Research Methods Flashcards
Dependant variable
The thing that you measure
Independent variable
The conditions that you change
Aim
Describes the purpose of the investigation
Hypothesis
States the relationship between the two control variables
Operationalisation
Defining the variables in a specific way so they can be measured
E.g. Gender affects sports performance
—> gender difference in time taken to sprint 100m
Randomisation
Randomly allocate participants, reduces the risk of participant variables influencing the result
Standardisation
Keep everything the same for each participant
Counterbalancing
Half and half in each condition then switch over
Random sampling
Equal chance of being selected
Stratified sampling
Divided into sub-groups then randomly selected
Opportunity sampling
Simplest form
Take anyone who is available
Self- selected sampling
Participants sign up
Independent groups
Different groups of people
Only take part in one condition
Repeated measures
Same group of participants
Take part in each condition
Matched pairs
Different groups that have been matched
Only take part in one condition
Lab experiment
Controlled environment
Easy to replicate
Extraneous variables minimised
Not realistic
Low mundane realism
Field experiment
Natural setting, manipulate IV
More realistic than lab
Higher ecological validity
Hard to control extraneous variables
Ethical issues if people dont know theyre being observed
Natural experiment
IV changes naturally and isnt influenced by the experimenter
High ecological validity
Studies ‘real’ problems
Random allocation isnt possible
Informed consent
Participants know the aims and accept conditions
Consent forms
Deception
Lie to participants
With hold info
Full informed consent not given
Protection from harm
Participants not at more risk than they would be in real life
No physical/psychological harm
Privacy and confidentiality
Make info private/anonymous
Must be confidential, data protection act
Naturalistic/controlled observation
Naturalistic - left as it is and researcher doesnt interfere
Controlled - variables are regulated in a lab setting, know theyre being observed
Overt/covert observations
Overt - informed consent, know theyre being observed
Covert - unaware theyre being observed e.g. in a public place
Participant/ non-participant observations
Participant - observer is a part of the group being observed
Non - observer watches from a distance and doesnt interact
Event sampling
Count the number of times a certain behaviour occurs
Time sampling
Recording what behaviours happen in a certain amount of time
Features of a good questionnaire
Clear questions Avoid biased/leading questions Answers can be analysed Open/closed questions (likert scale, rating scale, fixed choice options) Start with easy questions Filler questions to mislead
Structured interview
Questions designed in advanced
Asked in a set order
Unstructured interview
More like a conversation
No set questions
Have a general aim
Expand/elaborate on answers
Semi-structured interview
Questions set in advance
Can ask follow up questions to expand
What does a scatter diagram show?
Correlations
Positive - both increase
Negitive - one increase, one decrease
Zero - no relationship
Correlation coefficient
Indicated the degree of a correlation
Perfect positive is +1
Perfect negitive is -1
How research is used irl
Use social influence
Memory research for eyewitnesses and cognitive interview
Emotional care in early child development
Mental health treatments
Understanding the brain
What is peer review
Research is reviewed by experts to check for mistakes, improve accuracy, suggest improvements and decide where to allocate funding
Quantitative data
Numbers/quantities
Easy to analyse
Lacks detail
Qualitative data
Words/quality of info
More detailed
Harder to analyse
Primary data
Collected first hand by researcher
Plan, design & conduct own experiment
More control
Time consuming/expensive
Secondary data
Data collected for a purpose other than the current one
Used in meta-analysis
Quicker & easier
Less reliable
Mean (central tendency)
Representative of all data as all values are used
Unrepresentative if extreme scores
Median (central tendency)
Not effected by extreme scores
Not a sensitive as mean
Mode (central tendency)
Useful if data is in categories
Not good if theres multiple modes
Standard deviation
Spread around the mean
Less affected by extreme values
Can be compred
Raw data tables
Data is unprocessed
Organised with clear headings
frequency table
Normal distribution
Mean = median = mode
Positive skew
Mode < median < mean
Negative skew
Mean < median < mode
When to use the sign test
Looking for a difference
Repeated measures design
Data in nominal ( in separate categories)
Statistical significance
Probability must be less than 5% for the results to have occurred by chance
Steps for doing a sign test
1) record the scores for both groups
2) subtract the after from the before and record the difference and + or - sign
3) count up the amount of + and -, ignore any that are 0, the lowest value is the S value
4) find the number of participants and compare this is the critical values table for 0.05
5) if the S value is equal to or less than the critical value then it’s significant
