research methods - mainly year 1, terms etc Flashcards
what are the types of experiment?
laboratory
natural
field
quasi
what are lab studies? evaluate them
take place in highly controled environments where the researcher can control the conditions and variables
this is great for ensuring no extranuous variables but has poor generalisability because its not a real environment
what are natural studies? evaluate them
these take place in a naturally occouring event where the researcher takes advantage of a situation that wouldve happened regardles
youre likely to get authentic results because they act naturally in that environment and you can study thing =s you would never be able to do as its unethical, however you have no control over extranuous variables
what are field studies? evaluate them
these take place in a natural setting (meaning its usually covert) where the researcher manipulates variables
participants are more likely to be acting naturally but because of the setting its still hard to control the extranuous variables
what are quasi studies? evaluate them
the IV cannot be changed and no variables are manipulated but the researcher studies an existing change such as age or gender
these usually have the same strengths as a lab (control) but we cant randomly allocate participants
how do we always start when writing an aim?
to investigate whether
how can we decide which hypothesis to choose?
directional if previous research suggests a certain outcome so we can reasonably predict our study does the same
non-directional if theres contradicting research or no research pointing to a certain outcome
what are the two types of alternitive hypothesis
a directional / one tailed - this states the direction of the change eg itll be higher, lower, positive etc
a non-directional / two tailed - this states a difference but not how eg there will be a change…
what is the IV and the DV?
independent variable - thing thats changed
dependent variable - thing thats measured
what are extranuos variables?
any variable other than the IV that can effect the DV
what are the two levels of independent variable?
the experimental condition
the control condition
what are cofounding variables?
variables that interfere with the IV, resulting in having a second unintended IV. this means we cant be sure if our results are because of the actual IV or not
what are demand characteristics?
when the participant works out the actual aim of the study, so alters their behaviour and acting unnaturally
they may want to either please or screw the researcher, purposely answering the way they want them to or actively getting it wrong
what are participant variables?
variables the participant has that can affect research such as age or gender
what is randomisation and standardisation?
randomisation - assigning at random to reduce biases
standardisation - ensuring all procedures are the same for everyone
what is the difference between a correlation and experiment?
experiment: the researcher can manipulate and the iv that then effects the dv
correlations: no manipulation occurs and theres no cause and effect, just a relationship between co-variables
what types of correlations are there and what do they look like?
positive - as one increases so does the other
negative - as one increases the other decreases
no/zero - no relation is seen
what does the correlation coefficient +1, 0 and -1 mean?
+1 is a perfect positive correlation
0 is having no correlation
-1 is a perfect negative correlation
what are the different observational techniques and what do they mean?
naturalistic and controlled - natural setting vs lab setting
covert and overt - ptps are unaware vs unaware
participant and non-participant - researcher becomes part of the group vs stays seperate
what is time and event sampling?
time sampling: recording behaviours every eg 30 seconds
event sampling: recording everytime /counting the number of times the thing happens, everytime it does
what are the self report techniques?
questionnaires and interviews
what type of interviews are there?
structured (has pre determined questions)
unstructured (more conversational, no set questions)
semi-structured (mixture of both, has some guidence questions but can go off-piste when needed/ ask follow ups
what are the positives about each of the interview types?
structured interviews provide generalisable and analysable data
unstructured interviews are more casual meaning that social barriers are broken down, getting a more indepth understanding and rich data
what are open and closed questions? use an example
open ended questions produce qualitiative data as people are free to answer however, Eg why do you think that, describe, explain etc
closed questions produce quantitative data as answers have have fixed responses, Eg yes/no, scale of 1-10, are you likely, less likely, not likely etc
what are the pros and cons of using questionaires?
they are cost effective and can be distributed easily, gathering large amounts of information quickly
however answers may be untruthful, demand charicteristics like social desirability bias can effect the ptps answers or response bias’ that cause ptps to favour ‘yes’ answers
what are the three types of closed question questionnaires?
