Research Methods Flashcards
Define operationalism
The process of making variables quantifiable
What is the difference between an aim and a hypothesis?
An aim is what you hope to achieve from an experiment whereas a hypothesis is a prediction of the outcome of your experiment
Define internal validity
The extent to which evidence supports the experiment within the context of the experiment
Define external validity
The extent to which evidence supports the experiment outside of the context of the experiment
What is the experimental method?
Observation Hypothesis Designing a study Collecting data Analysing data Questioning validity Drawing a conclusion
What are the 3 types of internal validity?
Control
-> did the IV affect the DV or was it something else?
Construct
-> how logical is the hypothesis - does an existing theory disprove it?
Mundane Realism
-> how realistic is it?
What are the 3 types of external validity?
Ecological Validity
-> can it be generalised to different settings / places?
Population Validity
-> can it be generalised to different ppl / populations?
Historical Validity
-> can it be generalised to different times in history?
What are the 4 types of experiment?
Lab: -> IV + DV both controlled Field: -> IV controlled, DV not Quasi: -> DV controlled, IV not Natural: -> neither controlled
Define investigator bias
any cues from the investigator (not the IV) that encourages certain behaviours from the participant
Define demand characteristics
the participants are aware of the aims of the study and change their behaviour as a result
What are the 4 principles of the BPS code of ethics?
Respect
Competence
Responsibility
Integrity
Define social desirability bias
When participants alter their behaviour to seem more socially desirable -> biasing the results
Define extraneous variables
Any characteristic of a participant that may affect the DV
Define confounding variables
A feature of the research situation that may affect a participant’s behaviour
Define the Hawthorne effect
The tendency for participants to alter their behaviour because they know that they are being observed
Define sampling
The process of selecting a group to study
Define target population
The group of people that a researcher is interested in
What are the 5 sampling methods
Volunteer Opportunity Random Systematic Stratified
What is volunteer sampling?
When researchers their experiment and accept volunteers as participants
What is opportunity sampling?
When researchers ask the most conveniently found people to participate
What is random sampling?
When the researcher obtains a list of people and randomly selects a group of participants using computer algorithms
What is systematic sampling?
Where a researcher gains a list of a population, assigns each person a number and selects every nth person
What is stratified sampling?
When levels of the target population are identified (stratas) and then participants are selected randomly
Which are the most valid types of sampling?
Invalid to Valid: Volunteer Opportunity Random Systematic Stratified
What are the strengths of volunteer sampling?
- > gives access to a variety of participants
- > requires little effort
- > less chance of the ‘screw you’ phenomenon
What are the weaknesses of volunteer sampling?
- > sample is biased because participants are likely to be one particular aspect of the population
- > volunteers are eager to please which increases chances of demand characteristics
What are the strengths of opportunity sampling?
Easiest -> just have to take the first suitable participants you can find
What are the weaknesses of opportunity sampling?
Inevitably biased because sample is drawn from a small part of the population and therefore cannot be generalised
What are the strengths of random sampling?
- > Unbiased as all members of the target population have an equal chance of being selected
- > sample should be fairly representative
What are the weaknesses of random sampling?
- > need to have a list of all members of the population and then contact all of those selected, takes time
- > unbiased selection doesn’t always guarantee an unbiased sample
What are the strengths of systematic sampling?
- > unbiased as all members of the target population have an equal chance of being selected
- > sample should be fairly representative
What are the weaknesses of systematic sampling?
- > need to have a list of all members of the population and then contact all of those selected, takes time
- > the process of selection can interact with a hidden periodic trait within the population (eg every 5th person lives in a flat)
What are the strengths of stratified sampling?
- > unbiased as participants are selected using an objective system
- > likely to be more representative than other methods because there is a proportional and randomly selected representation of subgroups
What are the weaknesses of stratified sampling?
- > requires detailed knowledge of the population
- > very time consuming
When are directional hypotheses used?
When we have enough information from past research
They predict what will happen due to the IV
When are non-directional hypotheses used?
When we cannot be sure what will happen
They predict something will happen due to the IV
Define correlation
Correlations systematically show how strong a relationship is between 2 continuous co-variables
When are correlations used?
- > it would be unethical to do an experiment
- > it is impractical to manipulate the variables
- > no other methods are available
Define the 3rd variable problem
When there appears to be a casual relationship but there is actually a 3rd variable affecting the result
Define the correlation coefficient
The strength and direction of a correlation
-1 = perfectly negative 0 = no correlation 1 = perfectly positive
What are the types of observation?
