7 - research methods Flashcards
laboratory experiment key features
highly controlled, artificial environment, replicable, random allocation
laboratory experiment strengths
- high control over extraneous variables
- easy to find cause and effect
- can be replicated
- measured accurately
laboratory experiment limitations
- data may lack ecological validity
- high risk of demand characteristics
- alter behavior based on experiment
- experimenter bias
field experiment key features
natural setting, some variables cant be controlled, independent variable is changed
field experiment strengths
- high ecological validity
- less likely to have demand characteristics
field experiment limitations
- less replicable
- sampled bias
- lack of control on extraneous variables
- lack of informed consent (ethical)
natural experiment key features
independent variable cant be changed e.g. boys/girls, no random allocation
natural experiment strengths
- high ecological validity
- low demand of characteristics
natural experiment limitations
- sample bias
- ethical issues
quasi experiment key features
naturally occurring independent variable e.g. age, gender
quasi experiment strengths
- allows comparison between types of people
quasi experiment limitations
- can only be used for naturally occurring iv’s
- ecological validity may be reduced as task may be fairly artificial
what is the difference between a directional and non directional hypothesis?
directional - says which way the hypothesis will go e.g. more than
non-directional - there will be a difference
what does operationalised mean?
variables and how they will be measured must be clear in the hypothesis
what is a null hypothesis?
states there will be no significant difference/correlation in results
what are the 4 types of extraneous variables?
demand characteristics
investigator effects
participant variables
situational variables
what are demand characteristics?
participants may want to be helpful and act differently to fit in with the experiment
what are investigator effects?
cues from the investigator that encourage certain behaviours e.g. tone of voice
what are participant variables?
the natural differences between people e.g. age or gender
what are situational variables?
features of experiment may make people respond differently e.g. temp, noise, tiredness
What are the 6 ethical guidelines for experiments
1 - informed consent
2 - deception
3 - right to withdraw
4 - protection form psychological and physical harm
5 - confidentiality
6 - privacy
what are the ways of dealing with ethical issues in the UK?
1 - cost-benefit analysis
2 - ethics committee
3 - punishment
How is the Canadian approach to ethics different?
It stimulates debate, encourages psychologists to engage deeply with ethical rules rather that just follow guideline
Issues with informed consent?
giving away aims may reduce validity
presumptive consent - can other people consent for you?
Issues with deception?
Cost-benefit analysis is subjective
Debrief cant always undo harm
Issues with Right to Withdraw
participants may feel obligated to continue, especially if there is money
issues with protection from harm
harm may not be apparent until after
issues with confidentiality
may be identifiable even without names e.g. through schools or organisations
issues with privacy
there is no universally accepted definition of a public or private place
what is random sampling
positives and negatives
- when all participants have an equal chance of being selected e.g. names in a hat
positives - unbiased, all pp’s have equal chances
negatives - need to have a list of all members, time consuming
opportunity sampling method
positives and negatives
- recruit those who are easiest and most available e.g. people walking past
positives - easy, not time consuming
negatives - biased (small population) or only people who don’t work at certain times for example
What is stratified sampling
positives and negatives
subgroups are identified and a proportional number of pp’s are taken from each group
positives- more representative
negatives- very time consuming to identify, select, and contact participants
what is systematic sampling
positives and negatives
predetermined system to select people e.g. every 3rd person
positives- unbiased and uses an objective system
negatives- not truly random unless selected using a random method
what is volunteer sampling
positives and negatives
advertised in newspaper / notice boards / internet
positives - access to a variety of participants e.g. people who read certain papers
negatives- biased as some participants are more motivated (volunteer bias)
what are the benefits of a pilot study?
