research methods Flashcards
mainly definitions / mark scheme
directional hypothesis
a hypothesis that states the kind of relationship or difference between two variables eg coffee will have an effect on reaction time on ppts
non-directional hypothesis
a hypothesis that predicts that there will be a difference / relationship but does not state the kind of difference eg coffee will change reaction time on ppts
laboratory experiment
an experiment conducted in a controlled situation
validity
testing or measuring what you set out to measure
advantages of an experiment
- manipulation of IVS can indicate cause and effect relationships
- increased control and measurement so higher validity
- standardised procedures mean replication is possible, so higher reliability
disadvantage of an experiment
- artificial conditions = unnatural behaviour/demand characteristics so lack both external and internal validity
- results biased eg in sampling / experimental bias, reducing internal validity
- cannot control all variables
- ethical problems of deception
strength of lab experiment
high in internal validity as extraneous variables controlled, so can be confident that any observed change in DV is due to the IV
disadvantages of lab experiment
participants aware = demand characteristics, reducing ecological validity
- IV/DV may be operationalised and so doesn’t represent everyday experiences - so low in mundane realism
- ppts uncomfortable in unnatural environment
field experiment
an experiment conducted in a more natural environment, and participants are usually not aware that they are in an experiment
- IV is still manipulated so casual relationships between IV and DV still demonstrated
field experiment strengths
- high in ecological validity since ppts unaware that they’re being studied so no demand characteristics
- natural setting = ppts more relaxed
field experiment weaknesses
- IV still controlled so may lack mundane realism
- more difficult to control extraneous variables
- major ethical issues - ppts unknown of being studied, so difficult to debrief them - raises ethical issues with manipulating and recording their behaviour
- more expensive and time consuming
natural experiment
IV is not manipulated, instead it is naturally occurring but still a measure of what you expect to see
independent groups design
a ppt only does one experimental condition
repeated measured design
a ppt does both experimental conditions
matched pairs design
ppt matched on any variable of researchers choice eg iq, ethnicity, age, then randomly allocated to one of the two conditions
ABBA effect
- two groups splitting sample
- in group 1 each ppt does condition A (morning) then B (afternoon), in group two each ppt does B then A
- counterbalances order effects
RMD strengths
- same task used in both conditions so less time consuming and less money
- does not need as many ppts as IGD
- set variable to match ppts to
IGD strengths
- not subject to order effects
- set variable to match ppts to
IGD weaknesses
- you need twice as many ppts as RMD to end up with the same amount of data
- cannot control ppt variables
RMD weaknesses
- order effects eg practice effect
MPD weaknesses
time consuming and difficult to match all ppt variables as you can only match on variables known to be relevant
- very difficult to match key variables
avoid weaknesses of IGD
- make sure no one is more skilled in one variable over the other - ppts can be randomly allocated to distribute variables more evenly
avoid weaknesses of MPD
restrict number of variables to match to make things easier
RMD avoid weaknesses
counterbalancing/ABBA
random sampling
every member of the population has an equal chance of being selected
opportunity sample
selecting those from the population that are most easily available at the time of study eg walking by you on the street
stratified sample
identifying subgroups within a population eg gender/age, and obtaining ppts from each subgroup in proportion to their frequency in the population
ppts are randomly chosen
(division of a population into smaller subgroups known as strata)
systematic sampling
sampling method where researchers select members of the population at a set/regular interval (every nth person)
volunteer sampling
obtaining of ppts relies solely on volunteers to make up sample by advertisement
random sample - strengths
- unbiased as all members of population have an equal chance
random sample - weaknesses
- find every person of a population and contacting them is time consuming
opportunity sample - strengths
- finding convenient/suitable ppts takes a short time
opportunity sample - weaknesses
- biased as ppts are drawn from a small part of the pop. rather than a general pop.
stratified sampling - strengths
- representative as a proportional selection of subgroups
stratified sampling - weaknesses
time consuming to find sub groups as well as contact and randomly select ppts from them
systematic sample - strengths
unbiased as an objective system is used for selection
systematic sample - weaknesses
not fully unbiased unless you select a number using a random method and start with that person, then select every nth person after
volunteer sample - strengths
representative as gives access to a variety of ppts
volunteer sample - weaknesses
- volunteer bias: volunteers have more time on their hands or need money unlike randomly selected ppts, so may be more highly motivated in the experiment
structured interview
- open questions for qualitative data interviews
- pre-determined questions
- conducted in real time with no deviation from structured questions
unstructured interview
- pre-determined questions, but interviewer develops new questions based on the answers given by the interviewee
questionnaire strengths
- researcher can collect data from a large sample of people as questionnaires can be distributed to large numbers cheaply and quickly
- respondents may be more willing to give personal information as not face to face
- can be repeated as questions are standardised
questionnaire weakness
only apply to people who can read, write and have the time to complete them. researches need to be careful of design, distribution and bias
structured interview strengths
- can easily be repeated as questions are standardised so answers from different people can be compared
- answers are more predictable, so easier to analyse answers than in an unstructured interview
structured interview weaknesses
- comparability may be a problem as the same interviewer could behave differently in different occasions or different interviewers behave differently
- interviewer bias - interviewers expectations may influence answers interviewee gives - you can avoid this by having skilled interviewers
unstructured interview strengths
- more detailed info obtained than in structured interview as interviewer tailors further questions building on a specific response
unstructured interview weaknesses
- due to development of new questions on the spot, this requires more skilled interviewers so more expensive to produce
- in-depth questions more likely to lack objectivity due to their instantaneous nature with no time for interviewer to reflect on what to say - may cause unnecessary detail and harder to analyse (rich qualitative data)
social desirability bias
people answering questions in such a way to present themselves in better light
interviewer bias
effect of an interviewers expectations communicated unconsciously on a respondents behaviour
pilot study
- a small trial run prior its actual study
- to find out if certain aspects of the design do or don’t work
- to see what needs to be adjusted without wasting time and money on a full scale study
- this improves validity as well as making a behaviour checklist
participant observation
the researcher takes part.
