research methods Flashcards
types of sampling
- opportunity
- volunteer
- random
- systematic
- stratified
sampling
OPPORTUNITY
- most available / easiest to obtain
✔ quick + convenient method
✘ unrepresentative sample
- so cannot be generalised
sampling
VOLUNTEER
- self-selecting
✔ willing ppts
- likely to engage more
✘ volunteer bias
- ppts may share certain traits
- limited generalisation
sampling
RANDOM
- every person in target population has an equal chance of being selected
✔ unbiased
- researcher has no influence over who is selected
- free from researcher bias
✘ people selected may be unwilling to take part
sampling
SYSTEMATIC
- ppts selected using a sampling frame
- every nth person
✔ unbiased
- first item is usually selected at random
- free from researcher bias
✘ requires time + effort
- requires complete list of population
- random sampling would be more ideal
sampling
STRATIFIED
- strata within the population is identified (eg age, gender)
- sample reflects proportion of different strata
✔ representative
- characteristics of target population are represented
- allows generalisation
✘ imperfect
- strata can’t reflect all the ways in which people are different
research issues
- extraneous variables: nuisance variables
- confounding variables: change systematically w IV so can’t be sure if any observed change in DV is due to IV or CV
- demand characteristics: any cue that may reveal aim of study
- investigator effects: any effect of investigator’s behaviour on outcome of study
research techniques
- randomisation: use of chance to control effect of bias
- standardisation: same formalised procedure for all ppts - more repeatable
- control groups: to set a comparison
- single bind: ppt doesn’t know aim of study (to reduce demand characteristics)
- double bind: both ppt + researcher don’t know aims of study (reduces demand characteristics + investigator effect)
- counterbalancing: half of ppts in sample carries out conditions in 1 order + the other half in reverse order - eliminates order effect w repeated measures design
pilot studies
small-scale trial run of a study before doing the real thing to see if there’s any problems with:
- experimental design
- intructions for ppts
- measuring instruments
- to ensure time, effort + money aren’t wasted on a flawed methodology
- important to use a representative sample of target population
experimental methods
LAB
- controlled environment where extraneous variables can be regulated
- IV is manipulated + effect on DV is recorded
✔ high degree of control
- minimises effect of extraneous variables
- conclusions about cause + effect can be drawn confidently
- high internal validity
✔ standardised procedure
- can be easily replicated - can confirm validity
✘ lacks external validity
- limits generalisability
experimental method
FIELD
- natural setting
- IV manipulated + effect on DV recorded
✔ high ecological validity
- representative of behaviour in everyday life
- results are more generalisable
✔ ppts unaware they’re being studied
- no demand characteristics
- high internal validity
✘ less control over extraneuos variables
- more difficult to draw conclusions about cause + effect
- low internal validity
✘ ethical issues
- no informed consent
- invasion of privacy
experimental method
NATURAL
- IV not manipulated
- measures effects of existing IV (naturally occurring, eg flood/earthquake) on the DV
✔ greater ecological validity
- involve real-life issues
- findings are more relevant to real experiences
✘ natural event may only occur rarely
- reduces opportunity for research
- limits scope for generalisation
experimental method
QUASI
- IV based on pre-existing difference between people, e.g. age or gender
- measures effect of this IV on DV
✔ high control
- high internal validity
✔ comparisons can be made between people
✘ ppts are not randomly allocated
✘ causal relationships not demonstrated
experimental design
INDEPENDENT GROUPS
- ppts randomly allocated to different groups
- 1 does condition A; other does condition B
✔ no order effects
✔ reduces demand characteristics
✘ participant variables - may reduce validity
✘ more participants required - time-consuming
experimental design
REPEATED MEASURES
- same ppts take part in all conditions
- order of conditions should be counterbalanced to avoid order effects
✔ controls participant variables
✔ fewer participants
✘ order effects