- identifies potential issues early
- tests effectiveness and can make improvements
- more likely to get meaningful results
- reduces wasted time and money
- able to reduce demand characteristics
methods for quantitative data
- experimental methods with a quantitative dependent variable
- closed questions
- tally of behaviours
- content analysis of descriptive material
methods for qualitative data
- open questions and interviews
- description of behaviour
- descriptive material
strengths of quantitative data
- easy to analyse
- usually objective
- conclusions can be drawn
weaknesses of quantitative data
- over simplifies complex behaviour
- closed questions force answers
strengths of qualitative data
- reflects complexities and different experiences
- encourages description
- shows feelings
- data isn’t restricted
weaknesses of qualitative data
- subjective so it’s difficult to draw conclusions
- hard to analyse
- not as easy to compare results
what is primary data?
data collected first hand directly by the researcher
strengths of primary data?
researcher has control over the data
very specific so meets the aims of study
limitations of primary data
- long time to carry out
- involves a lot of expenses: designing, recruiting, conducting, and analysis
what is secondary data
data collected for a different purpose but is utilised and reused by someone
e.g. from government data
strengths of secondary data
- takes less time, equipment and money to access
- already been analysed so statistics have been decided
limitations of secondary data
- previous studies and data may not fit the aims of the study
what is a meta-analysis
a review of similar studies and combines results to form new results
strengths of a meta analysis
- increases validity of findings as there is a wide range of participants
- if some conclusions are very off, they can be controlled
limitations of a meta analysis
- research design may vary between studies involved, can’t be truly compared
- effect size may not be appropriate
- invalid conclusions
what are independent groups?
each condition has it’s own group of participants. The scores (DV measurements) for each group are compared.
problem with independent groups and how to control it
- natural variations between individuals in each group may affect the DV measurements, making it look as if the IV has had an effect when it actually hasn’t
control
- after the pp’s have been recruited, they should be randomly assigned to their groups
- they should ensure the groups are similar on average
what are repeated measures
only one group of participants is used.
The group completes both conditions and participant
scores in each condition are compared.
problems with repeated measures and how to control them
problems
Doing both conditions may
(1)give PPs practice on the task;
(2) make them bored or tired;
(3) allow them to work out the aim of the study, all of which might affect the DV measurement.
(4) Reuse of stimulus material is not possible
Control:
Divide the PPs into two groups.
Half does condition A first, then condition B. The rest do condition B then condition A.
DV measurements for the conditions A
and B are then compared
(counterbalancing).
what are matched pairs?
pairs of participants are selected, who are as similar as possible.
One member from each pair completes each condition.
The scores for the pairs of participants are compared.
problems with matched pairs and how to control them
Problem:
(1) time consuming;
(2) an exact match is rarely possible;
(3) if one PP drops out you lose 2 PPs’ data.
Control:
Members of each pair should be randomly assigned to conditions.
However, this does not solve all these problems.
what is a non-experimental research technique
- watching and recording behaviour
- natural behaviour in people’s natural environment
why use observations?
- study a particular behaviour
- study natural behaviour
- study behaviour where it would be unethical to manipulate the IV
what are controlled observations?
- pp’s are likely to know they’re being studied
- some variables are manipulated by researcher
- may take place in a lab
what are naturalistic observations?
- observing people in their natural environment
- people behave freely and are less likely to know they’re being observed
- researcher doesn’t interfere
what is the difference between a participant and non-participant observation
participant - observer acts as part of the group being watched
non-participant - observer isn’t part of the group being observed
what is the difference between overt and covert
overt - ‘open’ observations where pp’s know they are being observed and why
covert - pp’s don’t know they’re being observed, observer is ‘under cover’
general strengths for ‘observations’
- have high validity: record how people actually behave rather than how they say they behave
- capture spontaneous and unexpected behaviour
- can be used to measure the DV in an experiment, so are a key method of gathering data
general limitations of ‘observations’
- observer bias is likely as it is difficult to be objective
- only observable info is recorded so doesn’t provide insight about how people think or feel
what needs to be taken into consideration about clarity when it comes to questionnaires?
- the reader must be able to understand what is being asked
- no ambiguity
- double negative questions reduce clarity
- double barreled questions are bad e.g. do you experience pain AND headaches
what needs to be taken into consideration about bias when it comes to questionnaires?
- may leaf respondent to be more likely to give a particular answer
- hard to stop changing answers for social desirability bias
what needs to be taken into consideration about analysis when it comes to questionnaires?
- it is best to use closed questions with quantative data
what are 4 good extra things to add into a questionnaire?
