experiments Flashcards
experiment
an experimental research method investigates cause-and-effect relationships by determining if one variable effects another.
the researcher begins with an aim, then formulates a hypothesis to test the potential causal relationship.
types of experimental methods; lab, field, natural and quasi
aim
statement of what the researcher intends to find out
hypothesis
prediction of what they expect to find from the research
you need to know the operationalised independent and dependent variable.
psychologists start their research with both a null and experimental hypothesis and then reject the one that doesn’t support the research
independent variable
variable that the psychologist changes
will lead to two or more conditions
the cause
dependent variable
variable that is measured to see if it changed e.g. number of..
the effect
operationalisation
researcher must assure IV and DV are operationalised - they are both specific and measurable
allows researcher or anybody else to repeat the study the same way to check results
alternative/experimental hypothesis
predicts a difference
“there will be a difference (DV)(condition 1)(condition 2)
there are two types: directional and non-directional
null hypothesis
predicts no difference
“there will be no difference in (DV)(condition 1)(condition 2)
directional hypothesis (one-tailed)
states which specific direction the results will go in. contains words like ‘there will be more/less’ ‘higher/lower’
may be used when findings of a previous effect point to a particular direction
non-directional hypothesis
less clear of the direction
‘there will be a difference’
psychologists may use this if previous research is less or if findings show no consistent direction
extraneous variable
can affect the DV if not controlled. consists of 3 types: participant, situational, experimenter
participant variables
differences between participants e.g. age, gender, prior knowledge and experiences
situational variables
differences in the environment such as the time of day or temperature or noise
experimenter variables
the tone of voice or body language the experimenter uses or bias
standardised procedures or matched pairs
method to control situational variables
ensure all pts are tested under the same conditions - avoid situational variables
repeated measures design or matched pairs
control participant variables
repeated measures: same pts used in every condition
matches pairs: match a pt from each condition on the basis of a key variable e.g. age.
pairs are then compared rather than whole groups
generally, psychologists should just design a study that makes it hard for pts to guess the aims of the study
double blind technique
control of experimenter variables
the person who carries out the research is not the same as the one who designed it
demand characteristics
structure of the study could lead to a pt guessing the aim of the study - can lead to them changing their behaviour - artificial - lowers validity
laboratory experiment
investigates causal relationships between an IV and a DV under controlled conditions
conducted in a special environment where variables are controlled - usually a university lab
strengths of laboratory experiment
strengths:
manipulation of IV under controlled conditions allows cause & effect to be established as EVs are controlled
easy to replicate due to standardised procedure
weakness of laboratory experiment
weakness:
experiments are artificial which may produce artificial behaviour - low ecological validity
low mundane realism - whether the task relates to real life tasks
ethical problems - deception, potential harm etc
field experiment
investigates causal relationships between IV and DV in a more natural environment.
pts may not be aware they are participating
researcher still manipulates the IV e.g. at a school vs casino
strength of field experiment
strengths:
higher mundane realism
less chance of demand characteristics - pts are not aware - can’t change behaviour as they don’t know the aims of the study.
easy to generalise results to our settings
weakness of field experiment
weaknesses:
ethical issues; deception, invading privacy, potential harm etc
difficult to control EVs - less validity - cannot measure what we intended to measure
natural experiment
investigates relationship between IV and DV but IV cannot be directly manipulated - so the IV varies naturally
e.g. researching prisoners, kids in an orphanage - more extremal
strength of natural experiment
strengths:
allows psychologists to carry out research where the IV cannot be manipulated for ethical reasons or practical reasons
allows psychologists to study real problems - increases mundane realism and ecological validity
weakness of natural experiment
weaknesses:
cannot demonstrate casual relationships between IV and DV
random allocation not possible therefore there may be EVs that cannot be controlled - lowers validity
quasi-experiment
investigates relationships between IV and DV but IV cannot be directly manipulated - IV varies naturally.
IV is a characteristic of a person e.g. mental disorders, gender, personality - more internal
strength of quasi-experiment
strengths:
allows psychologists to carry out research where the IV can’t be manipulated for ethical/practical reasons
allows them to study real problems - increases mundane realism and ecological validity
weakness of quasi-experiment
weaknesses:
cannot demonstrate casual relationships as IV is not manipulated
random allocation is not possible - may be EVs that can’t be controlled - lowers the validity
experimental designs
once researcher has chosen experimental method that suits the nature of the study best, they choose the type of design
there are 3 types: independent groups, repeated measures and matched pairs
independent groups
one set of participants who only represent one condition of the IV
advantages of independent groups
eliminate order effects - pts only take part in one condition
pts are less likely to guess the aim of the study and show demand characteristics
disadvantages of independent groups
cannot control effects of participant variables
to improve this - randomly allocate pts to conditions to distribute variables easily
advantages of matched pairs
eliminate order effects - pts only take part in one condition
less likely to guess aim and show demand characteristics
disadvantages of matched pairs
cannot control all participant variables
to improve this, conduct a pilot study to consider key variables that are important when matching
advantages of repeated measures
eliminate PVs
disadvantages of repeated measures
order effects are likely
more likely to guess aim of the study and show demand characteristics
to improve this researchers may use 2 different tests to reduce order effects.
there are 2 techniques to tackle order effects; counterbalancing and randomisation
counterbalancing
alternating the order in which participants perform in different conditions.
e.g. group A does condition 1 first then condition 2, and group B do condition 2 then condition 1
randomisation
material for each condition is presented in a different order
e.g. same words are presented but in a different order for each pt.
random allocation (for independent groups design)
using a non-biased method to allocate pts to conditions.
this will evenly distribute PVs across all conditions, so PVs are less likely to affect the results
e.g. putting all pts names into a hat then the first person drawn is allocated to condition 1. then the second person drawn to condition 2. the third person to drawn condition 1 and so on.
this greatly decreases systematic error as well as researcher bias
reliability
how consistent the results are. if study is repeated, the same or highly similar results should occur again
validity
extent to which a test measures what it intended to measure
ecological validity
extent to which you can apply the findings of the study to other situations/environments outside the setting of the study
control
how well the researcher has controlled the variables - without control we cannot state that it was the IV that changed the DV
mundane realism
how well an experiment task reflects real life tasks - the higher the mundane realism, higher the ecological validity
internal validity
refers to things that happen INSIDE the study such as whether we can be certain that it was the IV which caused changed in the DV
internal validity can be affected by a lack of mundane realism - lead the pts to act in an unnatural way - less valid results
can also be affected by EVs
external validity
whether the study is a true representation of behaviour outside of the specific experimental setting
population validity
extent to which you can generalise the findings to the rest of the population
temporal/historical validity
extent to which we can generalise the findings to other time periods
pilot studies
a trial run on a small scale study conducted on a small sample and aims to test all aspects; procedure, flaws in design, check materials used.
a way to check that everything in your actual experiment will run smoothly and help to improve validity
find flaws before time and money is invested in carrying out a large study