Research methods Flashcards
State the difference between directional and non-directional hypothesis and when it’s used
D = States the direction of the relationship that will be shown between the variables. Used when previous research has been carried out.
ND = Doesn’t. Used when no previous research has been carried out.
Define independent variable
What is manipulated or changed naturally.
Define dependent variable
What is measured and caused by the IV.
Define operationalise
To be precise and clear about what’s being manipulated and measured. This makes the research testable and repeatable. The hypothesis should also show the operationalisation.
Define extraneous variables and state the types
Any other variables that affects the DV. They must be controlled for to prevent them from skewing the results and leading to false conclusions.
- Demand characteristics
- Investigator effects
Define cofounding variables
When extraneous variables aren’t properly controlled for.
Define aim
The purpose of the study.
Define hypothesis
A statement of the relationship between variables been investigated.
Define pilot studies and state its aim
A small-scale version of an investigation which is done before the real one.
They’re carried out to allow potential problems of the study to be identified and the procedure to be modified to deal with these. Also may give an early indication of whether the results will be statistically significant.
Define peer review
The assessment of scientific work by experts in the same field, to ensure it’s credible (high quality) before being published.
Define independent group designs state its S+L and solution
= The participants only perform in one condition of the IV.
STRENGTHS:
- No order effects
- Demand characteristics eliminated.
LIMITATIONS:
- No control over participant variables
Solution: Random allocation
Define repeated measures
state its S+L and solution
= The same participants take part in all conditions of the IV.
STRENGTHS:
- Eliminates participant variables.
- Fewer participants needed so not as time-consuming.
LIMITATIONS:
- Order effects
Solution: Counterbalancing means when half of the participants do conditions in one order and the other half do it in an opposite order.
Define matched pairs
state its S+L
= Pairs of participants are matched on characteristics then experiment is conducted as an independent groups design.
STRENGTHS:
- No order effects
- Demand characteristics less of a problem
LIMITATIONS:
- Time consuming & expensive
- Large group of participants needed.
Define demand characteristics
Any cue the researcher or situation may give which makes the participant feel like they can guess the aim of the investigation.
This can cause them to act differently by:
- In a way they think the researcher wants them to.
- Intentionally underperforming to sabotage the study’s results.
State the difference between experimental/alternative and null hypothesis
experimental/alternative (H1) - a prediction that changing the IV will cause a change in the DV
null (H0) - a prediction that it wont
State the difference between type 1 and type 2 errors
Type 1 (false positive) = when researchers conclude an effect is real (i.e. they reject the null hypothesis), but it’s actually not.
Type 2 - (false negative): When researchers conclude there is no effect (i.e. they accept/fail to reject the null hypothesis), but there actually is a real effect.
State the aim of inferential testing
To see whether a study’s results are statistically significant, i.e. whether any observed effects are as a result of whatever is being studied rather than just random chance.
State when results are statistically significant
If the observed value is less than or equal to the critical/table value. Then reject the null hypothesis and accept alt.
State the criteria for the type of inferential test
- Difference or correlation
- Nominal, ordinal or interval data
- Experimental design is related (repeated measures) or unrelated (independent)
State the difference between primary and secondary data and S+L
primary = original data collected for the study
S - targets exact info needed so aims are fulfilled
L- requires time & effort
secondary = data from another study previously conducted
S - requires minimal effort to collect as data is already accessed
L - may be outdated or incomplete
-
state the difference between qualitative and quantitative data and S+L
qualitative = non-numerical
S - more meaningful insight
L - researcher bias presented as conclusions depend on their interpretation
quantitative = numerical
S - data can be mathematically and objectively analysed
L - less detail
Define laboratory experiment
state it’s S + L
= An experiment conducted in an artificial, controlled environment.
S
- HIGH DEGREE OF CONTROL: IV has been precisely replicated which means any change in outcome must be a result of a change in the variable, leading to greater accuracy.
- EASILY REPLICABLE for experimenters to check results
L
- LOW ECOLOGICAL VALIDITY: Results obtained in an artificial environment might not translate to real-life.
- EXPERIMENTAL BIAS: means that participants can be influenced by demand characteristics by changing their behaviour to their expectations.
Define field experiment and state its S + L
= An experiment conducted in in a natural, real-world environment.
S
- NATURALISTIC: therefore high ecological validity
L
- ETHICAL CONSIDERATIONS: invasion of privacy and so likely to have no informed consent
- LOSS OF CONTROL: over extraneous variables hence precise replication not possible
Define quasi experiment and
state its S + L
= An experiment that compares between two variables that cannot be change i.e. naturally exist. (e.g. gender)
S
- CONTROLLED CONDITIONS: hence replicable, therefore likely to have high internal validity
L
- NO RANDOM ALLOCATION: to conditions so there may be cofounding variables presented. This makes it harder to conclude that the IV caused the DV.