- filler questions
- sequence for questions
- sampling technique
- pilot study
4 strengths of questionnaires
- once you’ve designed a questionnaire, you can use them for a lot of people, cheaply and quickly
- may be more likely to share personal info rather than if they are being interviewed
- reduces experimenter bias
- no special training needed to conduct
what is experimenter bias?
where the participant changes their answers due to the unintentional influence of the experimenter
4 limitations of using questionnaires?
- can take lots of time to design
- can only be filled by people who can read and write
- biased due to those willing to fill it in
- social desirability bias - not wholly truthful
what are features of structured interviews?
- predetermined questions
- no deviation from schedule or questions
- conducted in real time
2 strengths of structured interviews
- it can be easily repeated as the questions are standardised - so easier to compare
- easier to analyse than an unstructured interview as answers are more predictable
2 limitations of structured interviews
- if the same interviewer behaves differently on different occasions, there is low reliability and harder to compare
- interviewer’s may influence the answer given (interviewer bias)
features of an unstructured interview
- questions are developed during the course of the interview based on the answers given
- may start with predetermined questions
- sometimes linked to clinical interviews such as a doctors appointment
strength of unstructured interviews
- more detailed info can be obtained as the interviewer tailors further questions to the specific response
2 limitations of unstructured interviews
- requires interviewers with more skill as they have to develop on the spot. So more expensive to train
- in depth questions may lack objectivity because of their instantaneous nature
what are often effects of the interviewer?
the amount of interest and the presence of an interviewer may increase the amount of information provided
non verbal communication from interviewer
- arms crossed, frowning, yawning etc may discourage respondent from speaking freely
- nodding and smiling would encourage
features of correlations
- looking for a relationship between 2 sets of data
- most be continuous
- uses co-variables rather than IV/DV
- numbers relate to behaviour
- coefficient is between -1 and 1
what is a directional hypothesis?
correlation
a hypothesis that predicts the directions of the relationship or difference
what is a non-directional hypothesis?
correlation
two tailed hypothesis that doesn’t predict the direction of the difference/correlation
what is a null hypothesis?
a hypothesis that states there will be no correlation and the statistics are independant
strengths of correlations?
- procedure can usually be repeated easily and if findings are similar, research is likely reliable
- correlations usually have high ecological validity as data comes from natural settings and real life
- correlations can investigate trends/patterns in data
- do not require manipulation of any variables so less likely to cause ethical issues
limitations of correlations
- correlation does not show causation
- they only show linear relationships not curvilinear
- if a casual relationship is shown, it may be due to intervening variables which are unknown variables that can explain why the covariables are links (ice cream sales and crime)
characteristics of observations
- has a hypothesis
- IV isn’t manipulated
- operationaalise variables
- has a plan
advantages of unstructured observations
- provides a great deal of rich, qualitative data
- can be useful as a pilot study for a more structured observation
disadvantages of unstructured observations
- can be difficult to analyse
- can be a tendency for observers to notice the most eye-catching behaviours which may not be most relevant
what are features of a structured observation?
- design a type of coding system (behavioural categories) to record pp’s behaviour
- often made to record how often a type of human behaviours appears
- data collected is quantitative
what is the difference between event sampling and time sampling?
event sampling - recording every time a particular behaviour occurs
time sampling - recording observations at specific time intervals e.g. recording identified behaviours every 30 seconds (not including any behaviours outside the interval)
how to make an observation more consistent and reliable
- have another person doing the same observation at the same time then comparing data and results after
- +0.8 correlation = degree of reliability
inter-observer
what is intra-observation
same observer does the observation two or more times
how to improve the reliability of observations
- clearly operationalise categories
- train observers correctly
- pilot studies before
features of a positive skew?
- long tail on the positive side of the peak
- the mean is higher than the mode and median
features of a negative skew
- mean is lower than the mode and median
- long tail on negative side of the peak
what is nominal data?
- quantitative value
- represents discontinuous data
- no overlap between categories
e.g.
eye colour, house, marmite opinion
most useful = mode
what is ordinal data?
