the change on the
dependent variable
♡ we compare effects of IV by comparing control group and experimental group
♡ operationalisation- clearly defining variables in terms of how they can be measured
Control of variables
♡ it is key that changes in the DV are caused only by manipulation of the IV; any other variables that
could interfere with the IV are called extraneous variables
♡ extraneous variables should be identified and minimised before an experiment
♡ nuisance variables do not vary systematically with the IV, so do not confound the results
♡ confounding variables change systematically with the IV, creating a second IV so we cannot tell what
caused changes in the DV; results completely confounded
♡ participant reactivity- the extraneous variable of participants trying to make sense of the new situation
♡ demand characteristics- cues that help participants guess the aim of the study
♡ participants may act how expected or in a way to please the experimenter (Please you effect) or
deliberately underperform to sabotage the experiment (screw you effect); either way behaviour is no
longer natural
♡ investigator effects- any unwanted influence of the investigator on the research outcome, including
expectancy effects and unconscious cues
♡ randomisation- use of chance to reduce researcher’s influence on design of study, minimises
investigator effects
♡ standardisation- all participants should be subject to the same environment and information, so
procedures are made the same for all participants
Experimental design
♡ experimental design- how participants are used in experiments
♡ independent groups- 2 separate groups of participants experience different conditions, all participants
undergo 1 condition of the IV, performance of groups compared to establish effects of IV on DV
♡ s- no order effects as participants only have to undergo 1 condition of iv
♡ s- less demand characteristics as participants don’t see both conditions of iv
♡ w- less economical as you need 2x as many participants
♡ w- participant variables occur, differences may be caused by differences between participants; random
allocation used to evenly distribute participant characteristics, so the effects are extraneous
♡ repeated measures- all participants undergo both conditions of the IV, then their data from one
condition is compared to their data from the other condition
♡ s- no participant variables can affect DV
♡ s- more economic
♡ w- participants get more demand characteristics, less authentic behaviour
♡ w- order effects, participants may be bored or fatigued by the second task, or performance improved by
the second task as participants are better practised; becomes confounding variable, so counterbalancing
used where ½ undergo condition A first and ½ undergo condition B first, so effects aren’t confounding
♡ matched pairs- participants are paired together on a variable relevant to the experiment (e.g. memory
test participants matched on IQ); participants with 2 highest IQs matched together, then 1 participant from
each pair randomly allocated to condition A, the other goes to condition B to control for participant
variables
♡ s- no order effects, demand characteristics less of a problem, reduced participant variables
♡ w- time consuming
♡ w- less economical than repeated measures
Types of experiment
, ♡ lab experiments- takes place in highly controlled environment where researcher manipulates IV and
records effect on DV while maintaining strict control over extraneous variables
♡ s- high control of extraneous variables, clear cause and effect and high internal validity
♡ s- replication more possible due to high control, allows us to check that results aren’t anomalous
♡ w- lack generalisability as lab setting not representative and too artificial, low ecological validity
♡ w- lack mundane realism and more likely that participants can guess the aim of the experiment, lacks
internal validity
♡ field experiment- takes place in a natural setting where researcher manipulates the IV, records effects
on the DV
♡ s- more mundane realism so behaviour is more authentic and natural, so can be better generalised to
real life, higher external validity
♡ w- loss of control over extraneous variables, cause and effect difficult to establish and precise
replication not possible
♡ w- ethical issues as if participants unaware they are being studies they cannot consent, invasion of
privacy
♡ natural experiment- change in IV not brought about by researcher but researcher takes advantage of
pre-existing Iv
♡ s- can study things you can’t manipulate, such as institutionalisation (Rutter 2011)
♡ s- high external validity as involve study of real-life issues as they happen
♡ w- IV may happen very rarely, reducing opportunity for research and limiting scope for generalisation to
other issues
♡ w- no random allocation to groups, difficult to establish cause and effect as DV may have caused IV,
Rutter- more intelligent children adopted first
♡ Quasi-experiments- IV is based on existing difference in populations such as age or gender
♡ s- carried out under controlled conditions so high control
♡ w- no random allocation to groups
Sampling
♡ target population- the subset of a general population the researcher is interested in studying
♡ sample- smaller group of target population which is representative of the target population (whole
population too large to be studied)
♡ sample should be representative of target population so findings are generalisable
♡ random sample- all members of target population have equal chance of being chosen, complete list of
members of target population obtained, all names on list assigned a number, sample generated through
lottery method (computer-based randomiser)
♡ s- completely free of researcher bias, avoiding them choosing people who they think would fit
hypothesis
♡ w- difficult and time-consuming to obtain full list of target population
♡ w- sample may still be unrepresentative
♡ w- participants may refuse to take part so we end up with a volunteer sample
♡ systematic sampling- every nth member of the target population is selected, sampling frame is a list of
people from target population organised into for example alphabetical order, interval determined
randomly, researcher works through sampling frame until sample complete
♡ s- completely avoids researcher bias
♡ s- usually quite representative
♡ w- can end up as a volunteer sample
♡ stratified sample- composition of sample reflects proportions of people in sub-groups (strata) in target
population, researcher identifies strata that make up population, proportions of strata in population
worked out, participants that make up each stratum selected by random sampling
♡ s- completely avoids researcher bias
♡ s- produces a representative sample
♡ w- time consuming
♡ opportunity sample- researcher selects anyone willing and available, researcher may stand on street
and takes chance to ask anyone that passes them by
♡ s- convenient and saves time and money
♡ w- unrepresentative sample as draw from one very specific area so cannot be generalised
♡ w- researcher has complete control over selection of participants allowing researcher bias
♡ volunteer sample- participants select themselves to be part of the study, researcher places advert or
participants volunteer when the researcher asks
♡ s- easy and requires minimal input from the researcher
♡ w- volunteer bias, attracts people who are helpful and curious so not representative
Ethical issues
♡ ethical issues- conflict that exists between participants’ rights and researchers’ needs to gain
meaningful findings