Define natural and state its S +L
= An experiment where the variable changes naturally e.g. reactions to earthquakes and the researcher seizes the opportunity to study the effects.
S
- HIGH EXTERNAL VALDITY: as ur dealing with real life issues
- PROVIDED OPPORTUNITIES: for research that would’ve otherwise been impossible due to practical/ethical reasons
L
- NATURAL OCCURING EVENTS: may be rare therefor unlikely to be replicable hence hard to generalise findings.
- DIFFICULT TO RANDOMISE: confounding & extraneous variables become a problem
Define investigator effects
when a researcher unintentionally influences the results of a study
Define meta-analysis and state its S+L
When a researcher combines results from many different studies and uses all the data to form an overall view of the subject they’re investigating.
S
- GENERALISABILITY is possible as a larger amount of data is studied. TMT the researcher is able to view the evidence with more confidence.
L
- PUBLICATION BIAS: such as the file drawer problem may be presented. This is when researchers intentionally leaves out the negative results. This gives a false representation of what the researcher was investigating.
Differentiate between nominal ordinal and interval data
NOMINAL = Discrete date in the form of categories.
ORDINAL = Whole numbers that can be ordered but not necessarily precise measurements.
INTERVAL = Standardised units of measurement
Define randomisation
The use of chance to reduce the effects of bias from investigator effects.
Define standardisation
Using the same exact same formalised procedures for every single participants.
explain informed consent and its solution
explanation:
participants must be told the research’s aims, the data being collected and any risks associated with participation
solution:
Prior consent = informing participants that they will be deceived without telling them the nature of the deception.
Retrospective consent = Informing participants that they were deceived after the study is completed and asking for their consent. The problem with this is that if they don’t consent then it’s too late.
Presumptive consent = Asking people who aren’t participating in the study if they would be willing to participate in the study. If these people would be willing to give consent, then it may be reasonable to assume that those taking part in the study would also give consent.
explain protection from harm and its solution
explanation: Participants must be protected from physical and psychological harm.
solution:
counselling
cost-benefit analysis - before the study, the ethics committee weigh up the procs and cons of the study to determine whether it will be ethical.
explain deception and its solution
explanation: the act of deliberately misleading participants about the nature of the study or withholding information from them.
(only acceptable to avoid demand characteristics or if wont cause distress)
solution:
debriefing (written or verbal)
- the true nature of the study must be said and participants should be told what their data will be used for
- after the debrief participants have the right to withhold or withdraw their data
explain privacy and confidentiality and its solution
explanation: refers to the right that participants have to controlling personal data about themselves - what is disclosed and how its used.
solution:
- Anonymity can be maintained by the researched not recording and personal details so that the data results can’t be traced back to them.
- This can be done by referring them as numbers or initials.
Define objectivity
Observations are made from a neutral perspective without bias, rather than the scientist’s subjective viewpoint.
e.g. a set of scales is a more objective way of determining which of two objects is heavier than a researcher lifting each up and giving their opinion.
Define replicability
The extent to which scientific methods and their results can be repeated by other researchers across other contexts and circumstances
- It’s used to asses the validity and reliability of results from the study.
Define falsifiability
When a theory or hypothesis is tested/replicated, there must be some possible observation that could prove it false – even if that never actually happens.
define the empirical method
when evidence is collected through making direct observations and through direct experiences.
define deduction
refers to the process of deriving new hypotheses from an already existing theory
define paradigm and paradigm and shift
paradigm = a set of shared ideas and assumptions within a scientific discipline
paradigm shift = a significant change in these assumptions resulting from a scientific revolution.
Differentiate between naturalistic and controlled observation and its S + L
Naturalistic = Observations made in a real-life setting
S
HIGH ECOLOGICAL VALIDITY as the researcher records naturally occurring behaviour in a natural environment, without any interference.
L
CAN’T REPLICATE as the researcher isn’t in control of variables
Controlled = Observations made in an artificial setting set up for the purposes of observation
S
LOW ECOLOGICAL VALIDITY
L
CAN BE REPLICATED
Differentiate between covert and overt observations and its S + L
Covert = Participants are not aware they are being observed
S
INVESTIGATOR EFFECTS unlikely meaning that participant’s behaviour is genuine
L
ETHICAL ISSUES as participants can’t give informed consent if they’re unaware
Overt = Opposite
S
INFORMED CONSENT obtained as participants are informed in advance
L
INVESTIGATOR EFFECTS likely which distorts participants behaviour due to social desirability bias
Differentiate between participant and non-participant observations and its S + L
Participant = Where the researcher/observer is actively involved in the situation being observed
S
IN DEPTH DATA obtained as they’re in close proximity to the participants and so unlikely to overlook or miss any behaviours
L
EVALUATION APPREHENESION means participants’ behaviour is distorted as they fear being judged by observers.
