research methods Flashcards
What are experiments AO1
- one variable is manipulated and its effect is measured
- determine cause and effect
- DV and IV manipulated
- field and lab experiments eg Steyvers and Hemmer or assessing reading ability in a classroom
field experiments strengths and weaknesses
- ecological validity
- if covert, less demand characteristics
- extraneous variables reduce reliability
- consent if covert
lab experiments strengths and weaknesses
- more reliable
- ensures causality - validity
- lack ecological validity
- often volunteer sample - demand characteristics
independent groups design strengths and weaknesses
- less likely to develop practise or fatigue effects if only doing one condition
- recruiting twice as many people - individual differences between groups
repeated measures design strength and weakness
- individual differences are reduced
- more likely to guess the aim and so show demand characteristics
matched pair design strength and weakness
- allow the groups to be compared fairly on relevant factors - reliable
- may not be possible to perfectly match ppts on all relevant factors
what are order effects and how can they be avoided
- performance of one condition is influenced by completing the previous condition - practise or fatigue effects
- avoided by counterbalancing, randomisation or leaving a time gap
what are extraneous variables and name 2 types
factors that may end up having an undesirable effect on the variable being measured, eg participant and situational variables
what are confounding variables
when EVs are not controlled and therefore have an effect on the DV , reducing the internal validity of the results
what are ppt variables and give an example
where ppts themselves affect the study results, for example if a ppt has a photographic memory and skews the results of a memory test
what are situational variables and give an example
environmental variables that may affect the results of the study for example the noise in the room while in a memory test
what is objectivity
the need to be impartial or judgement free regarding the analysis of results of a study, avoiding personal opinion.
what is subjectivity
the need to base a judgement on personal opinions, feelings or experience
what is reliability
the consistency of findings from research determines the reliability of the results
case studies AO1
- carried out on unique cases eg HM who was unable to make new long-term memories due to brain surgery to reduce his epilepsy
- aim is to see how brain damage affects cognitive function eg memory or perception
- used when unable to use experiments
- data is gathered from lots of sources and is then triangulated
case studies AO3
- hard to generalise
- lots of different methods make it reliable
- provides a full picture - valid
- researcher bias - get to know person
ethics dodge bc brain funny person cant give full consent all the time
2 strengths and weaknesses of using the mean
- sensitive and powerful measure as all scores used
- extremities are included - representative
- extremities may skew the results
- may not be an actual score eg 0.5 of a word cant be recalled
2 strengths and weaknesses of using the median
- not affected by extreme scores
- easy to work out
- not useful in a small group of numbers - inaccurate representation
- large sample of numbers will take a long time to work out
2 strengths and weaknesses of using the mode
- easy to calculate
- can be used on non-numerical data
- more than 2 modes makes it meaningless
- only takes into account common scores - unrepresentative
a strength and weakness of using standard deviation
- uses all scores to accurately measure dispersion, showing normal and skewed distributions
- hard to calculate manually
MODE IS ON THE LEFT, MODE IS ON THE LEFT, MODE IS ON THE LEFT, THAT’S A….
positive skew
difference between ordinal, interval and ratio data
- ordinal data can be ordered, the gap between values has no value
- interval data gaps have meaning, 0 has value eg temperature 0 degrees
- ratio data gaps have meaning, 0 has no value eg 0 seconds is nothing
if its a repeated measures design, the stat test will be ….
the wilcoxon
- experiment, repeated measures, at least ordinal data
if its an independent group design, the stat test will be …
chi square if nominal data or mann whitney if ordinal data, spearmans rho if its a corrolation
corrolation AO1
- looks for a relationship between two variables
- data gathered via self-report measures
- use the same ppt for two measures eg aggression levels and TV consumption
- two variables are of equal importance
- positive/ negative correlation
- coefficients show strength - +1 perfect pos correlation, -1 perfect neg corrolation
corrolations AO3
- can do unethical ones eg smoking/ lung cancer
- easy to see if there’s a relationship via scattergraph
- cant imply causation
- could be an extraneous variable instead
- self-report data
What’s the difference between a type 1 and 2 error?
type 1 = too lenient
Type 2 = too harsh
CAT scans AO1
- x-rays take b/w images of the brain
- person lies on a bed that moves through the gantry
- images can be compiled to form 3d image