Experiments
Lab experiment
Conducted in an artificial environment
Researcher has control over the variables – they manipulate the IV and measure the DV. As it is a structured environment extraneous variables are controlled
An example is Craik & Tulving who used a lab experiment to measure levels of processing by measuring the number of words recalled and whether they were processed semantically, phonetically or structurally.
+ Extraneous variables are controlled, so cause and effect can be established
+ Reliable as a standardised procedure if followed, so it can be replicated
+ Ethical because pp’s can give consent, be given the right to withdraw and be given a brief and debrief
- Lacks ecological validity because the situation is unrealistic to everyday life
- As pp’s are aware they are being studied they may show demand characteristics
- There may be experimenter effects where the experimenter influences the results
Field experiment
Conducted in a natural environment
The researcher still has control over the IV and seeks to measure the DV
An example is Hofling as this study used nurses who were on duty at the time and measured whether nurses obeyed a phone call from an unknown doctor
+ PP’s will act naturally, as they may be unaware they are being studied – so no demand characteristics
+ High ecological validity as the experiment takes place in a real life environment so can be generalised to real life situations
- No control over extraneous variables so the study may not be reliable
- More time consuming to set up
- Less ethical as pp’s may be unaware they are in a study, so do not give consent or have the right to withdraw
Natural experiment
Conducted in a natural environment
The IV is naturally occurring and not manipulated by the researcher; and the effects are then measured
Charlton’s study measured the behaviour of children before and after satellite TV was introduced to a community in St Helena.
+ High ecological validity, this is highly realistic as the variable occurs naturally
+ More ethical, the IV is not being controlled by the researcher and the situation would be occurring naturally
- Lack of control over variables and ability to generalise – the researcher has no control over the pp’s
- Difficult to replicate as natural experiments are often unique situations
Key Terms
Independent variable = what is changed to see if there is a difference e.g. males/females
Dependent variable = what the researcher measures to see the effects of the IV e.g. a score
These should be operationalised – defined precisely e.g. score out of 100, number of hours
Extraneous variables – any variable except the IV that affects the findings
Confounding variables – a factor not controlled that does affect findings
Situational variables – something in the environment that could affect findings e.g. noise, light, time of day, location, heat
Participant variables – something about the participant that could affect findings e.g. fatigue, previous knowledge, age
Participant Designs
Independent groups: Pp’s only do one experimental condition
+ No order effects
+ No demand characteristics
- No control over pp variables
- Large sample needed
Repeated Measures: PP’s do all experimental conditions
+ No pp variables/ no individual differences
+ Smaller sample can be used
- Order effects
- Demand characteristics
Matched Pairs: Pp’s only do one experimental condition, but pp’s in each condition are matched on characteristics important in the study e.g. age, ability
+ PP variables are reduced
+ No order effects
- Time consuming to find a large sample and to match the pp’s
- Although pp variables are reduced there may still be differences between groups
Sampling Methods
Random: Everyone in the target population has an equal chance on being chosen e.g. all names in a hat, name generator
+ unbiased as no control over who is chosen
- Time consuming and difficult to conduct unless a small target population
Opportunity: Whoever is available at the time of the research
+ Quick and easy to do, as do not need to spend time looking for pp’s
- Likely to be biased to specific groups who are free at the time
Volunteer: Advert is placed and those who want to take part sign up to do so
+ PP’s will be motivated and willing to take part, so engage fully
- Likely to be biased as may be motivated to take part for specific reasons
Stratified: Sample is proportional to each group within the target population e.g. gender, age
+ Highly representative so it can be generalised
- Very time consuming and difficult to do
Lab experiment
Conducted in an artificial environment
Researcher has control over the variables – they manipulate the IV and measure the DV. As it is a structured environment extraneous variables are controlled
An example is Craik & Tulving who used a lab experiment to measure levels of processing by measuring the number of words recalled and whether they were processed semantically, phonetically or structurally.
+ Extraneous variables are controlled, so cause and effect can be established
+ Reliable as a standardised procedure if followed, so it can be replicated
+ Ethical because pp’s can give consent, be given the right to withdraw and be given a brief and debrief
- Lacks ecological validity because the situation is unrealistic to everyday life
- As pp’s are aware they are being studied they may show demand characteristics
- There may be experimenter effects where the experimenter influences the results
Field experiment
Conducted in a natural environment
The researcher still has control over the IV and seeks to measure the DV
An example is Hofling as this study used nurses who were on duty at the time and measured whether nurses obeyed a phone call from an unknown doctor
+ PP’s will act naturally, as they may be unaware they are being studied – so no demand characteristics
+ High ecological validity as the experiment takes place in a real life environment so can be generalised to real life situations
- No control over extraneous variables so the study may not be reliable
- More time consuming to set up
- Less ethical as pp’s may be unaware they are in a study, so do not give consent or have the right to withdraw
Natural experiment
Conducted in a natural environment
The IV is naturally occurring and not manipulated by the researcher; and the effects are then measured
Charlton’s study measured the behaviour of children before and after satellite TV was introduced to a community in St Helena.
+ High ecological validity, this is highly realistic as the variable occurs naturally
+ More ethical, the IV is not being controlled by the researcher and the situation would be occurring naturally
- Lack of control over variables and ability to generalise – the researcher has no control over the pp’s
- Difficult to replicate as natural experiments are often unique situations
Key Terms
Independent variable = what is changed to see if there is a difference e.g. males/females
Dependent variable = what the researcher measures to see the effects of the IV e.g. a score
These should be operationalised – defined precisely e.g. score out of 100, number of hours
Extraneous variables – any variable except the IV that affects the findings
Confounding variables – a factor not controlled that does affect findings
Situational variables – something in the environment that could affect findings e.g. noise, light, time of day, location, heat
Participant variables – something about the participant that could affect findings e.g. fatigue, previous knowledge, age
Participant Designs
Independent groups: Pp’s only do one experimental condition
+ No order effects
+ No demand characteristics
- No control over pp variables
- Large sample needed
Repeated Measures: PP’s do all experimental conditions
+ No pp variables/ no individual differences
+ Smaller sample can be used
- Order effects
- Demand characteristics
Matched Pairs: Pp’s only do one experimental condition, but pp’s in each condition are matched on characteristics important in the study e.g. age, ability
+ PP variables are reduced
+ No order effects
- Time consuming to find a large sample and to match the pp’s
- Although pp variables are reduced there may still be differences between groups
Sampling Methods
Random: Everyone in the target population has an equal chance on being chosen e.g. all names in a hat, name generator
+ unbiased as no control over who is chosen
- Time consuming and difficult to conduct unless a small target population
Opportunity: Whoever is available at the time of the research
+ Quick and easy to do, as do not need to spend time looking for pp’s
- Likely to be biased to specific groups who are free at the time
Volunteer: Advert is placed and those who want to take part sign up to do so
+ PP’s will be motivated and willing to take part, so engage fully
- Likely to be biased as may be motivated to take part for specific reasons
Stratified: Sample is proportional to each group within the target population e.g. gender, age
+ Highly representative so it can be generalised
- Very time consuming and difficult to do