✘ increases demand characteristics
experimental design
MATCHED PAIRS
- 2 groups of ppts used but are paired on participant variables that matter for the experiment
✔ controls participant variables + demand characteristics
✔ no order effects
✘ matching is time consuming and can’t control all relevant variables
✘ more participants required - time consuming
observational techniques
naturalistic vs controlled
naturalistic: takes place where target behaviour would normally occur
✔ high ecological validity
- more generalisable
✘ low control
- low internal validity
controlled: some manipulation of variables including control of EVs
✔ replicable
-standardised procedures
✘ low external validity
-limits generalisation
observational techniques
covert vs overt
covert: ppts unaware they’re being studied
✔ eliminates demand characteristics
- high external validity
✘ ethically questionable
- ppts right to privacy
overt: ppts are aware they’re being studied
✔ more ethically acceptable
- ppts have given consent
✘ demand characteristics
- may influence behaviour
- lower external validity
observational techniques
participant vs non-participant
participant: when researcher becomes part of the group they are studying
✔ greater insight
- enhances validity of findings
✘ less objective
- more bias
non-participant: when researcher remains separate from group they are studying
✔ more objective
- less chance of bias
✘ loss of insight
- may reduce validity of findings
observational methods
behavioural categories: target behaviour to be observed broken up into set of observable/measurable categories
✘ difficult to make clear + unambiguous
- have to be self-evident + must not overlap
time sampling: target behaviour is recorded at prescribed intervals
✔ reduces no. of observations
- more structured + systematic
✘ may be unrepresentative
- may not reflect the entire behaviour
event sampling: target behaviour is recorded every time it occurs
✔ may record infrequent behaviour which could be easily missed during time sampling
✘ complex behaviour oversimplified
- may affect validity of findings
self-report techniques
QUESTIONNAIRES
- made up of a pre-set list of written questions to which a ppt responds
✔ can be distributed to a lot of people
- large amounts of data gathered quickly
- cost-effective
✔ respondents may be willing to open up
- less chance of social desirability bias
✘ may not be honest
✘ acquiescence bias (response bias)
- tendency to agree regardless of their beliefs
- or not reading the question properly
questionnaires
open vs close
closed questions: respondent has limited choices
+ data is quantitative
✔ easier to analyse + draw conclusions
✘ respondents are restricted
- may be unrepresentative
- reduces validity of findings
open questions: respondents provide own answers expressed in words + data are qualitative
✔ not restricted
- more detailed answers
- more insightful
✘ difficult to analyse
self-report techniques
INTERVIEWS
interview schedule
standardised list of qs for interviewer to cover
can reduce interviewer bias
quiet room
will increase likelihood of ppt opening up
rapport
beginning w neutral qs to make ppt feel relaxed
ethics
remind ppts that answers will be treated in confidence
interviews
structured vs unstructured
structured: list of pre-determined qs asked in a fixed order
✔ easy to replicate
- standardised format
✘ interviewees cannot elaborate / deviate from topic
unstructured: general topic to be discussed, free-flowing interaction
✔ greater flexibility
- more likely to gain insight
✘ difficult to replicate
- risk of interviewer bias
- semi-structred
correlations
case studies
- detailed, in-depth + longitudinal analysis of an individual/group/institution/event
- often involves analysis of unusual individuals or events
- eg person w rare disorder
- usually involves qualitative data
✔ rich, detailed insight
- increases validity of data
✔ enables study of unusual behaviour
- some conditions are very rare + cannot we studied using other methods
- some cases can help understanding of normal functioning
✘ low ecological validity
- studying a single person/event
- unique cases
- difficult to generalise to wider population
✘ prone to researcher bias
- based on the subjective interpretation of the researcher
- may reduce validity
✘ ppts accounts may be biased
- personal accounts prone to inaccuracy/memory decay
- evidence may be low in validity.
