Chi Squared
- tests a difference or association
- independent data
- nominal data
Degrees of freedom- (no. rows - 2) x (no. columns - 1)
Pearsons R
- parametric
- correlation
- related/ repeated measures
- interval data
Df= N-2
Spearmans Rho
- correlational
- non parametric
- repeated measures/ related data
- ordinal data
Unrelated T Test
- independent groups
- tests a difference
- interval data
Df= total no. people - 2
Related T test
- repeated design
- interval data
- tests a difference
Df= total no. people - 1
Wilcoxon T Test
- ordinal data
- non parametric
- repeated measures/related data
- tests a difference
Mann Whitney Test
- non parametric
- tests a difference
- independent groups
- ordinal data
Content analysis
Structured observation to make qualitative data quantitative by counting themes. \+ good ethics \+ high ecological validity \+ can operationalise \+ easier to compare numbers \+ can replicate - observer bias - have to train observers - reduces complex processes to pre-determined ideas
Time sampling
Observations at regular intervals
Event sampling
Tally each time a behaviour occurs
Features of science
Empiricism Objectivity Replicability Theory construction Hypothesis testing
Type 1 error
False positive
Accept hypothesis as significant when its not
Type 2 error
False negative
Reject hypothesis, accept null hypothesis when its significant
The scientific process
Induction Deduction Hypothetico deductive method Falsification Kuhn's paradigms
Induction
Development of general theories based pn what you see. Reason from particular to general.
Deduction
Start with a theory and look to confirm this.
General to particular.
Hypothetico deductive methods
Proposed by Karl Popper (1935): theories first then generate expectations/hypotheses which can be falsified
Falsification
One observation that disproves the theory.
E.g. All flamingoes are pink/one white in disproves theory.
Kuhn’s Paradigms
Theories people believe.
E.g. World is flat.
- Normal science
- People question dominant paradigm
- Evidence
- Paradigm shift
Inferential statistics
Allow us to draw conclusions from the data, about the significance of the results.
Descriptive statistics
- measures of central tendency
- graphs
- summary of data
Probability
Chance of any difference occuring.
Total no. outcomes x happens
/
Total no. of possible outcomes
Significance
How sure we are that any difference/correlation is meaningful (not down to chance).
Correlation
Correlation coefficient shows direction:
-1 = perfect negative correlation
+1 = perfect positive correlation
Correlation coefficient shows strength:
0= weak
1= strong
