-
Data
-
Defn.
- A series of observations, measurements, or facts that can be analysed
-
Variable
- Has a possible range of values
-
Analysis
- Gathering, modelling and transforming data with the goal of highlighting useful information, suggesting conclusions and supporting decision making.
-
Types
-
Nominal
-
Categorys
- No relationships
- Least powerful
-
Ordinal
- Rank
- Has a relationship (1st, 2nd etc.
- Non mathematical relationship
-
Interval
-
No real 'ZERO'
- eg. temperature
- Has a mathematical relationship
-
Ratio
- Has a tue 'ZERO'
- Eg. distance, height.
- Most powerful
-
Research Methods
-
Which method?
- Depends on the question
-
Quantitative
- Experiments
- Surveys
- RCTs
-
Numbers
- Tells you what happened
-
Qualitative
- Focus groups
- Interviews
- Case studies
- Tells you why it happened
-
Validity
-
Internal
-
to do with study design
- Is it ok?
- Are we measuring the right thing
- Eg measuring height for as a measure of intelligence is wrong.
-
External
-
Can it be applied outside?
- Can the results be generalised?
-
Replicability
- Can it be done again?
-
Reliability
- If experiment done again will the same results be given?
- Easier for lab based work
- Objective
- Unbiased
-
Variables
-
Must be operational
- Be explicitly stated
-
Constructs
- Defined by theoretical definitions
-
Variables
-
Quasi-independent
-
Characteristics that cant be randomly assigned
- Eg sex, age
-
True experimental variables
- Can control these in a true experiment
-
Can be randomly assigned
- eg Give Drug A or Drug B
-
Independent variables
-
The ones we control
- To bring about change in DV
-
Levels
-
At least TWO
- eg. gender- male/female
- AKA 'condition'
-
Dependent variables
- The ones we measure
- The ones that depend on the IV
-
Research design
-
Experimental design
-
True experiment design
- Design where researchers can randomly assign participants to experimental condition.
- Eg randomly assign normal participants to consume different amounts of alcohol
- Randomised
-
Quasi experimental design
- Design where researcher cant randomly assign participants to groups
- eG Compare heavy vs light drinkers, as you are either in one group or the other
-
Randomisation
-
Reduces confounding variables
- When groups to be compared differ in ways other than what the researcher has manipulated
- As they are distributed equally among the groups
- Prevents (un)intentional bias.
- Ensures participant is equally likely to be assigned to either group
- Enables use of powerful statistics.
-
Subjects design
-
Independent groups design
- Comparing BETWEEN groups
-
Potential problems
- Confounding factors
- Solutions
- Randomisation
- Matched groups
-
Matched groups
- Make sure subjects in both groups are matched as closely as possible on potential confounding factors
-
Repeated measures design
- Testing WITHIN groups
-
Advantages
- Fewer participant
- Each participant is their own control
- Removes some confounding factors
-
Disadvantages
- Order effects
- Cant return ppnt to original state
- Practice effect
- better performance due to practice
- Fatigue effect
- SOLUTION
- Counterbalancing
- Randomly assigning order to group
- Therefore we can know whether the order has made any difference
-
Causation
- How correct is our claim of A being the cause of B?
-
SOLUTION
-
Have a comparision group
- Eg treatment vs placebo
-
Could do O-X-O
- Eg. test, give alcohol, test.
- Therefore we know if alcohol is the cause
-
Forms of validity
-
Face
- Does it measure what it says it does?
-
Criterion
-
Concurrent
- Comparison of new test with established test
-
Predictive
-
Does the test have predictive value?
- Eg Does blood pressure value now predict heart attack in 5 years?
- Does the measured results agree with other measures of same thing?
-
Construct
- How well does the design tap into the underlying construct
-
Ecological
-
Does study reflect naturally occuring behaviour?
- Eg does mouse in box reflect its behaviour in wild?
-
Population
- Is our sample adequate for the claims we make about the population?
- What population are we interested in?
-
Sampling
- A sample is a selection or subset of individuals from the population
-
Why sample?
- Time
- Money
-
Sufficiency
- Maybe we dont need that much data as we feel that the sample gives an accurate data
- Access
-
How
-
Random sample
- No pattern
-
Systematic
- Drawn from the population at fixed intervals
-
Stratified
- Specified groups appear in numbers proportional to their size in population
-
Opportunity/Convenience
- People who are easily available
- Leads to bia
-
Snowball
- Get current participants to recruit more for the research
-
Useful if you want to recruit very specific population
- eg drug users might know other drug users