ethical issues
INFORMED CONSENT
ppts should be aware of aims of research, procedures, their rights (incl to withdraw) + what their data will be used for in order no make an informed decision about whether to take part
dealing w it:
- should be issued w a consent letter, detailing all relevant info that may affect their decision to participate + a signature must be obtained
ethical issues
PROTECTION FROM HARM
ppts should leave research in the same psychological + physical state as they entered; should be at no more risk than they would in everyday life
dealing w it:
- should be reminded of their right to withdraw at each stage of the research process
- should be debriefed at the end + reassured that their behaviour was normal
- may be referred to counselling in some cases
ethical issues
DECEPTION
deliberately misleading or witholding info from ppts meaning consent is not informed
dealing w it:
- at the end of the study, ppts should be given a full debrief, including the true aims + nature of the research what their data will be used for
- should be given the right to withold data if they wish
- should also be reassured that their behaviour was normal
- in extreme cases, they may be offered counselling
- contact details of experimenter should be given if they have any further questions of queries
ethical issues
CONFIDENTIALITY
a ppt’s personal info is protected by law under the Data Protection Act both during + after the experiment
dealing w it:
- any personal info should remain undisclosed to protect their identity + assure anonymity
ethical issues
PRIVACY
the right of individuals to decide how info about them will be communicated to others
dealing w it:
- should be provided w informed consent + right to withdraw at any stage
- should be explained the ways in which their info will be protected + kept confidential
peer review
- before publication, all aspects of the investigation are scrutinised by experts (peers) in the field
- these peers should be objective + unknown to the researcher
aims to:
* funding - allocate research funding
* validation of the quality + relevance of research
* improvements + ammendments are suggested
*
economical implications
types of data
quantitative vs qualitative
quantitative: numerical data
✔ easier to analyse + draw conclusions
- more objective + less open to bias
✘ oversimplifies behaviour
- doesn’t provide meaningful insight
qualitative: non-numerical data expressed in words
✔ more detailed
- thoughts + feelings can be explained
✘ difficult to analyse + draw conclusions
types of data
primary vs secondary
primary: firsthand data collected for purpose of investigation
✔ info is directly relevant to research aims
✘ requires time + effort
secondary: taken from journal articles, books, websites or gov records
✔ already exists
- inexpensive + minimal effort
✘ quality may be poor
- may be outdated/incomplete
- challenges validity
types of data
meta-analysis
- type of secondary data that involves combining data from a large no. of studies
✔ increases validity of conclusions
- much larger sample size
- more generalisable
✘ publication bias
- may not select all relevant studies
descriptive statistics
measures of central tendency
mean: arithmetic average
✔ sensitive
- includes all values from data set in calculation
✘ may be unrepresentative
- easily distorted by any extreme values
median: middle value
✔ unaffected by extreme values
- only focussed on middle value
✘ less sensitive than mean
- not all values included in calculation
- extreme values may be important
mode: most frequently occurring value
✔ relevant ro discrete/categorical data
- sometimes the only appropriate measure
✘ overly simple
- may be many modes in a data set
- not useful in this case
descriptive statistics
measures of dispersion
range: difference between the highest + lowest values
✔ easy to calculate
✘ only takes into account 2 most extreme values
- unrepresentative of data set as a whole
- doesn’t account for distribution of values
standard deviation: measure of average spread around the mean
✔ more precise than range
- includes all values in calculation
✘ may be misleading
- extreme values may not be revealed
presentation + display of quantitative data
distributions
content analysis
type of observational research where people are studied indirectly via qualitative data (their communications, eg texts, emails, TV, film and other media)
coding: 1st stage of content analysis
- data sets may be extremely large so info needs to be categorised into units eg counting up no of times a word/phrase appears in a text
- produces quantitative data
thematic analysis: produces qualitative data + refers to any idea that is recurrent
- more descriptive than coding units
✔ many ethical issues may not apply
eg obtaining consent
- as material already in public domain
✔ flexible method
- can be adapted to suit aims of the research
✘ communication is studied out of context
- likely to reduce validity
reliability
- measure of consistency
- whether a measurement is repeated + same result is obtained
reliability
ASSESSING RELIABILITY
TEST-RETEST: same test/questionnaire given to the same person on 2 / more different occasions
- result should be similar
INTER-OBSERVER: compare observations from different observers in a pilot study of the same event
measuring reliability: 2 sets of data are correlated
- correlation coefficient should exceed + .80 for reliability
reliability
IMPROVING RELIABILITY
rewriting questions in questionnaines:
- replacing some open qs (can be misinterpreted) w closed qs
- improve training / use same person for interviews
- standardised procedures in experiments - strict control over instructions + conditions
- operationalisation of behevioural categories in observation
- behavioural categories should be measurable + not overlap
validity
is result legitimate?
