Research Methods Flashcards
Types of hypothesis
- null
- alternative
Null hypothesis
- IV will not affect the DV
- no relationship between observed variables
Alternative hypothesis
IV will affect DV
Types of variable
- independent
- dependent
- extraneous
Independent variable
What the researcher changes/manipulates to test its effect on the DV
Dependent variable
The outcome/effect that is being measured in a study
Extraneous variable
Unwanted variable that can affect the DV
Target population
Group of people that researcher wants to generalise findings to
Sample
Small portion of Ps from target population being studied
Types of Sampling
- random
- opportunity
- systematic
- stratified
Random sampling
- selection of participants is random
- e.g- names out of a hat or numbered and selected by computer
Random sampling STRENGTH
- unbiased results
- all target population have equal chance of being picked
- sample is more representative of target population, more validity
Random sampling WEAKNESS
- takes much time and effort
- have to obtain list of all members of target population and number them, then may not want to take part
- effort may not be with it
Opportunity sampling
Asking those who are most easily available
Opportunity sampling STRENGTH
- quick and easy
- choose people who are nearby and available
- less time consuming
Opportunity sampling WEAKNESS
- greater chance of being biased
- sample drawn from narrow part of target population, may only have 1 certain type of person - may be participant variables
- reduces generalisability/reliability of results
Systematic sampling
Selecting every ‘nth’ member of the target population as a participant
Systematic sampling STRENGTH
- avoids researcher bias
- researcher has no say over who’s - selected
- more representative
Systematic sampling WEAKNESS
- need bigger sample size
- e.g- if you require 100 Ps for study and picked every 10 Ps, 1000 Ps would be needed
- may be time consuming method
Stratified sampling
- sub-groups within target population identified
- Ps chosen from each sub-group in proportion to their occurrence in target population
Stratified sampling STRENGTH
- most representative sampling method
- all subgroups represented in proportion to target population
- findings should have high reliability/validity to make generalisations to target population
Stratified sampling WEAKNESS
- time consuming method
- have to identify sub-groups, select necessary Ps and attempt to get them proportionate
- difficult/impractical method to use
Experimental designs
- independent groups
- repeated measures
- matched pairs
Independent groups
- Ps divided into (usually 2) subgroups
- groups take part in different experimental condition
- used for comparison
Independent groups STRENGTH
- no practice effects
- Ps only do 1 condition so won’t benefit from practice, less likely to display demand characteristics as it will be harder for them to figure out aim of study
- results will be more representative of how members of target population usually behave
Independent groups WEAKNESS
- may be participant variables
- each group may contain more of a type of person than another, as each group only does 1 condition of IV, this may affect results
- results may be less representative, can’t be generalised to all of target population
Repeated measures
All Ps do all conditions of the experiment
Repeated measures STRENGTH
- requires less effort when gathering Ps
- don’t have to split Ps into groups of pairs based on certain characteristics
- set-up of experiment will be less time-consuming
Repeated measures WEAKNESS
- higher risk of practice effects
- as Ps do all conditions of experiment, they may improve with practice or display demand characteristics if they figure out aim of study, potentially avoided with counterbalancing
- researcher may view improvement as result of IV rather than demand characteristics
Matched pairs
- Ps paired on relevant variables to study (eg: ethnicity, disabilities, gender)
- one P goes into each group
Matched pairs STRENGTH
- reduced participant variables
- Ps taking part matched on variable relevant to experiment
- results will have more validity, more representitive of target population
Matched pairs WEAKNESS
- difficult to match people of personality variables
-can generally only match people on fixed traits (sex,race,age), personality factors may be relevant to experiment - may produce results that aren’t a result of IV
Types of experiment
- lab
- field
- natural
Lab experiment
- takes place in a controlled environment
- researcher manipulates IV
Lab experiment STRENGTH
- limits role of EVs
- researchers have full control of environment, also easier to replicate and reliability checked
- can establish cause and effect relationship, can trust results
Lab experiment WEAKNESS
- artificial environment
- results gathered in controlled lab setting may not reflect real-world situations, Ps may behave unnaturally
- lack ecological validity
Field experiment