likert scales: strongly disagree - strongly agree
rating scales: from 1-10 etc
fixed-choice option: a list of tick boxes red, green etc
what types of biases can affect questionnaires and interviews?
social desirability bias
how can we ensure were writing good questions for interviews/questionnaires?
not overusing jargon (technical terms people dont generally know)
not using emotive language or leading questions, this prompts people to answer inauthentically
not using double barrelled questions or double negitives - this confuses people, they may not answer in the way they intended
what is content analysis?
a type of observational research where people are studied indirectly, through their communications (texts, graffiti, magazines etc)
what is ‘coding’ in content analysis? use an example
when the communication being studdied (the texts etc) are analysed. catergories are identified and everytime the thing occours its recorded.
eg in advertising. catergories could be proffessonal roles and family roles, its then recorded how many men and women are seen in each catergory.
what are case studys?
an in-depth investigation of one person or a group of people
what sampling techniques are there
volunteer: self picking from eg an ad
oppotunity: whoever is there and willing, eg classmates
random: randomly choosing from everying in thr target population
stratified: identifying your subgroups in the target population then choosing ptps to reflect the proportions
systematic: choosing every nth ptp from a list
evaluate these sampling techniques
volunteer: quick and easy but attracts spesific people so not representitive
opportunity: again quick and easy
random: fairly simple, although less than others can still be unrepresentitive
stratified: very representitive
systematic: very objective but time consuming
what 3 experimental designs are there
independent groups - two group do 2 seperate tasks
repeated measures - one group do both tasks
matched pairs - twins or similar people do seperate tasks
order effects can be a problem in repeated measures designs, why is this and how can we combat it?
participants can learn, get tired or work the study out
we can fix this with counterbalaning, this is when people do the conditions in different orders than eachother, so any effects are balanced out
what is a pilot study and why do we use them
a smaller ‘trial-run’ study, testing the design of the study is the most effective/ identifying potential issues
what is a peer review and why are they useful
the peer review is when experts/peers assess the methods and findings used, the validity of the research and how original it is. essentially anything from how its written to what it all means, to see if its worth publishing and what the implications of publishing may be.
what are standardised procedures
a way to control situational variables by ensuring that every participant recieces the same conditions, questions etc.
what is validity and in what two ways can we test it
this is if the research actually measures what it was meant to we can test it through:
concurrent validity - how much results from previous research correspond with eachother
face validity - does it measure the right thing on face value, ‘eyeballing it’
what are the 4 ethical guidelines
informed consent (this includes right to withdraw), deception, protection from harm, privacy and confidentiality
how can we get around the ethical guidelines on consent/ what other consents can we get instead
presumptive - asks a simular group
prior general - ptps give a general agreement before
retrospective - getting consent after
how can we deal with or get around the other ethical guidelines
do a cost-benifit analysis to see if the positives of the findings outweigh the negitives
ensuring ptps know they have the right to withdraw
doing a full debried after and giving access to further help if required
what is reliability and how do we test it
a measure of consistency, if we repeat the study do we get the same results, we can test with:
inter-rater reliability - the extent to which different observers agree
test-retest retesting the same ptp at a later time to see if they do the same thing
how can we improve reliability?
questionairres: test re-test
interviews: structured interviews and no leading questions
observations: operationalise behavioural catergories
experiemtns: use standardised procedures
what are the five features of science/ the things that we need to have in order for it to be scientific
have paradigms - a shared set of assumptions in society
have paradigm shifts - a change in these assumptions
be objective - no personal biases
use empiricle methods
be replicable - be replicable in other studies
be falsafiable - admits the possibility of being wrong, dand is able to be tested on this.
what are type I and type II errors
type one: a false positive, when the null is wrongly rejected and alternitive accepted
type two: a false negitive, when the null is wrongly accepted and the alternitive is rejected
how do we test for a causal difference between variables/ how can we test for what is causing the other and not just an assosiation
- must have an IV (the one thats changed, identifies a cause) and DV (the one thats measured, shows the effect)
- must be different IV’s (different conditions for the groups) so you can make comparrisons
- extranuous variables must be controlled
what is a single blind and double blind study?
