research methods Flashcards
what is an aim?
a general statement of what the researcher intends to investigate - purpose of the study
what is a hypothesis? what is a directional and non-directional hypothesis?
- statement made at start of a study and clearly describes relationship between variables as stated by a theory
- needs to make it easy to tell what IV and DV are
- directional - researcher makes clear sort of difference anticipated between 2 conditions or two groups of people - include words like more/less, higher/lower, faster/slower
- non-directional - simply states difference between conditions/groups
what are independent and dependent variables? how should other variables potentially affecting the DV be treated? why?
- in experiment - researcher manipulates independent variable + records/measures effect of change on dependent variable
- all other variables that might potentially affect DV should remain constant in a properly run experiment
- so researcher can be confident any change in DV due to IV alone
why should we check levels of IV? how would we?
- in terms of energy drink. if simply gave ptps speedup, how would we know how talkative they were?
- in order to test need different experimental conditions - need a comparison
- could either: compare ptps talkativeness before + after drinking
- compare two groups: those who drink energy drink + those who drink water
- the two levels of IV - control condition (no energy drink/drink of water) and experimental condition (energy drink)
how do you operationalise variables in a hypothesis?
- make the variables measurables
- ex. after drinking 300ml of energy drink, ptps say more words in next 5 mins than ptps who drink 300ml water
what are extraneous variables? what are examples? what are their effects?
- independent variable is manipulated to see how it affects the DV
- only thing that should affect DV is IV
- any other variables that might potentially interfere with IV should be controlled/removed
- additional unwanted variables called EV
- age of ptps
- lighting in lab
- ‘nuisance variables’ do not confound findings - may just make it harder to detect result
what are confounding variables?
- do change systematically with IV
- can’t tell if any change in DV due to IV or confounding variable (almost like a second unintended IV)
what are demand characteristics?
- any cue from researcher/from research situation - may be interpreted by ptps as revealing purpose of investigation
- may lead ptp changing behaviour within research situation
- may act in a way they think is expected + overperform (please-U effect) or opposite to sabotage results (screw-U effect)
- unnatural ptp behaviour - extraneous variable may affect DV
what are investigator effects?
- any effect of investigator’s behaviour (conscious/unconscious) on research outcome (DV)
- can include everything from design of study to selection of + interaction with ptps during research process
what is randomisation?
- minimise effect of extraneous/conf variables on outcome
- use of chance methods to reduce researcher’s unconcscious bias when designing investigation
- controls investigator effects
- in exp where ptps involved in number of conds - order of conds should be randomly determined
- if all ptps were to take part in all conds - order of completion would need to be randomised for each ptp
what is standardisation?
- all ptps should be subject to same environment, info + experience
- to ensure all procedures standardised - list of exactly what will be done in the study
- standardised instructuctions that are read to each ptp
- such standardisation means non-standardised changes in procedure do not act as extraneous variables
what is an experimental design?
different ways in which ptp’s can be organised in relation to the experimental conditions
what are the different experimental designs?
- independent groups design
- repeated measures
- matched pairs
what is an independent groups design?
- when two seperate groups of ptp’s experience two different conditions of the experiment
- if there are two levels of the IV - means all ptps experience one level of IV only
- performance of two groups then compared
what is an example of an independent group?
- one group of ptp drinking energy drink (condition A - experimental condition)
- a different group (group2) drinking water (condition B - control condition)
- performance compared : mean number of words spoken in the 5 min period after drinking for each group/condition
what is a repeated measures design? use the energy drink example
- all ptp’s experience both conditions of experiment
- ex. each ptp would first experience condition A (energy drink condition)
- ptp would then later be tested again in condition B (glass of water condition, control condition)
- following this, the two mean scores from both conditions would be compared to see if there was a difference
what is a matched pairs design?
- ptp’s paired together on a variable/variables relevant to experiment
- ex. memory study - might be matched on IQ - good indicator on ability to recall
- two ptps with first + second highest IQ paired together + the one with 3rd and 4th highest and so on
- then one ptp from each pair allocated to a different condition of the experiment
- attempt to control for the confounding variable + ptp variables
how would matched pairs be used in the energy drink example?
- observe ptp’s interacting in a room before ex begins + select 2 which most chattiest - one pair placed in condition A + other condition B and so on
- then ex will run in same way as indep group design
what is an issue with independent groups when it comes to ptp variables? how can it be dealt with. what is an issue of it compared to RMD?
- ptp’s who occupy the different groups not the same in terms of ptp variables
- if researcher finds a mean difference between groups on dependent variable - may be more to do with ptp variables than effects of IV
- may act as confounding variable - reducing validity
- to deal with it they use random allocation
- less economical than repeated measures - each ptp contributes a single result only - twice as many ptp needed to produce equivalent data to that collected in a repeated measures design - increases time + money spent on recruiting ptp
what are strengths of independent groups? compared to RMD?
- order effects not a problem - whereas problem for repeated measures design
- ptp’s less likely to guess aims
what are the issues of the RMD? explain effects of order using energy drink example. how else can order effects happen? how can an order effect be a confounding variable?
- each ptp has to do at least two tasks + order of these tasks may be significant (order effects)
- energy drink: having energy drink first may have a continuing effect when ptp drinks water afterwards
- could arise because repeating two tasks could create boredom/fatigue - might cause deterioration in performance on second task - order matters
- alternatively - ptps performance may improve through effects of practice - esp on a skill-based task - in this case ptps will perform better on second task - order effect confounding variable
how is demand characteristics a problem for RMD?
- more likely ptp will work out aim of the study when experience all conditions of experiment
- hence why DC tend to be more of a feature of RMD than indep groups
what is a strength of RMD?
- ptp variables controlled - therefore higher validity
- fewer ptp needed (less time recruiting)
what are strengths of matched pairs?
ptp only take part in single condition - order effects less of a problem
what are disadvantages of matched pairs?
- although some attempts to reduce ptp variables - ptp can never be matched exactly - even when identical twins used - will be important differences between them may affect DV
- matching may be time-consuming + expensive - particularly if pre-test required - less economical than other designs
what are the different types of experiments?
- laboratory experiments
- field experiments
- natural experiments
- quasi-experiments