- shows order of values, but the difference between the values isn’t fixed
e.g.
the difference between 1st and 2nd position isn’t the same as 2nd and 3rd
most useful = mode and median
what is interval data?
- numeric, we know the order and exact differences between the values
e.g.
time in seconds, or distance in cm
mode, median, or mean can all be used
what are advantages and limitations of using a mean
advantages
- takes into account the exact distances between all the values of the data
limitations
- can’t be used with nominal data
- can be easily distorted by one or a few extreme values
what are advantages and limitations of using a median
advantages
- not affected by extreme scores
- appropriate for ordinal data
- easier to calculate than the mean
limitations
- not as ‘sensitive’ as the exact values are not reflected in final calculation
what are advantages and limitations of using a mode
advantages
- unaffected by extreme values
- useful for discrete data
- can be used for category data
limitations
- not useful when there are several modes
- tells us nothing about the other values
what is peer review?
- an assessment of scientific work by others wo are experts in the same field
- judge the scientific quality of research
- enable poor scientific practice, or fraudulent work to be identified and prevented from being published
what are parts of the scientific process?
- assess the appropriateness
- check the researches validity
- to judge the significance of the research
- check if the research is original
- provide or suggest recommendations
- publish
why are peer reviews needed?
- ensure it is high quality research
- helps allocate research funding
- contributes to the research rating of university departments
- spot out fraudulent work
2 examples of badly reviewed work
claim that MMR vaccine leads to autism
Cyril Burt claimed that intelligence is inherited
why was Cyril Burt’s work said to be fraudulant?
- he made up lab assistants who didn’t actually exist
- made up data
AO3 of scientific processes
peer review
Limitation
Already published research
P - there are some issues with peer review
E - Although it is very important to prevent false information, it can’t solve the issue of already published research.
Once it has been published, the results remain in the public view, even if it were found to be fraudulent
- therefore, it doesn’t ensure information we are exposed to is valid for example Cybil Burt
L - this emphasises the importance of the general public to be more critical about research claims and no always to accept them
AO3 of peer review
preserving the status quo
P - peer review results in a preference for research that goes with existing theory rather than dissenting or unconventional work
E - Richard Horton made the following comment ‘the mistake is to have thought that peer review was any more than a crude means of discovering the acceptability - not the validity - of a new finding
E - Journals should be aware of the damaging effects of such bias
AO3 of peer review
finding an expert
P - it isn’t always possible to find an appropriate expert to review a research proposal or report
E - This means poor research may be passed on if the reviewer didn’t really understand it. Also, reviewers may be bias towards the prestigious researchers rather than the less well known names
E - Emphasis the need for rigorous approaches
AO3 of peer review
Anonymity
P - reviewers may have their identity kept a secret
E - This aim lets them to be completely honest and objective
But, reviewers may use the veil to settle old scores or bury rivals
E - to avoid this, some journals favour open reviewing
How can you apply psychology to the economy to psychopathology?
McCrone report estimated the direct costs of mental health in England is £22.5 billion per year
Using drugs rather than psychotherapies provides more economic gain as it helps patients to return back to work sooner
How can you apply psychology to the economy to memory?
Eye witness testimony research improves crime detection and reduces expenses spent on wrongful arrests
How can you apply psychology to the economy to social influence?
Campaign to reduce drink driving or smoking discussed where attitudes and behaviour was changed making people aware
Leads to potential to bring about positive change so people fit into the norm
Tax forms
How can you apply psychology to the economy to attachment?
Bowlby’s theory on attachment showed the importance of emotional care in early child development. Continuing and ensuring the healthy development of children to become more productive members of society and thus improving world economy.
Why is statistical testing used in psychological research?
- to infer differences or relationships between 2 or more groups
- helps draw conclusions
- test if a difference is significant
What is the probability level on psychological research unless otherwise stated?
0.05 (5%) chance that the result will occur if the null hypothesis is true
Use 0.01 in more life impacting research
What is N on the critical values table?
Number of scores or participants
What is the p level on a critical values table?
probability level
what is a one tailed hypothesis?
A directional hypothesis
What is a two tailed hypothesis?
A non-directional hypothesis
What are the 3 D’s when deciding and justifying the choice of inferential statistical tests?