- takes place in natural/everyday setting,
- experimenter manipulates IV
Field experiment STRENGTH
- more realistic environment
- thought to make responses and behaviours of Ps more realistic as not always aware they are being observed
- high internal validity as behaviours of Ps can be generalised to wider population
Field experiment WEAKNESS
- difficult to replicate
- environment of experiment may be difficult to recreate, Ps may be members of public with personality factors influencing the results which are unaccounted for
- replication and reliability issues
Natural experiment
- takes place in natural/everyday environment
- IV occurs naturally
Natural experiment STRENGTH
- Ps often produce no demand characteristics
- Ps unaware of experiment so won’t be able to work out aim of study and behave falsely
- results most likely indicative of behaviour that can be generalised to wider population
Natural experiment WEAKNESS
- ethical issues
- consent - Ps often not aware they are being studied so can’t give informed consent, may not wish to take part and be monitored, may sometimes be okay if Ps debriefed after
- not always ethically right to natural experiments
Interview
- researcher being in direct contact with P
- face to face or phone/video call
- researcher records responses of P
Types of interviews
- structured
- semi-structured
- unstructured
Structured interviews
- interviewer reads prepared questions only
- follow up questions prepared before
Structured interviews STRENGTH
- can be easily replicated
- questions all preset so can simply be said by another researcher
- can easily check results for reliability and consistency is drawn conclusions
Structured interviews WEAKNESS
- often lack qualitative data
- usually have mostly closed questions so give little insight into thoughts/feelings, maybe less relevant follow-up questions
- less useful and detailed results
Semi-structured interviews
- some questions prepared before
- follow up questions come from answers
Unstructured interviews
- Interviewer has general aim
- few/no questions prepared before
- new questions based on previous answers
Unstructured interviews STRENGTH
- allow Ps to explain responses
- researcher asks new questions based on previous answers so can be asked to expand/explain
- results more accurate/valuable than from structured
Unstructured interviews WEAKNESS
- require more skilled interviewers
- need to articulate themselves better and quickly think of good questions, not like structured where questions can simply be read out
- less opportunities to be carried out
Questionnaire
- researcher designs set of questions for Ps to answer
- Ps taking part called ‘respondents’
Types of questionnaire questions
- open
- closed
Open questions
- Ps can answer how they want
- produce qualitative data
Closed questions
- fixed range of possible answers
- produce quantitative data
Questionnaires STRENGTH
- much data can be gathered quickly
- can be sent out to many people physically/electronically to acquire many Ps responses
- more data means results can be generalised to wider population
Questionnaires WEAKNESS
- Ps may misunderstand questions so answer them incorrectly
- unlike in interview with researcher, Ps can’t clarify meaning of question so may respond in a way not representative of their views
- results may be less reliable/valid
Case studies
- in depth investigation into individual/group/event
- usually involve unusual situations
- can be used in everyday experiences
- produce mostly qualitative data
- usually longitudinal
Case studies STRENGTH
- info collected over long period of time
- usually longitudinal so gradual changes and much data can be recorded over time
- can provide in depth and accurate results
Case studies WEAKNESS
- target single/few individuals
- situation/factors influencing outcomes unlikely to be relatable to others
- results less representative of wider population, can’t be generalised
Observation
Researcher watches/listens to Ps engaging in behaviour being studied and records behaviour
Types of observations
- naturalistic
- controlled
- covert
- overt
- participant
- non-participant
Naturalistic observations
- recorded in a place where it would naturally occur
- nothing controlled or changed
Controlled observations
Part of environment controlled by researcher
Covert observations
Ps not aware they are being observed/recorded
Overt observations
Ps told in advance they will be observed/recorded
Participant observations
Researcher becomes part of group they are studying
Non-participant observations
Researcher remains separate from group they are studying
Observations STRENGTH
- true to how Ps actually act
- what people say is often different to what they really do, if Ps asked about behaviour, they may give socially desirable responses, observations allow true behaviour to be recorded
- high ecological validity, representative of real-world behaviour
Observations WEAKNESS
- ethical issues