single - when the ptp is unaware who gets the placebo
double - both ptp and researcher dont know
what are the two main types of hypothisis
the null hypothisis - predicts no difference
the alternitive hypothisis - predicts there will be a significant difference
when writing a hypothisis what feature should be included
must operationalise your conditions/variables.
this means you have to be spesific say ‘there will be a difference between those who do 10 minutes of meditation everyday for 2 weeks and those who do no meditation’ not just there will be a difference between conditions.
what are situational variables?
variables in the situation that can affect research such as time of day, room temperature etc
what are investigator effects?
any affect the researcher has on the participants, this could be smiling by accident or sighing, giving away what answers theyre looking for
what are correlation coefficients and what do they tell us?
a number between -1 and +1, telling us the strength and direction of the relationship between variables
how do correlation coefficients tell us the strength and direction of the correlation?
the + or - tells us if its a positive or negitive correlation
the number (-1,0,+1) tells us how strong.
the closer a value is to 1, the stronger the correlation
the closer the value is to 0, the weaker the correlation
what are the two ways of recording data in an observation?
structured: the researcher has a set list of things to observe, marking them with a tally
unstructured: the researcher writes down everything they observe
what are behavioural catergories?
when we make a list of observable catergories, such as if were studying the behaviour affection, we could break it into hugging, holdding hands etc
how should an interview room be set up in a 1-1 interview?
the interview should be in a quiet room away from other people to help them open up
the interviewer can ask some neutral questions to ease the interviewee in and build rapport
they should be reminded that their answers are confidential
what is thematic analysis?
this is a form of content analysis that produces qualitative data instead of quantitative.
it is a method of identifying patterns or themes found in the content analysis (recurrent catergories)
what three data types are in the ‘choosing a stat test’ table and what do they mean
nominal - catergorical data
ordinal - in a ranked order worst-best etc
interval - data with accepted intervals such as time, heart rate or some accepted eg depression tests
what is meta analysis?
examining data from a large range of studies all on the same/simular subject to identify any patterns or trends
what are the three descriptive statistics - caluclate and evaluate them
mean: add all and divide by the amount - affected by extreme values but most representitive
median: the middle number/halfway point - less representitive than the mean
mode: most common: good for catergorical data, not representitive
what are the two measures of dispersion and how do we caculate them
range: diff between highest and lowest - highest - lowest +1
standard deviation: tells us how far the scores deviate from the average, low value means the values are fairly simular, high value means there very spread out
what 3 graphs are there and when should the be used
bar chart: for catergorical data
scattergram: for correlations
histogram: for continuous data
describe the statistical/inferentical tests table
data on the left: nominal, ordinal, interval
headings at the top: unrelated, related, correlation
(unrelated - ind groups, related - matched+repeated)
chi squared - sign test - chi squared
mann whitney - wilcoxon - spearmens rho
unrelated T test - related T test - pearsons r
the correlation coefficient must be ‘what’ to show valid/significant data
+0.8 / +.8
in statistical tests what level of significance do we use
0.05 or 5%, this means theres a 95% chance its true and only a 5% chance it isnt.
how do you calculate a percentage
value/total x 100
what features are there in a report
abstract - short summary of the paper including aims, methods, results etc
introduction - lit review of the general theories in that area
method - all procedures, design, sampling, matereals used, ethics
results - descriptive statistics, data in graph form etc
discussion - verbal description of results including things you would better next time and any IRL implications
refferencing - all citations
by looking at data alone, how can we tell if it would show a normal or skewed distribution
in normal distribition the mode, median and mean are all at the point of the curve. the data would present as eg: 73.5, 74, 74.2
in skewed distribution the mode is always at the top, mean at the lower end and median in the middle
for a positive skew the data would look something like: mean:66, median:54, mode: 44 aka. the mean is the highest number and mode is the lowest
for a negitive skew the data would look something like: mean:32, median:40, mode:51 aka the mean in the lowest, and mode is the highest