Non-participant = Opposite
S
INVESTIGATOR EFFECTS AND EVALUATION APPREHENSION LESS LIKELY
L
LACK OF IN DEPTH DATA as research isn’t in close proximity therefore might overlook/miss behaviours
differentiate between time and event sampling
time sampling = Recording participant behaviour at regular time intervals.
E.G. making notes of the participant’s behaviour after every 1 minute has passed.
event sampling = counting how many times the participant behaves in a certain way.
E.G. tallying it over a full hour to gather an overall impression of the amount of co-operative behaviour
define behavioural categories
an organisation of observational data by operationalising the behaviours being recorded
Define reliability
= A measure of whether the same results can be consistently replicated under the same circumstances.
define internal reliability and how its measured
= describes the internal consistency of a measure.
E.g. Whether different questions in a questionnaire are all measuring the same thing.
measure: SPLIT HALF METHOD
- where data is split randomly in half and compared, to see if results taken from each part of the measure are similar.
describe external reliability and how its measured
= assess the consistency when different measures of the same thing are compared
measure 1: TEST RE-TEST
- When you give the same test to the same person on two different occasions. If the results are the same or similar both times, this suggests they are reliable.
measure 2: INTER-OBSERVER
- Whether different observers obtain the same data from the same study
- Can also be assessed mathematically by looking for correlation between observers’ score.
- formula: total no. of agreements / to. no. of observations >= 0.80 = high
explain how reliability is improved in questionnaires
Replace some of the open questions where they may be more room for misinterpretation, with closed, fixed choice alternatives which may be less ambiguous.
explain how reliability is improved in interviews
use the same interviewer each time or properly train them to not ask leading or ambiguous questions
explain how reliability is improved in observations
behavioural categories are:
- operationalised
- measurable
- self-evident
- not overlapping
define validity
refers to whether a study results accurately measures what they’re supposed
define internal validity and state the type
whether the results of an experiment are due to the manipulation of the IV and not any other factor
- concurrent = the extent to whether the results obtained are similar to the results of the well established test
define face validity
The study just looks like it measures what it’s supposed to at face value.
e.g. People with with high exam scores score higher on IQ test, however you can’t be sure that these are directly linked.
Define external validity and state the types
= a measure of whether data can be generalised to factors outside the investigation
- ecological
- temporal
- population
define ecological validity
generalisability to other situations and setting.
define temporal validity
generalisability to other historical times and eras
define population validity
generalisability to different populations of various ages.
define participant variables
characteristics or aspects of a participant’s background that could affect the study’s results
define researcher variables
factors such as researchers behaviour, gender or appearance that could affect the study’s results
define situational variables
factors in the environment that could affect the study’s results
state the types of investigator effects
- Non-verbal communication
e.g. a raised eyebrow can make the participant aware they might have said/done something that surprised the researcher - Physical characteristics such as gender
e.g. a male interviewer in all-female group of participants - Bias in interpretation of data (if the method of measurement is subjective)
state the ways of improving validity in experimental research
- using a CONTROL GROUP to better assess whether changes in the dv were due to iv
to reduce demand characteristics and investigator effects:
- STANDARDISATION
- SINGLE-BLIND PROCEDURE = participants aren’t made aware of the aims of the study
- DOUBLE-BLIND PROCUDERE = a third party conducts the experiment without knowing its aims
state the ways of improving validity in questionnaires
- incorporate a LIE SCALE
within the questions in order to assess the consistency of the participants response and to control for the effects of social desirability bias - assuring respondents that all data submitted will remain ANONYMOUS
state the ways of improving validity in observations
high ecological validity if there’s minimal intervention by the researcher, especially in COVERT therefore behaviour of participant is likely to be natural and authentic
state the ways of improving validity in qualitative methods
- high ECOLOGICAL VALIDITY as depth and detail associated with case studies and interviews, for instance is better able to reflect participants reality
- researcher has to demonstrate the INTERPRATIVE VALIDITY of their conclusions = the extent to which the researcher’s interpretation of events matches those of their participants
e.g. including direct quotes from participants within the report
- TRAINGULATION - the use of a no. of different sources as evidence
e.g. data compiled through interviews with friends and family, personal diaries ect.