non-ppt observation
researcher just watches the ppts, does not take part
covert
one-way mirror is used to prevent ppts being aware that they are being watched, researcher can also hide
overt
ppts are aware that they are being watched
time sampling
recording behaviours for short time intervals
event sampling
all occurrences of behaviour are recorded constantly - decision only affected by how busy the situation is. if less busy, time sampling is easier
reliability
how consistent or accurate the findings of the study are
peer review
scientific work is judged prior application by others who are experts in the same field to ensure any research conducted is of high quality
having inter-rater reliability
two or more observers must be in agreement with their observations at 0.8 or greater
components of peer review: parliamentary office and science and technology (2002)
- allocation of research and funding
- publication of research in academic journals and books
- assessing the research rating of university departments
allocation of research and funding (PR)
research paid for by gov/charitable bodies - budget - 5.8 bil. so stuff like medical research council can decide which research is worth spending on, so require reviews - spending is done responsibly
publication of research in academic journals and books (PR)
peer review used in scientific academic journals as a means of preventing incorrect or faculty data entering the public domain
assessing the research rating of university departments (PR)
all university science departments’ research is assessed in terms of quality (by “research excellence framework”)
- future funding for the department depends on good ratings from the peer review
peer review evaluation:
- finding an expert to review isn’t always possible - this may lead to poor research being passed (smith 1999)
- reviews can be biased towards prestigious researchers
- publication bias - journals publishing positive results to increase standing of their journal due to positive implication - leads to a misperception of the true facts eg Ritchie et al (2012) submitted a rep of a study on paranormal phenoma and found the study was not even considered for peer review
- cannot ensure that all data we are exposed to is valid - results will always remain in the public even after being assessed, so rather be careful of what is put out there
meta-analysis
when a researcher combines results from many different studies leading to one conclusion
advantage of meta-analysis
- more generalisability as a larger amount of data is studied
- lots of evidence
disadvantage of meta-analysis
- publication bias: intentionally leaves out negative results and does not publish all data from relevant studies, meaning a false representation is given of what the researcher was investigating - lowers validity
nominal data
- data in the form of categories
ordinal data
- data that is represented in ranking form eg 1= hating maths and 10= loving maths
- lacks precision as it is based on the subjective opinion of people
interval data
- data that is based on numerical scales - based on objective measures
strengths and weaknesses of primary data
- targets the exact info of which the researcher needs so fits aims and objects
- it requires time and effort
- can be expensive
strengths and weaknesses of secondary data (collected by someone else other than researcher but used for investigation)
- data is accessed so requires minimal effort to collect
- data could be outdated or incomplete
- unreliable - researcher not there when study was conducted so likely to be unsure of validity of results
strengths and weaknesses of qualitative data
- richness and depth of detail
- allows ppts to further develop opinions - external validity
- meaningful insight
- difficult to analyse
- difficult to make comparisons
- researcher bias as conclusions rely on subjective interpretations of researcher (interpretative bias)
strengths and weaknesses of quantitive data
- can be analysed statistically so converted into graphs/charts
- easy comparisons
- lack of depth in detail
- no meaningful insight into ppts views
- ppts don’t develop their opinions so results are low external validity
S+W of questionnaires
- cost effective
- large amounts of data quickly
- researcher not needed to be present
- can analyse
- difficult to know whether target population was answered eg if online
- take long to design
- ppt bias eg age. time, gender
- response bias instead of putting in effort
dealing with ethical issue of informed consent
- prior general consent: ppts given to take part in many studies whereby one of them involves deception, so consenting to getting deceived
- presumptive consent: when a researcher gathers opinions from a group like the ppts but does that inform actual ppts so demand characteristics eliminated
- retrospective: researcher asks for consent after participating in the study
dealing with deception
- debriefing: all ppts debriefed after the study - can be written or verbal. during it, true nature of study must be said and ppts should be told what their study will be used to. after, ppts have the right to choose to withhold or withdraw their data
dealing with protection from harm
- researcher could provide counselling if ppts subject to any stress or pyschological harm
- cost-benefit analysis should be done before study - ethics committee weigh pros and cons to determine whether study is ethical
dealing with privacy and confidentiality
- anonymity can be maintained - so not recording an personal details of ppts so data isn’t traced back to them - could refer to ppts with numbers or initials instead when writing up investigation
- ppt should be reminded of briefing and debriefing of investigation that their data will be protected
nomothetic approach
where general laws and principles on human behaviour are made - done using a scientific method reducing objective data which is quantifiable, very often using large samples
research AO3 - what to remember when writing A03 on research
- reliability - consistency
- ethics
- determinism vs free will
- reductionist / hollistic
- nomothetic (does it agree with science?)
what ethics to use during evaluation
- informed consent
- protection from harm
- deception
- privacy (human rights) and confidentiality