Data (what level of data? nominal, ordinal, interval?)
Difference (difference or relationship?)
Design (what experimental design?)
What D’s do you use when using the sign test?
Data = nominal
Difference = difference
Design = repeated measures or matched pairs
What are the 5 steps of the sign test?
- Work out the ‘sign’ - whether the score has gone up or down
- calculate the value of ‘s’ - add up pluses and minuses and select the smaller value. This is the calculated value
- Calculate the value of N - the total number of scores excluding any nil scores
- find critical value of S - using critical values table
- determine whether results are significant - if the calculated value of S is less than or equal to the critical value of S, it is significant
What are descriptive and inferential tests?
descriptive - measures of central tendency (mean, median, mode) and dispersion (range, standard deviation)
inferential - used when comparing results with probabilities to decide if results are significant. They can accept or reject null hypothesis
What is the parametric criteria?
- level of data must be interval
- data population with a normal distribution (can’t be skewed)
- similar range and standard deviation (variance of 1 sample should not be more than 4 times the variance of the other
what is content analysis?
- quantifying qualitative information through the use of coding units
- used to statistically analyse verbal or written or visual material
what are the steps involved in content analysis?
- decide on sample used
- read source, identify any themes, break it down into a coding system
- read again and tally number of times coding unit occurs
- assess and improve reliability
- analyse findings through quantitative means e.g. descriptive statistics
strengths of using content analysis
- replicable method as coding units aren’t open to interpretation
- easily obtained materials, no pp’s needed
- high ecological validity as it observes people in real life settings or communications
- allows statistical analysis to be conducted
limitations of using content analysis
- observer bias if coding categories are ambiguous
- time consuming if carried out correctly
- cultural bias where content is interpreted differently
- causation can’t be established, it only reports data
what is thematic analysis?
qualitative way of describing and interpreting data through the use of themes
what are the intentions of thematic analysis
- impose some kind of order on the data
- ensure order represents the pp’s perspective
- ensure order emerges from data rather than any preconceptions
- summaries data so hundreds of pages/ hours can be reduced
- enables themes to be identified and general conclusions to be drawn
when is thematic analysis useful?
when finding out people’s views, knowledge, experiences, values
what are the stages of thematic analysis?
- familiarisation ( get to know the data)
- coding (look for/ break down data into useful units and codes)
- reviewing themes ( return to data and compare themes against it to ensure they are a good representation of the data)
is anything missing? are these themes really visible? - defining and naming themes (formulate exactly what is meant by each theme)
strengths of thematic analysis
- shows key themes
- can extend analysis
- high ecological validity
- no pp’s needed, no problem with availability
- replicable
limitations of thematic analysis
- subjective nature so open to bias from researcher
- time consuming
what is a case study?
- in depth study of individual, group, event, or institution
- involves input from family, and others with contact
- findings are organised to represent the individuals’ thoughts emotions, experiments and abilities
- longitudinal
strengths of case studies
- rich and detailed qualitative data as it focuses on individual throughout a long time
- validity and ecological validity - high for that one individual or group as the research is carried out in their real life setting
- avoids practical/ ethical issues - case studies can shed light on aspects of behaviour that couldn’t be set up as research
limitations of case studies
- subjectivity - researcher’s interpretation of findings may be different to one another
- generalisability - hard to generalise to a wider group
- time consuming - often longitudinal studies which are carried out over a long period of time
- lack of replication - results can’t be validated by other studies since they are so personal and unique
what is reliability?
the extent to which a test produces consistent findings every time it is done
what are the 2 methods of testing reliablitiy?
- interobserver reliability - getting another observer to do the same thing and compare the correlation between results (must be over 0.8 to be significant)
- test-retest - repeat the test with the same pp’s at a different time
how can reliability be improved in lab experiments?
standardise procedures
how can reliability be improved in observations?
- operationalise behavioural categories more clearly
- practice using categories so observers can respond more quickly
how can reliability be improved in questionnaires?
- reexamine questions and avoid/remove and ambiguity
how can reliability be improved in interviews?
- interviewer should be trained to avoid asking leading questions
- reexamine questions asked
What is validity?
