Please enable JavaScript.
Coggle requires JavaScript to display documents.
Research methods - Coggle Diagram
Research methods
Probability
- hypothesis must be falsifiable
- if we look for a difference in the DV fro two conditions of the IV in an experiment the null hypothesis is “there is no difference” and the alternative is that there is
- samples may have small differences due to random variation or chance
- inferential statistical tests permit how to work out how probable it is that a pattern in data could have been by chance
- alternatively, the effect may represent a real difference/correlation in the population from which the samples were drawn
P levels
- generally use a p level of 5% as the cut-off
- means there's a less than 5% chance of results occurring if the null is true, given as p<0.05
- so there is at least a 95% chance that the effect observed is a real one in the population
- In some studies like drug testing researchers that want to be more certain use a more critical probability of p<0.01 means a less than 1% chance of the results occurring when there is no real difference/correlation between the populations
Errors
-
- when you reject the null when it should’ve been accepted
-
- when you accept the null when you should’ve rejected it
Features of a science
5 key features
- Empirical methods
- Empirical evidence is gained through direct observation or experiment
- means claims can be tested and used to make predictions
- Objectivity
- objective data collected systematically in controlled conditions
- so it’s not affected by the expectations of the researcher
- Replicability
- procedures are carefully recorded so other scientists can replicate them with different groups to test their validity
- theory construction
- explanations of observations and findings
- theories emerge from observations or from hypothesis testing
-
Falsifiability
Karl Popper 1934 🦢
- argued that it is only possible to disconfirm a theory
- pointed out that however many confirmed sightings of white swans there are you can’t conclude all swans are white as only one black swan will disprove it
- therefore research tests the null hypothesis e.g not all swans are white
- if you can reject the null hypothesis (with reasonable certainty) we may accept the alternative hypothesis
- a theory must be falsifiable
- some Freudian psychoanalysts can be criticised as lacking falsifiability
Paradigms
- Thomas Kuhn 1962
- proposed that scientific knowledge develops through revolutions not the process of gradual change suggests by Popper’s swan theory of falsification
- Said there are 2 main phases in science
- Normal science- which existing theory remains dominant while disconfirming evidence gradually accumulates
- Paradigm shift- a revolutionary overthrowing of existing theory and its replacement with a new set of assumptions and methods
Ao3
- The empirical approach teaches us to question claims
- must always look for evidence
- evidence should be directly observed or collected using controlled objective methods
- means we can reject pseudoscience beliefs based on weak or subjective evidence
- Kuhn’s theory of paradigm shift
- Kuhn described scientific progress as more like a religious conversion than a systematic logical process of hypothesis testing
- therefore according to him science is socially constructed through dialogue with other people not just a logical process of evolving theory based on empirical evidence
- the theory of paradigm shift is itself a paradigm shift from the previous ways of understanding how the scientific process works
Types of data
Nominal
-
-
-
e.g men vs women, smokers vs non-smokers
Ordinal
-In rank/order
e.g 1st 2nd 3rd, low medium-high, scale of 1-5
-
Ratio
-
-
e.g distance, income, weight, time
-
Case studies
Case studies
- detailed study of a single individual, institution or event
- variety of research techniques can be used to gather data such as observations interveiws and cognitive or personality tests
- they provide a rich detailed description of a person’s life
- can be longitudinal, following individuals over time
- findings are presented in a qualitative way being organised into themes
- can also include quantitative data like scores from tests
Individuals
- HM- hippocampus removed due to epileptic seizures resulting in an inability to form new memories
- Little Hans- Freud used him to illustrate the principles of psychoanalysis
- Phineas Gage- 1848, survived an iron rod passing through his brain but suffered personality changes
events
- London riots (2011) were studied by Reicher and Stott to re-examine explanations of ‘mob’ behaviour from a social psychological perspective
- mass suicide of a cult- reverend Jim Jones was responsible for the deaths of 900 followers in the 1970s. illustrated process of conformity and obedience
Ao3
- Case studies can be used to study very rare experiences
- some situations could not be generated experimentally for practical or ethical reasons but can be studied qualitatively using case studies
- David Reimer's case study, a boy whose pp was accidentaly removed and so raised as a girl is a fascinating example
- it is possible to provide insight into the complex interactions of many factors in people’s experience after horrific events like accidents or extreme deprivation
- However, case studies study unique individuals
- the subjects of case studies often have particular or unusual characteristics
- e.g HM suffered epilepsy for many years as well as brain damage from his hippo removal
- we don’t know how different factors interacted to affect the indiv behaviour so it is difficult to generalise from indiv cases
- confidentiality and informed consent must be carefully considered
- as many cases are unique they may be easily identifiable so they’re given a false name or use initals.
