Research methods Flashcards
6 types of observations
naturalistic vs structured
overt vs covert
participant vs non-participant
event sampling
observer records the number of times a certain behaviour occurs
Time sampling
observer records behaviour at perceived intervals
content analysis
typically used as a research tool to analyse the certain content for the incidence of certain words books filmks
usually use coding units
- enable comparisons and contrasts
- done in own time
- can be repeated
- gain inter-rater reliability
- some results subtle and missed
- subjective
Observations
no control
look for patterns
no cause and effect
inter rate reliability
extraneous variable
variable that needs to be controlled
situational variable
variable such as the environment
participant variable
affected by mood of participants
confounding variable
variables not controlled and effects results
counterbalancing
RMD when there is a problem with order effects
randomisation
IGD when there is a problem with ppts variables
internal validity
testing what you intent to test
external validity
extent to which findings can be generalised to other settings
ecological validity
extent to which findings still explain behaviour in different situations
predictive validity
able to accurately predict same future behaviour from results
concurrent validity
comparing questions to another test to see if they agree
face validity
seeing if the question makes sense
inter-rater reliability
assessed by comparing results from 2 or more raters
internal reliability
consistency of measure within itself
test-retest
measure of whether something varies from one time to another
researcher effects
researchers expectations (bias) may encourage certain behvaiours
demand characteristics
behaving in a way that they think should fit with what they perceive the aim of the study to be
nominal data and example
where data forms discrete categories e.g. hair colour
ordinal data
level of measurement when numbers are rankings
stratified sampling
when sample contains a proportional representation of target population
social desirability
respondent does not give genuine answer but one thats desirable to social norms
survey
self-report method to gather info
questionnaire
consist of questions
closed questions
fixed answers
4 BPS ethical guidelines
respect
responsibly
competence
integrity
hypothesis
statement about the effect of the IV on the DV
directional hypothesis (1 tailed)
specific outcome is clear
non-directional hypothesis (2 tailed)
not sure what the effect will be
null hypothesis
a statement of no difference
range
difference between the highest and lowest
- affected by extreme scores
measures of dispersion define
cluster of results
standard deviation
value that shows how scores are spread out around the mean
pros and cons of the mean
- takes into account all values
- affected by extreme scores
pros and cons of the median
- is not affected by extreme scores
- doesn’t take into account all values
pros and cons of the mode
- useful is data is in categories
- bad if there are many modes
pros of standard deviation
uses all the data collected in a set of results to calculate how much variation there is from the mean
therefore outliers have less of an effect on the dispersion
pilot study
small trial versions of proposed studies to test their effectiveness and make improvements
peer review
assessment of scientific work by others who are experts in the same field
- to ensure any research that is conducted and published is of high quality
the decision about whether to publish a research article depends on:
- the article being seen to make significant contribution to knowledge
- the strength of the methodology and results analysis
- usefulness of the conclusion that are drawn
- whether the study follows ethical guidelines
structured interview
pre-determined questions
semi-structured
some questions, pre-determined, others developed
unstructured
no questions are decided in advance
pros and cons to a random sample
each person has equal chance of selection
- unbiased
- can use a subgroup of population
- takes time
- may be biased if some decline to take part
pros and cons to stratified sampling
participants selected from subgroups in proportion to frequency in population
- most representative
- select subgroups, reduces EV
- selection of subgroups may be biased
- lengthy process and some may decline
pros and cons to volunteer sampling
advertise for willing participants
- committed participants
- specialised group
- volunteer bias
- more likely to response to demand characteristics
pros and cons to opportunity sampling
most easily available
- convenient and quick
- only option if can’t list whole population
- biased because only drawn from part of target population
- refusal means we end up with a volunteer sample
structured observation
staged observations that are usually carried out in an environment where researcher has some control
naturalistic observation
observing behaviour of participants in a natural environment
participant observation
observation when the researchers takes an active role in the situation being observed
non-participant observation
where researcher observes behaviours of others but does not form part of the group study
overt observation
a form of observation where those being observed are aware of the presence of the observer
covert observation
a form of observation where those being observed do not know
why is SD better than the range
The standard deviation takes account of all the figures in the data set individually whereas the range does not, it just takes the bottom score from the top score
The standard deviation can take account of the effect of outliers, and the range takes the most outlying scores into account but just the top and bottom scores
control group
a group of participants that does not experience the experimental situation but acts as a baseline against which to judge any chance
order effects
the problem with presentation order of stimulus material.