- consent - Ps often not aware of observation so can’t give informed consent, informing before may lead to altered behaviour as they are aware they are being observed
- not informing Ps raises ethical issues of privacy + lack of consent
Categories of behaviour
- behavioural categories created to make it clear what behaviours are being recorded
- number of times behaviour is observed is recorded
Inter-observer reliability
- 2 independent observers with same categories of behaviour
- compare data for consistency
- should produce same results
- discuss differences
Correlation
A relationship between 2 variables
Types of correlation
- positive
- negative
- zero
Positive correlation
- cause and effect relationship between 2 variables
- as 1 variable increases/decreases, so does the other
Negative correlation
- cause and effect relationship between 2 variables
- as 1 variable increases/decreases, the other does the opposite
Zero correlation
No relationship between variables
Correlations STRENGTH
- good starting point for research
- if two variables are related, gives researchers ideas for future investigations to find true cause of correlation
- more scientifically useful
Correlations WEAKNESS
- doesn’t provide all detail
- only tells us if 2 variables are related, doesn’t tell which variable caused relationship or if there is 3rd unknown variable influencing the other 2
- usually require further experimentation to establish all details
Standardised procedures
Set of sequences that apply to all Ps when necessary to ensure the experiment is unbiased
Types of instructions to participants
- briefing
- debriefing
Briefing
Ps being encouraged to participate with a log of what is discussed to gain consent
Debriefing
- Ps being given detailed explanation of aims of a study at end
- ethical issues raised and Ps can withdraw their data/contributions
Purpose of randomisation
To make sure there are no biases in a procedure
Methods of reducing researcher bias
- random allocation
- counterbalancing
Random allocation
- when using independent groups design, Ps are randomly selected for the condition of the IV they do
- e.g- names chosen out of a hat
Counterbalancing
- used to deal with order effects in repeated measures
- sample divided in half, half completes conditions in one order, half the other
Ethical Issues
- concerns about what is morally right/wrong when using participants in research
- when there is conflict between needs of research and rights of Ps
Ethical considerations
- consent
- confidentiality
- deception
- privacy
- protection from harm
Consent
- must be informed - given info and purpose of study
- Ps given right to withdraw consent and leave
Confidentiality
- personal data protected and respected
- researcher must guarantee anonymity and with hold Ps’ names
Deception
- Ps not lied to about aims of study without justification
- mild deception justifiable, major deception if benefits outweigh ethical cost
- Ps must be debriefed and given true aims at end of study
Privacy
- Ps have right to control info about themselves
- researcher mustn’t observe Ps without informed consent unless in public space
Protection from harm
- Ps physical and psychological safety protected, including stress + embarrassment
- Ps reminded they can leave any time
Quantitative data
Data in numerical form
Quantitative data STRENGTH
- easy to analyse
- can be converted to averages/graphs/charts to check for patterns/correlations
- can be easily compared, more useful
Quantitative data WEAKNESS
- lacks depth and detail
- obtain little info about thoughts/feelings
- doesn’t reflect real-world complexity
Qualitative data
Data is descriptive/non-numerical form
Qualitative data STRENGTH
- more depth/detail
- gain insight into thoughts + feelings, can better understand attitudes/beliefs
- more psychologically useful
Qualitative data WEAKNESS
- difficult to check for reliability
- information from Ps mostly subjective
- results may be biased, can’t be generalised
Primary data
collected firsthand by researcher
Primary data STRENGTH
- suits aims of research
- authentic, comes first hand from Ps so can be trusted
- data more useful
Primary data WEAKNESS
- takes time and effort to collect
- must design and carry out study
- slows down process, increases expense
Secondary data
- comes from other sources/studies
- already published
- collected by other researcher with different aim
Secondary data STRENGTH
- convenient to use
- results already collected + checked
- reduces expense + effort
Secondary data WEAKNESS
- may not fit what researcher wants
- from different study with different aim, may have been poorly designed
- may reduce validity
Validity
Whether results reflect real world behaviour
Reliability
- consistency
- if you can repeat measurement and get same results, measurement is reliable
Bar chart
- displays data in categories
- spaces between bars
Histogram
- displays continuous data
- no spaces between bars
Characteristic of normal distribution
Mean, median and mode all close together