- however some may not have been able to give informed consent like HM who obviously was too messed up to give consent
- so researchers must not reveal too many personal details that enable the person to be located
- We often lack data from before a particular event
- the interest in an individual often begins aster an event such as the brain damage to Phineas Gage or HM
- E.g we don’t know if HM’s epilepsy and previous drug treatments may have affected his brain prior to surgery
- we can’t compare before and after results which makes it difficult to draw conclusions and cause and effect
Reliability
Reliability- consistency of measurements
procedures should be standardised in any research study to improve reliability and replicability
-
-
-
Validity
Validity- whether an observed effect is genuine
Test may be reliable but lacks validity if it doesn’t measure the concept the researcher is aiming to measure
-
Ecological validity
- method used to measure DV can be artificial giving poor ecological validity- lab, field or natural experiment
- consider the mundane realism- whether a study reflects real-world experiences rather than just the location
- e.g Godden and Baddley 1975
- scuba diver study of forgetting in context
- learnt word on land or in water and had to recall them on land/in water
- situation was real-life but the task was very artificial and the divers were aware o being studied so may not have behaved naturally
- Demand characteristics can effect ecological validity if Ps are aware of being studied so alter behaviour to “look good” (social desirability bias) or to fit what they think the researcher expects. means they’re not behaving accurately to how they would IRL
-
Content analysis
Content analysis
- content analyss- indirect form of observational study, analysing materials produced by people e.g bookd, films, photos
- deciding on a sample e.g TV ads over a week
- soding tha data using behavioral categories e.g men or women using household products
- recording the occurences of each coding category
- in a quantative analysis the instances of each coding category are tallied and can be representened using descriptive statistics and graphs
- in a qualatative analysis examples in each category are described
Thematic analysis
- thematic analysis- a type of qualitative content analysis, summarises the data descriptively. aims to identify underlying theme in the data rather than spotting obvious words or phrase
- aims to allow themes to emerge from data and maintani Ps perspectives
- read and rereade the data transcripts
- break the data into meaningful units
- assign a label or code to each unit
- combine codes into larger themes
- ensure that these emerging themes represent all of the data
Ao3
- A stregnth of content analysis is its high ecological validity
- baed on oberservation produced by peoplein thier real lives
- inlcudes newpapers, books, paintings, photos, flms, videos
- these cimmunications can be current andrelevent to a specific research question
- Content analysis is replicable
- materials are often publicly available and can be access by another researcher
- means that observations can be tested for inter-rater reliability
- if sevreal researchers identify similar thees or occurance of coding categories the findings are reliable
- Content anaysis oversimplifies the data
- summarises rich qualitative data into a simplified form
- means tha data loses detailed desprictive flavour
- observer cannot truly be objective and may impose meanign on people’s behaviors because of preconceptions
- so coding categories may lack validity as they don’t represent peolpes understanding of their own behavior or creative output
- Thematic analysis is very time-consuming and painstaking to carry out
- because it involves examining and re-examining huge amounts of data with themes emerging interatively
- enables datat to be summarised and conclusions to be drawn
- but is not always suitabel when researchers have limited time available as it takes so bloody long
-
-
-
-
-
-
-
Mean, mode, median, range, etc
-
-