ppts may become practised at the test so improve performance or become tired so performance deteriorates
type 1 error
when the null hypothesis is rejected and alternative hypothesis supported when the effect was not real
(too lenient)
type 2 error
when the null hypothesis is accepted and alternative hypothesis rejected when there was actually a real effect
(too strict)
impact of levels of significance and errors
in psychology we set level of significance at 0.05 (5%)
- at a 0.1 (10%) level of significance there is a risk of type 1 error being too lenient
- at a 0.01 (1%) level of significance (very significant) there is a risk of type 2 being too strict
researcher bias
researchers performing the research influence the results, in order to portray a certain outcome
researcher effects
researchers expectations (bias) may encourage certain behaviours
what is thematic analysis
a way of analysing qualitative data and can be done inductively or deductively so record patterns within the data
inductively = researcher would read and reread the qualitative data and themes would emerge without researcher imposing any of their own ideas or expectations from it.
deductive = researcher specifying the themes that they will look for before analysing data
how do researchers conduct thematic analysis
researcher will develop themes into ‘codes’ which represent the categories of themes found and use codes to analyse data and search for instances where it appears in data (reviewed continually and changed if necessary until themes can be stated and supported).
pros and cons of thematic analysis
- more detailed and meaningful info than quantitative
- very little control over how thematic analysis is conducted
- unscientific as themes are highly dependant on the subjective opinions of researcher = researcher bias
e. g. preconceived ideas that affect theme choice and interpretation
response bias
tendency of a person to answer questions on a survey untruthfully or misleadingly
correlation
method used to assess the degree to which 2 co-variables are related
pros and cons of a correlation
- provides a means of looking at relationships and whether relationships are significant
- useful way to conduct a preliminary analysis as a weak correlation = no causal relationship
- cannot show a cause and effect relationship as there is not independent variable that has been deliberately altered
- If co variables are correlated one may be causing the changes in the other but we do not know the direction of the possible effect
- there may be intervening variables that can explain why the co- variables are linked
co- variables
variables investigated in a correlation
+1 correlation coefficient
perfect positive correlation
pros and cons to peer review
- allows for replication by other researchers
- criticisms of methodology mean original researchers revise their procedures and enhance validity
- anonymous so honest and subjective
- anonymous so researchers may try to bury or sabotage rival group
operationalisation
how you intend to measure the DV e.g. time taken
pros and cons to a lab experiment
- prevents EV affecting behaviour
- replication is easy
- high control
- no natural behaviour
- lacks ecological validity
pros and cons to a field experiment
- high ecological validity
- reduced demand characteristics
- replication is hard
- sample bias
- time consuming
- ppts unaware of participation
independent measures design
ppts allocated to different conditions
repeated measures design
ppts take part in all conditions
matched pairs
ppts matched in some way e.g. age, gender
pros and cons of quantitative data
- easier to analyse
- more objective
- may not express ppts precise thoughts
- oversimplifies human experience
pros and cons of qualitative data
- details of how people think and behave as they express
- represents true complexities of behaviour
- harder to detect patterns and draw conclusions due to large variety of data
- very subjective
primary data
directly by researcher
- up to date info
- high reliability
- time consuming
secondary data
evidence by other researchers
- quicker
- unaware of validity and reliability
Advantages of adoption studies
Disadvantage of adoption studies
- Removes extraneous variable; environment
- Useful in showing twin studies overestimate genetic factor
- Children adopted in to families similar to biological ones
- Adopted parents ‘unusual’ so not representative
Advantage of twin studies
Disadvantages of twin studies
- Look at influence of genes of MZ compared ti DZ
- Twin registries so large amounts of data
- Twins share same environment so nature VS. nurture
- Cannot tell which specific gene
Adoption study
leve et al 2010
A: see interaction between adopted childs genes and adoptive environment whilst seeing what interventions might be put to help child and families
S: 360 sets of triads
- children began school
- child adopted before 3 months
- no medical conditions
- families given chance to opt out
M: longitudinal study \9infancy/toddler
- face-face interviews, questionnaires, observations
- saliva samples 3-6months - 7yrs
R: children kept attention when adoptive mother had anxiety/depression
C: genetic issues in an adoptive child interacted with environmental issues. Used to inform interventions to help children and adoptive families
Twin study
Brendgen 2005
A: examine genetic and environmental effects on social and physical aggression using 6 year old twins
S: 234 twins, longitudinal, Canada, opportunity sample
84% european decent 16 ethnic minoriteis
P: 25.2% twins attended same classroom
Teacher = 2 questionnaire; social and physical
Social: ‘tries to make other child dislike child’
Physical: 3 point scale questionnaire ‘hit’ ‘bites’
Peer review= children given pics of classmates checked if students could identify pics and asked if which child fit description
R: 82% twins 1 nom for social
62% twins 1 nom for physical
- teacher more likely to find boys more physically aggressive girls more socially
- peers more likely to find boys more physically and social