state the S + L of mean as measure of central tendency
S
- uses all the data in the set
- good for interval data
L
- influenced by outliers so it can be unrepresentative
state the S + L of median as measure of central tendency
S
- not effected by outliers
- good for ordinal data
L
- not as sensitive as mean as it doesn’t use all data
state the S + L of using mode as a measure of central tendency
S
- useful for nominal data
L
- Not useful when there’s several modes
define measures of dispersion and state the S + L of range and SD
= quantifying how much scores in a data set vary
Range
S - Easy to calculate
L
- Affected by outliers which may give an exaggerated impression of the data distribution
- Doesn’t account for how common scored are for e.g. if data sets have the same range but different distributions
SD = a measure of how much numbers in a data set
S - less affected by outliers
L - takes longer/difficult to calculate
state the characteristics of a normal distribution
- majority of scores cluster or/near the mean average
- symmetrical
- scores become rarer as they deviate from the mean
differentiate between a positively skewed and negatively skewed distribution
+ most of the mean is distributed on the right
- most of the mean is distributed on the left
define curvilinear relationship
as one variable increases so does the other but only up to a certain point after which as one variable continues to increase the other begins to decrease ( u shape)
state the type of data used for bar charts, histograms and line graphs
bar - discrete
histogram - continuous
line - continuous
state the S + L of correlations
S
- used as STARTING POINTS to asses patterns between co-variables before committing to conducting an experiment
- SECONDARY DATA used which which makes it less time consuming
L
- difficult to establish a CAUSE AND EFFECT relationship as only an association is found
- THIRD VARIABLE PROBLEM presented in which the researcher is unaware it’s responsible for the relationship between co-variables
Define content analysis and how to conduct one
= a method used to analyse qualitative data and transform into quantitative data
- uses coding units (varies widely depending on data)
e.g. the number of swear words in a film
- Data is collected
- Researcher reads through and examines the data, familiarising with it
- Researcher identifies coding units
- The data is analysed by applying coding units
- A tally is made of the number of the times that a coding unit appears
state the S + L of content analysis
S
- reliable as the coding units aren’t open to interpretation and so are applied in the same way over time with different researchers
- easy to analyse and isn’t time-consuming
L
- Only describes the data it cannot extract any deeper meaning or explanation for the data patterns arising
state the process of thematic analysis
(alt to content analysis)
- data is reviewed repeatedly so that the researcher can identify trends
1.
State the types of peer review
- Open review: The researchers and the reviewers are known to each other.
- Single-blind: The researchers do not know the names of the reviewers preventing them for being able to influence
- Double-blind: The researchers do not know the names of the reviewers, and the reviewers do not know the names of the researchers, preventing researching bias.
State the strength of peer review
Helps to prevent scientific fraud as the work submitted is scrutinised
State the limitations of peer review
- prevents progress of new ideas: reviewers of papers are typically older and established academics who have made their careers within the current scientific paradigm. As such, they may reject new or controversial ideas simply because they go against the current paradigm rather than because they are unscientific.
- plagiarism: In single-blind and double-blind peer reviews, the reviewer may use their anonymity to reject or delay a paper’s publication and steal the good ideas for themself.
State the strengths of questionnaires
S
- REPLICABLE since uses standardised procedures (pre-set questions for all participants) therefore results can be confirmed by other researchers strengthening certainty in the findings.
- QUANTIFIABLE: closed questions provide quantifiable data in a consistent format, which enables to statistically analyse information in an objective way.
state the limitations of questionnaires
- DISHONESTY: social desirability bias
- LESS DETAILED: descriptions of private experiences
- BIASED SAMPLES: questionnaires handed out to people at random will select for participants who actually have the time and are willing to complete it
state the strengths of interviews
MORE DETAIL: particularly unstructured interviews conducted by a skilled interviewer - enable researchers to delve deeper into topics of interest, for example by asking follow-up questions.
REPLICABLE since uses standardised procedures (pre-set questions for all participants) therefore results can be confirmed by other researchers strengthening certainty in the findings.
state the limitations of interviews
LACK OF QUANTIFIABLE DATA: may produce difficulties in comparing data between participants (statistical analysis) due to nature of open discussion
INTERVIEWER EFFECTS: The interviewer’s appearance or character may bias the participant’s answers e.g. gender
State the features of a case study
- detailed investigations into an individual, a group of people, or an event.
- use observations, questionnaires, interviews
- qualitative data as its study of single subject (S)
- data is then used to build a case history of the subject
- longitudinal
- Allows for investigation into issues that may be impractical or unethical to study otherwise. (S)
state the limitations of case studies
- LACK OF SCIENFTIC RIGOUR: Because case studies are often single examples that cannot be replicated, the results may not be valid when applied to the general population.
- RESEARCHER BIAS