Naturalistic or controlled
Covert or overt
Participant or non-participant
Define a naturalistic observation
Nothing about the setting is changed
Usually watching infants or animal interactions
More ecologically valid than controlled
Define controlled observation
We change / control the situation
eg Bandura’s bobo dolls
Less extraneous variables than naturalistic
Define covert observation
Participants are unaware that they are being watched
Ethical issues - deception, informed consent
More natural, realistic behaviour can be observed
Define overt observation
Participants know that they are being watched
We try to be subtle but they know we’re there
We can ask for informed consent
May cause demand characteristics
Define participant observation
Observer is part of the group that they’re watching
Investigator effects and confounding variables more likely
Observer gains insights
Subjects may become aware of study
Define a non-participant observation
The observer watches from a distance
Less investigator effects + confounding variables
Less valid as less data than experiencing it themselves
Define self reports
Any technique where the data that’s collected comes directly from the participant
eg questionnaires or interviews
Define a structured interview
All questions are pre-determined
No deviation from the list of questions
Results are recorded exactly
What are the advantages of structured interviews?
More fair, unbiased and reliable
Easily repeatable as questions never change
What are the disadvantages of structured interviews?
Less valid
Less detail and less exploration of the participant’s thoughts
Lower reliability if interviewer treats participants differently
Social desirability bias
Define an unstructured interview
General targets for discussion -> not predetermined questions
Interviewer free to make decisions
Similar to clinical interviews
What are the advantages of unstructured interviews?
Details can be explored -> allows access to what people really think
More valid
What are the disadvantages of unstructured interviews?
High risk of investigator bias -> interviewer’s expectations could affect answers given
All interviewers must be skilled and trained
High risk of social desirability bias
What are the advantages and disadvantages of questionnaires?
Advantages:
- > respondents may feel more willing to reveal personal info
- > once designed + tested, can be distributed quickly + cheaply
- > impersonal nature reduces social desirability bias
Disadvantages:
-> data collection limited by sampling bias -> only people willing and with spare time will complete
What are the advantages and disadvantages of open questions?
Advantages:
- > respondents can expand on answers
- > provide unexpected answers, gaining new insights into thoughts + behaviours
Disadvantages:
- > can be hard to understand answers
- > difficult to summarise
- > social desirability bias
What are the advantages and disadvantages of closed questions?
Advantages:
-> quantitative data is hard to analyse
Disadvantages:
- > limited range of answers
- > demand characteristics
- > may result in false / best fit answers
- > acquiescence bias
Define primary data
Information that is observed first hand
Define secondary data
Information that doesn’t come directly from the source
What are the 3 measures of central tendency?
Mean
Median
Mode
What are the 2 measures of dispersion?
Range
Standard deviation
Define inferential statistics
Interpreting the meaning in relation to your hypothesis
What techniques are used in inferential statistics?
Correlation
Graphical display
Parametric and non-parametric difference tests
Define standard deviation
The average distance from the centre
What are the 4 levels of data / measurement?
Nominal
Ordinal
Interval
Ratio
Define nominal data
Data that can be put into discrete categories
Define ordinal data
Data that can be put into orders
Define interval data
Data in categories that can be divided into subsections
Define ratio data
Data in no categories and decimal format
What data can be used in parametric tests?
- > interval or ratio data
- > data that can normally distributed (no outliers)
- > similarity of variance is essential
What data can be used in non-parametric tests?
- > nominal or ordinal data
- > data must be transformed
What does a statistical test calculate?
The probability that chance has caused a difference
What confidence level is most commonly looked for?
p < 0.05 or 95%
Define the sign test
Used with repeated measures or matched pairs design
When we are looking for the difference in scores
Non-parametric test -> nominal data
Define an alternate hypothesis
A significant difference has been found
Define a null hypothesis
A significant difference has not been found
Define a type 1 error
A false positive result -> accepting the alternate hypothesis instead of the null hypothesis
Define a type 2 error
A false negative result -> accepting the null hypothesis instead of the alternate hypothesis
What are the 4 stages of the sign test?
- is the hypothesis directional or non-directional?
- signs -> + or -
- S/N variables
- Match to confidence level
How can reliability be tested?
By repeating the experiment and seeing if the results are the same as before.