- whether
ECOLOGICAL validity: do findings generali re no other rething (eneryday life?
TEMPORAl validity: do findings remain true oner vime ?
validity
ASSESSING VALIDITY
FACE VALIDITY - achieved by ‘eyeballing’ measuring instrument OR passing it on to an expert to check
CONCURRENT VALIDITY - test is administered to group of ppts + scores are compared w findings of a well-established test
- correlarion should exceed + .80 for validity
validity
IMPROVING VALIDITY
- in experiment - [control group, - increases confidence that chayes in DV anedue weffec of IV + srandardisarion.
- questionnaines - vie scales control Grether of social desirabilin bior + confidenialing is alsoned for all data submitted. pox ambigusol/ofes
- observarions - behavioural caregories ane operarionalised + mell defined
- qualitarine nesearch - interpretive validin + tiangularion
reporting psychological investigations
- ABSTRACT: short summary of the study including all major elements:
- aim, hypotheses, method, results + conclusion.
- INTRODUCTION: literature review - a look at relevant theories, concepts + studies related to study
- METHOD: should include sufficient detail for replication incl:
- design, sample, apparatus, procedure, ethics
- RESULTS: summary of key findings from investigation, including:
- descriptive stats: eg tables, graphs + charts, measures of central tendency + dispersion
- inferential stats: reference is choice of stats test, calculate + critical values, significance level + final outcome
- DISCUSSION:
- REFERENCING:
features of science
PARADIGM
- shared set of assumptions + methods
- KUHN suggested this was what separated scientific + non-scientific principles
- argued that psychology lacks a universally accepted paradigm + is best seen as a pre-science, unlike natural sciences eg biology
- paradigm shifts occur when there is a scientific revolution
- where a handful of scientists begin to question the accepted paradigm when there is too much contradictory evidence to ignore
features of science
THEORY CONSTRUCTION
- theory = set of general laws/principles that have the ability to explain particular events or behaviours
- testing a theory depends on being able to make clear + precise predictions on the basis of the theory (ie to stae a no of possible hypotheses)
- a hypothesis can then be tested using scientific methods to determine whether it will be supported or refused
- the process of deriving a new hypothesis from an existing theory is known as deduction
- bottom up + top down
features of science
falsifiability
- popper argued that the key criterion of a scientific theory is its falsifiability
- genuine scientific theories could hold themselves up for hypothesis testing + the possibility of being proved false
- popper distinguished between theories which can be challenged + what he called ‘pseudosciences’ which couldn’t be falsified
- he believed that even when a scientific principle had been successfully + repeatedly tested, it was not necessarily true
- instead, it had simply not been proved false (yet)
eg psychodynnamic approach
features of science
REPLICABILITY
- testing the validity of research results
- if a scientific theory is to be trusted, the findings from it must be shown to be repeatable across a no of different contexts
- by repeating a study, we can see the extent to which the findings can be generalised
features of science
OBJECTIVITY
to reduce bias in research
* scientific researchers must keep a ‘critical distance’ during research
* musn’t allow personal opinions/biases to ‘discolour’ the data or influence the behaviour of ppts
* methods associated w greatest level of control, eg lab studies, tend to be th most objective
features of science
EMPIRICAL METHODS
- emphasise the importance of data collection based on direct, sensory experience
- experimental + observational method are good examples of empirical method in psychology
- early empiricists, eg John Locke saw knowledge as determined only by experience + sense perception
- a theory cannot be claimed to be scientific unless it has been empirically tested + verified
the sign test
levels of measurement
probability + significance
statistical tests