Resharch Methods Flashcards
Name the 4 experimental methods
- Lab
- Field
- Natural
- Quasi
Lab experiment
Controlled study carried about in an artificial setting such as a university office
Lab experiment strengths ?
. Controlled so more reliable
. Fewer extraneous variables
Lab experiment cons?
. Lacks ecological validity as its unrealistic
. Risk of demand characteristics
Field experiment?
A controlled study carried out in a realistic setting – testing people where they would naturally be
Field experiment strengths?
. Controlled so more reliable and fewer extraneous variables
. Realistic setting so higher ecological validity
Field experiment weaknesses
- Harder to control as many extraneous variables as in a lab
- Less ethical (more deception)
Natural experiment
The researcher does not manipulate the independent variable. Instead, it is a naturally- occurring experience or behaviour
Strengths of natural experiment
+ Total ecological validity as it has already happened
+ Can test things you couldn’t if you wanted to manipulate variables (e.g. trauma)
Weaknesses of natural experiment
- Lacks reliability as there is no control or standardisation
- Limited to the people who have the experience/behaviour
Quasi experiment?
The researcher does not manipulate the independent variable instead its a naturally- occurring characteristic
Strengths of quasi experiment
+ Total ecological validity as it has already happened
+ Can test things you couldn’t if you wanted to manipulate variables (e.g. intelligence)
Weaknesses of quasi experiment
- Lacks reliability as there is no control or standardisation
- Limited to the people who have the characteristic
3 non experimental methods
. Questionnaire (self report technique)
. Interview ( self report technique)
. Observations
Questionnaires?
Participants are given a set of written questions, either ‘closed’, ‘open’ or ‘scale’
Strengths of questionnaire
+ Quick (can be sent to thousands instantly via email)
+ Easy to quantify and analyse
Weaknesses of questionnaire?
- Not a good method for detailed info
- Can’t build a rapport with the participant
Interviews?
Live questioning (usually face:face). Could be structured (set questions) or unstructured (conversational)
Interviews strengths?
+ Can build a rapport and make them feel at ease
+ Can ask follow-up questions for detail
Interviews weaknesses
- Higher risk of social desirability bias as it is
face:face
Observations?
Watching the actual behaviour of participants. Controlled vs. naturalistic; overt vs. covert; participant vs. non-participant
Strengths of observations?
+ Observe the actual behaviour so lower risk of demand characteristics/social desirability bias
Weaknesses of observations?
- Cannot observe a person’s thoughts
- Could be affected by researcher bias
4 more non experimental methods
. Content analysis
. Case study
. Correctional analysis
. meta analysis
Content analysis?
Quantitative analysis of qualitative data. Researchers identify specific behavioural categories and then tally their occurrence
Strengths of content analysis
+ Quantitative, meaning more reliable
+ Can use secondary data with high ecological validity
Weaknesses of content analysis
- By quantifying the data you lose its meaning, so
conclusions are less valid
Case study’s ?
An in-depth study of one person or a small number of people, usually those who have a unique characteristic or experience
Case study strengths
+ Able to ‘triangulate’ and study the person using multiple methods + In-depth
Case study’s weaknesses
- Lacks reliability as it is naturally-occurring
- Risk of subjectivity over time
Correctional analysis
Statistical analysis of two or more co variables to see id there is a realsonships between them often secondary data
Correlational analysis strength
+ Useful for identifying potential areas of research
Correlational analysis weakness
- Cannot identify causation, only correlation
Meta analysis?
Compiling data from multiple studies that investigate the same behaviour to create overall conclusions
Strengths of meta analysis
+ Eliminates individual researcher bias
+ Large sample sizes
+ Can cover a range of methods
Weakness of meta analysis
- Secondary data so cannot confirm all studies were standardised
Three types of hypotheses
. Directional
. Non directional
. Null
When should you only use a directional hypothesis
If the is published past research .
Directional hypothesis?
Predicting that there will be a difference between conditions/correlation between variables, and stating the direction.
Non directional hypothesis ?
Predicting that there will be a difference between conditions/correlation between variables, but not stating the direction
Null hypothesis
Predicting that there will be no difference/correlation.
Name the 5 sampling methods.
- Random
- Opportunity
- Volunteer
- Systematic
- Stratified
Random sampling?
Number people in your sampling frame; use a random number generator to decide who should take part.
Strength of Random sampling?
+ No investigator effects in deciding who to choose
Weaknesses of random sampling?
- Could be randomly unrepresentative
- Needs a sampling frame
Opportunity sampling
Choosing people who are in the right place at the right time (e.g. select people who are on the high street)
Strengths of sampling methods?
+ Easier to carry out since the participants are already available
Weaknesses of opportunity sampling
- Risk of investigator effects in choosing who they like
- Likely to be unrepresentative
Volunteer sampling
Place an advertisement and participants self-select
Volunteer sampling strengths
+ Ensures participants will be motivated
+ More ethical as participants know they are being tested
Weaknesses of volunteer sampling?
- High risk of being unrepresentative
- Hard to incentivise people to take part
Systematic sampling ?
Gather a sampling frame of people; choose every nth person (n = sampling frame size/sample size)
Systematic sampling strengths?
+ If you list people in order (e.g. by age), you can ensure a diverse sample
+ No investigator effects
Systematic strengths weaknesses
- Needs a sampling frame
Stratified sampling?
Calculate proportions in the population (e.g. 60:40 Male:Female). Randomly select proportionate numbers for your sample
Strengths of stratified sampling
+ Most representative of the population
+ No investigator effects
Weaknesses of stratified sampling?
- Needs a sampling frame
- Might not be able to gather
the data you would like
Name three experimental designs?
. Independent groups
. Repeated measures
. Matched pairs
Independent groups?
Each participant only takes part in one condition of the independent vat level
Strengths of independent groups?
+ low risk of demand characteristics
+ no order effects
Weaknesses of independent groups
- participant variables mean it is hard to compare groups
- requires more participants
Repeated measures?
Each participant takes part in all conditions of the independent variable
Strength of repeated measures?
+ easier to compare performance on conditions validity because they are the same people. Both times
Weaknesses of repeated measures ?
- high risk of demand charactertics
- high risk of Order effects
Matched pairs ?
Each participant only takes part in one conduction however the researcher tests partipents and sorts them into equal groups
Matched pairs strengths?
+ no order effects and limited demand characteristics.
+ reduces the effect of participant variables
Weaknesses of matched pairs?
- often impossible/impractical to match and sort paratipents beforehand
- still risk of participant variables
Independent variable ?
The difference between the conditions its the thing you change
Dependant variable?
The variable that you measure
Control variable
Anything you keep the same in all conditions like the same room or time of day
Extraneous variables ?
Anything other than the independent variable that has an affect on the dependant variable
Situational - any changes in the environment tests etc . Between conditions
Partipant - differences between participants in condition 1 and 2 that affect dependant variable
Confounding variables?
A third variable within a correlation that affects both the other two variables consistently
What is counterbalancing ?
During a repeated measures design half of the participants carry out condition A first and then B; the other half carry out B first and then A. This helps to eliminate order effects by cancelling it out.
Standerdarstatidon
Keeping everything the same from one condition to the next. For example, writing down your instructions for participants instead of trying to be consistent in how you explain the study to them.
