RESEARCH METHODS
PILOT STUDIES
Definition:
Pilot studies are small scale trial runs of the investigation carried out
before the actual investigation
They are used with all research methods
They use a smaller number of ppts who are similar to those who
take part in actual study
Aim of pilot studies
To identify any potential issues by checking all aspects of research
are clear and EVs are controlled for
To modify the design and procedure by changing the things that
don’t work
Pilot studies therefore improve quality of research by increasing validity
and reliability of procedure and findings. Help avoid unnecessary work,
save time + money
CONTROL OF VARIABLES
Variables: any object/behaviour being studied
Extraneous variables: all other variables/features of research, other
than IV, that have potential to affect DV if not controlled
Confounding variables: any EVs, other than IV that has effected the DV
to the extent where the researcher is uncertain whether the IV was the
source of change to the DV
Participant variables
Any individual differences between ppts that may affect the DV.
Must be relevant to research
o Age, intelligence, motivation, IQ, memory, gender etc
Control
o Repeated measures or matched pairs design
o Random allocation of ppts to experimental conditions
Participant effects
Ppt reactivity (demand characteristics)
Ppts actively seek cues from environment or investigator on how to
behave and change their behaviour as
o They may be unsure about what’s expected
o They may want to be helpful
o They may want to ruin the research
, Social desirability bias
o Ppts may wish to present themselves in the best way and
therefore not behave according to personal preference but the
way thy consider most socially acceptable
Hawthorne Effect
o Ppts performance may change because they’re responding to
extra attention that they’re receiving
Control
o Single blind technique: ppts not told which condition they’re in
or whether there’s a condition at all
o Experimental realism: make task as interesting as possible in
order to engage ppts
Situational variables
Any features of experimental situation surrounding environment
that may affect DV
o Time of day, temperature, noise, weather etc
Control
o Good research design based on pilot study
o Standardised procedures to ensure conditions the same for all
ppts
Demand characteristics
Cues in research environment that tell ppts what is expected from
them and what investigator hopes to find (investigator effects)
May cause ppts to change behaviour
Possible cues include
o Communication during study
o What ppts may have already heard
o The way the ppt was approached and asked to volunteer
o Type of person researcher is
o Setting or surroundings
Control
o Use design based on pilot study and aim for experimental
realism
o Single bling technique
o Standardised procedures + instructions
o Matched pairs
Investigator effects
When person collecting data has knowledge of what the aim is and
that affects data obtained
o Direct effects
Unwanted influence of investigator on the research
outcome due to investigators expectancy effects and
unconscious cues
Includes personal attributes of experimenter
o Control
, Double blind technique
Standardised procedures + instructions
o Indirect effects
Due to design of research which ensues desired results
because of bias on part of researcher, e.g. selection of
ppts
o Control
Pilot study for good design
Repeated measures design to avoid order effects by
counter balancing
Use randomisation
SAMPLING
Population: entire group of individuals with specified characteristics
that a researchers interested in studying. Often referred to as target
population. A smaller sample selected from target population but results
are generalised to target population
Sample: group of people selected from target population who take part in
research and presumed to be representative of that population
Representative sample: group of people who form part of sample that
accurately reflects characteristics in target population due to similar
distribution of relevant characteristics. Allows researchers to generalise
findings with confidence
Sampling techniques: methods used to selects a sample from target
population
Random sampling
Systematic sampling
Stratified sampling
Opportunity sampling
Volunteer sampling
Bias: when certain groups may be over or under represented within
sample. E.g. too many younger people or too many people of one ethnic
origin. Problem as limits extent to which findings can be generalised
Generalisation: extent to which findings can be broadly applied to
population
Technique Definition Method Strengths Weaknesses
Random Every member Identify everyone As everyone has Difficult + time
sampling of target in target equal chance of consuming-
population has population selection sample complete list on
equal chance of Name/assign is more target population
being selected number to each representative- may be hard to
member findings can be obtain or
, Place all numbers generalised. population might
in a hat and Free from be too large.
select which researcher bias- If sample size is too
members will be researcher has no small may be
part of sample control over who unrepresentative
is selected for , cannot be
study generalised.
Some ppts may
refuse to take part
Systemati When every nth Sampling frame If sampling frame If a periodic trait
c member of of target has been exists In population
sampling target population randomised, or if list not
population is organised in sample should be produced
selected some way unbiased making randomly, sample
produced it will not be
Sampling system representative, representative.
chosen results Time consuming +
deliberately or generalised. may be difficult to
randomly Free from achieve, may still
Researcher researcher bias- be
selects from researcher has no unrepresentative if
sampling frame control over who sample size is too
using system is selected for small
study
Stratified Composition of Researcher Most Identified strata
sampling sample reflects identifies different representative cant reflect all the
proportions of strata that makes sample- ways people are
people in up population. population is different- complete
certain sub- Proportion of each proportionally representation is
groups within stratum based on reflected, results not possible.
target proportion on the are valid + Time consuming
population. population is generalised w dividing population
calculated. confidence. into strata +
Sample selected Free from randomly selecting
at random from researcher bias- from each
each stratum researcher has no
using control over who
randomisation is selected from
the study
Opportuni Consists of Researcher Convenient cost Influenced by
ty sample of approaches + time effective researcher bias-
sampling people people and asks method. complete control
available at them to With natural over selection of
time study is participate experiments, this ppts- findings
carried out- method has to be unlikely to be
willing to used as accurate reflection.