Refers to whether the study or measuring test is measuring what it claims to measure
What are the 2 types of validity?
internal - questions whether the results are due to the IV
external - questions whether the results can be generalised
what are the ways which internal validity can be affected?
Demand characteristics - cues that communicate the aims of the study to participants
Investigator effects - anything the investigator does that impacts the pp’s e.g. encouraging them to try harder
confounding variables - a variable that changes alongside the IV so it is hard to tell what impacts the DV
Social desirability bias - pp’s not being completely to show themselves in a more socially acceptable way
what are the examples of external validity?
Ecological validity - being able to generalise findings to everyday life and making sure the conclusions would be made in a real life situation
Temporal validity - ability to generalise a research effect beyond the particular time period of the study
Population validity - ability to generalise findings to a wider group of people
What are the 2 ways that validity can be tested?
Face Validity - looking at something and assessing the extent to which a test or measure appears, on the service, to assess the intended construct
Concurrent validity - comparing new procedure with a similar one that has already been done. If scores correlate as a strong positive (0.8+) the test is probably valid
How to improve validity in experimental methods
Use a double blind procedure to reduce demand characteristics
How to improve validity in observations
Non-participant and covert observations should be used so the observed behaviour would be more natural and valid
How to improve validity in questionnaires
- questions should be revised so they are related more obviously to the topic
- If concurrent validity is low, remove irrelevant questions
- Ensure anonymity
- Use filler questions to avoid DC
How to improve validity in interviews
use the same interviewer to reduce investigator effects
What is empiricism?
Data collected through direct observation or experiments
what is objectivity?
researchers should not be influenced by personal feelings, opinions, experiences
Remains unbiased
High levels of control
Not open to interpretation
what is replicability?
- the ability to repeat a method
- relies on standardised procedures
- helps determine validity
what is falsibility?
- can’t be scientific unless it admits the possibility of it being found false
- not necessarily true if it hasn’t been found false
- ‘fine tunes’ theories to make gradual changes
What did Popper say about falsibility?
It is not possible to confirm a theory, it is only possible to disconfirm it
what is a paradigm
An unidentified set of assumptions and agreed methods within a theory e.g. theory of evolution
what is a paradigm shift?
Gradually disconfirming evidence accumulates and the dominant theory is overthrown
‘scientific revolution’
what is theory construction?
A theory needs to be a logically organized set of propositions that defines events, describes relationships among events, and explains and predicts the occurrence of events. A scientific theory should also guide research by offering testable hypotheses that can be rigorously tested.
what is the difference between inductive and deductive theories?
deductive - theory proposed before hypothesis testing
inductive - theory proposed after hypothesis testing
- important in entirely new research
What should be included in an abstract?
approximately 150 words of a summary / snapshot of what is covered in the work
what should be included in the introduction of a psychological report?
literature of previous studies which are relevant
What is the funnel technique?
narrowing down evidence / relevant studies
what should the introduction of a psychological report end with?
aim and hypothesis
summary of which evidence will be used to back up and why
how is an aim different to a hypothesis?
aim depicts what they plan in an experiment to do and a hypothesis states a prediction of the results
what should the method section of a psychological report include?
- design
- participants
- material
- procedure
- ethics
what should the discussion part of a psychological report include?
- summary of findings
- compare results
- any limitations and possible suggestions
- implications in real life
how should book references be written?
Author A.A. (year of publication) title of work. Location : puublisher
how should journal references be written?
Author A.A. and Author B.B. (date of publication) title of article. Title of journal, volume number, page range
how should a website reference be written?
retrieved from ‘full website address’
what should be attached in the appendix?
all the research and data collected e.g. consent form, debrief form, questionnaire, diagrams, statistical calculations, raw data
What is concurrent validity?
the extent to which the results of a particular test or measurement correspond to those of a previously established measurement for the same construct
what is probability?
the likelihood that a particular event or outcome will occur
what is a type I error?
when we accept the experimental hypothesis even though the results were actually due to chance (false positive)
what is a type II error?
when we reject the experimental hypothesis when the results were actually significant (false negative)