research methods (P2) Flashcards
Aims, Hypothesis, and Variables
Aim - the purpose of research
Directional hypothesis - expected effect is stated (previous research suggests an outcome)
Non directional hypothesis - there will be a difference (no previous research)
Null hypothesis - there will be no difference
IV - manipulated
DV - measured
EVs - variables other than the IV that may have an affect on the DV
Standardisation - keeping EVs the same across all conditions
Operationalisation - clearly defining variables so they can be measured
Control condition - used as a baseline
Randomisation - using chance to control the affects of bias
1 - names in a hat
2 - draw out one and allocate to A then B
3 - continue until all are assigned
Experimental designs
Repeated measures - PPTs take part in both conditions
+ less time consuming/ costly as less PPTs required
+ reduces PPT variables (individual differences) as same PPTs taking part in all conditions
– order effects as PPTs may become tired
– demand characteristics as they are more likely to guess the aim
Independent measures - PPTs take part in different conditions
+ avoids order affects as PPTs only take part in one condition (not tired)
+ reduces demand characteristics as they are less likely to guess the aim
– more time consuming/ costly as more PPTs required
— PPTs variables as different characteristics (age) act as EVs
Matched pairs - pairs of PPTs are matched based on variables such as IQ (involves pretesting)
+ reduced PPT variables as PPTs are matched based on characteristics
+ reduced order effects as they only take part in one condition
– time consuming/ costly as more PPTs required
– PPT variables may still occur as it is difficult to find exact matches
Data types
Qualitative - data expressed in words
Quantitative - data that can be counted
Evaluation:
- rich in detail
- easy to analyse and draw conclusions
- less open to bias/ opinion
Primary - data collected first hand by researcher for the purpose of this research
+ perfectly fits the study
– more time and effort
Secondary - data collected from pre-existing research
+ less time consuming and costly
– quality cannot be controlled and may not perfectly match the study
Ethics
Ethical issue:
- Informed consent (PPTs agree to take part based on info) - Debriefing (at the end of the study PPTs are made aware of the aims)
- Psychological harm (stress, embarrassment) - Right to withdraw
- Deception (misleading PPTs - Counselling
- Privacy and Confidentiality (right to control info about themselves and have data protected) - Anonymity (numbers)
Consent form:
- Details of experiment
- ethical issues (right to withdraw, confidentiality)
- signature
Debrief form:
- details of the study PPTs could not know beforehand (what the researcher expected to find)
- ethical issues (right to withdraw, confidentiality)
- offer counselling
- opportunity to ask questions
- signature
Features of science
Objectivity - no bias? personal opinion
Replication - repeating the same procedure to get the same findings
Empirical method - collecting data based on direct experience
Falsifiability - proving something is wrong
Theory construction - a theory is a set of general laws that have the ability to explain particular behaviour and is tested using hypotheses
Paradigms - a set of beliefs about how behaviour is explained and a paradigm shrift occurs when the scientific community change from one established way to a new one (smoking)
Validity and Reliability
Validity - accuracy of data
Internal - are results solely affected by IV
External - can data be generalised
Face - does it appear to measure what it says it measures - conducted by specialist in that area
Concurrent - performance of test is compared with pre-established test (0.8+)
Ecological - the extent to which findings can be applied to other settings (everyday life)
Temporal - the extent to which findings can be applied across time
Demand characteristics - PPTs change behaviour due to guessing aim
Social desirability bias - underreporting undesirable behaviours and overreport desirable ones to be seen in a more positive light
Investigator effects - unwanted influences the investigator communicates to PPTs that affects their behaviour
Confirmation bias - tendency to seek out information that supports our pre-existing beliefs
Reliability - consistency
How to test:
- Test-retest - giving PPTs the same tasks on more than one occasion (results should be the same) - scores are correlated (0.80+)
- Inter-rater - agreement between 2 observers (correlate observations, 0.80+)
Peer review
Peer review - research is sent to independent scrutiny by other psychologists to see if it should be published
Fabrication - exaggerating
Plagiarism - stealing someone else’s work
Purpose:
- allocate funding - usefulness to wider society
- validate research - check accuracy/ originality
- suggest improvements - check its high quality to protect psychology’s scientific status
Experiments
Lab - researcher directly manipulates the IV in a controlled environment
+ control over EVs - cause and effect
+ standardised - replicated
– lack ecological validity - generalised
– lack mundame realism - generalised
Field - researcher directly manipulates the IV in a realistic environment
+ high ecological validity
+ standardised - replicated
– ethical issues - unaware of observation
– less control of EVs - cause and effect
Natural - Iv is not directly manipulated as it is naturally occurring
+ unique events that would otherwise not be able to be research - ethical issues
+ high ecoligcal validity
– hard to replicate
– less control over EVs
Quasi - Iv based on pre-existing differences (age)
– PPTs are not randomly allocated and so we cant limit the effects of bias
Self report techniques
Interviews:
- Structured - pre-prepared questions
- unstructured - conversation-like
- semi-structured - mostly prepared questions
Open questions - allow PPTs to answer however they wish
+ rich in detail - expanding
– social desirability bias
Closed - restricted to a set of answers
+ easy to analyse
– unable to explore responses
Observations
Covert - observe people without their knowledge
+ reduces demand characteristics
– unethical - privacy
Overt - know you are being observed
+ ethically acceptable - informed consent
– demand characteristics
PPT - observer takes part in activity
+ gain greater insight, increasing validity
– researcher may become too involved - objective
Non-PPT - observer does not take part
+ more objective - less bias
– miss important details - reduces validity
Naturalistic - setting is where the behaviour naturally occurs
+ high ecological validity
– lack of control over EVs - cause and effect
Controlled - where EVs are controlled
+ easily replicated - standardised
– reduced ecological validity - artificial
Sampling
Time sampling - observations made at regular intervals
– infrequent behaviours are missed
Event sampling - target behaviour is recorded every time it occurs
– complex behaviours are not recorded accurately
Case studies
- involves interviews and observations
+ give unique insight into human functioning
+ high ecological validity - no manipulation
– cannot be replicated - less reliable
– small sample - cant be generalised
Sampling
Random - each member has an equal chance of being selected - random number generator
+ no researcher bias - representative - generalisable
– time consuming to ensure everone has an equal chance of being selected
Systematic - every nth person
+ no researcher bias - predetermined system
– not truly unbiased as every nth person could have a particular trait
Stratified - the makeup of the sample reflects the makeup of the target population
+ no researcher bias
– timeconsuming to identify subgroups
Opportunity - who is available and fits criteria
+ quick, easy and cost affective
– bias - specific location
Volunteer - PPTs put themselves forward from an advert
+ minimal effort - quick and easy
– bias - particular type of person to approach
Content analysis
Indirectly observing the presence of words, images or concepts within the media (films, books)
Coding takes place (data is categorised into meaningful units) and the codes are then counted
Measures of central tendency
Mean
+ more representative (includes all data)
– anomalies can skew the data
Median
+ not affected by extreme values
– less representative (doesn’t account for all values)
Mode
+ appropriate with nominal data
– not representative (of the data as a whole)
Measures of spread
Range
+ easy to calculate
Standard deviation
+ more accurate and not effected by extreme values