participate + fit researcher has no Unlikely to be
criteria control over who representative as
is being studied. ppts likely to share
PILOT STUDIES
Definition:
Pilot studies are small scale trial runs of the investigation carried out
before the actual investigation
They are used with all research methods
They use a smaller number of ppts who are similar to those who
take part in actual study
Aim of pilot studies
To identify any potential issues by checking all aspects of research
are clear and EVs are controlled for
To modify the design and procedure by changing the things that
don’t work
Pilot studies therefore improve quality of research by increasing validity
and reliability of procedure and findings. Help avoid unnecessary work,
save time + money
CONTROL OF VARIABLES
Variables: any object/behaviour being studied
Extraneous variables: all other variables/features of research, other
than IV, that have potential to affect DV if not controlled
Confounding variables: any EVs, other than IV that has effected the DV
to the extent where the researcher is uncertain whether the IV was the
source of change to the DV
Participant variables
Any individual differences between ppts that may affect the DV.
Must be relevant to research
o Age, intelligence, motivation, IQ, memory, gender etc
Control
o Repeated measures or matched pairs design
o Random allocation of ppts to experimental conditions
Participant effects
Ppt reactivity (demand characteristics)
Ppts actively seek cues from environment or investigator on how to
behave and change their behaviour as
o They may be unsure about what’s expected
o They may want to be helpful
o They may want to ruin the research
, Social desirability bias
o Ppts may wish to present themselves in the best way and
therefore not behave according to personal preference but the
way thy consider most socially acceptable
Hawthorne Effect
o Ppts performance may change because they’re responding to
extra attention that they’re receiving
Control
o Single blind technique: ppts not told which condition they’re in
or whether there’s a condition at all
o Experimental realism: make task as interesting as possible in
order to engage ppts
Situational variables
Any features of experimental situation surrounding environment
that may affect DV
o Time of day, temperature, noise, weather etc
Control
o Good research design based on pilot study
o Standardised procedures to ensure conditions the same for all
ppts
Demand characteristics
Cues in research environment that tell ppts what is expected from
them and what investigator hopes to find (investigator effects)
May cause ppts to change behaviour
Possible cues include
o Communication during study
o What ppts may have already heard
o The way the ppt was approached and asked to volunteer
o Type of person researcher is
o Setting or surroundings
Control
o Use design based on pilot study and aim for experimental
realism
o Single bling technique
o Standardised procedures + instructions
o Matched pairs
Investigator effects
When person collecting data has knowledge of what the aim is and
that affects data obtained
o Direct effects
Unwanted influence of investigator on the research
outcome due to investigators expectancy effects and
unconscious cues
Includes personal attributes of experimenter
o Control
, Double blind technique
Standardised procedures + instructions
o Indirect effects
Due to design of research which ensues desired results
because of bias on part of researcher, e.g. selection of
ppts
o Control
Pilot study for good design
Repeated measures design to avoid order effects by
counter balancing
Use randomisation
SAMPLING
Population: entire group of individuals with specified characteristics
that a researchers interested in studying. Often referred to as target
population. A smaller sample selected from target population but results
are generalised to target population
Sample: group of people selected from target population who take part in
research and presumed to be representative of that population
Representative sample: group of people who form part of sample that
accurately reflects characteristics in target population due to similar
distribution of relevant characteristics. Allows researchers to generalise
findings with confidence
Sampling techniques: methods used to selects a sample from target
population
Random sampling
Systematic sampling
Stratified sampling
Opportunity sampling
Volunteer sampling
Bias: when certain groups may be over or under represented within
sample. E.g. too many younger people or too many people of one ethnic
origin. Problem as limits extent to which findings can be generalised
Generalisation: extent to which findings can be broadly applied to
population
Technique Definition Method Strengths Weaknesses
Random Every member Identify everyone As everyone has Difficult + time
sampling of target in target equal chance of consuming-
population has population selection sample complete list on
equal chance of Name/assign is more target population
being selected number to each representative- may be hard to
member findings can be obtain or
, Place all numbers generalised. population might
in a hat and Free from be too large.
select which researcher bias- If sample size is too
members will be researcher has no small may be
part of sample control over who unrepresentative
is selected for , cannot be
study generalised.
Some ppts may
refuse to take part
Systemati When every nth Sampling frame If sampling frame If a periodic trait
c member of of target has been exists In population
sampling target population randomised, or if list not
population is organised in sample should be produced
selected some way unbiased making randomly, sample
produced it will not be
Sampling system representative, representative.
chosen results Time consuming +
deliberately or generalised. may be difficult to
randomly Free from achieve, may still
Researcher researcher bias- be
selects from researcher has no unrepresentative if
sampling frame control over who sample size is too
using system is selected for small
study
Stratified Composition of Researcher Most Identified strata
sampling sample reflects identifies different representative cant reflect all the
proportions of strata that makes sample- ways people are
people in up population. population is different- complete
certain sub- Proportion of each proportionally representation is
groups within stratum based on reflected, results not possible.
target proportion on the are valid + Time consuming
population. population is generalised w dividing population
calculated. confidence. into strata +
Sample selected Free from randomly selecting
at random from researcher bias- from each
each stratum researcher has no
using control over who
randomisation is selected from
the study
Opportuni Consists of Researcher Convenient cost Influenced by
ty sample of approaches + time effective researcher bias-
sampling people people and asks method. complete control
available at them to With natural over selection of
time study is participate experiments, this ppts- findings
carried out- method has to be unlikely to be
willing to used as accurate reflection.
participate + fit researcher has no Unlikely to be
criteria control over who representative as
is being studied. ppts likely to share
