inferential statistics Flashcards
Alternative hypothesis
A clear, precise, testable statement that states the relationship between variables to be investigated. Variables must be operationalised.
Null hypothesis
Stating there is no difference or correlation between variables and that if one is found it is due to chance
Operationalisation
Clearly defining variables in terms of how they can be measured
Independent group designs INCLUDING STREGNTHS+CONS
each participant takes part in one condition of the IV-randomisation of groups
+lower risks of demand characteristics- naive participants only do one condition
+no order effects
-individual differences makes it harder to compare groups
-requires more participants than repeated measures
Repeated design measures INCLUDING STRENGTHS+CONS
Participants take part in all conditions of the IV and counterbalancing is required
+less participants needed than Independent
+controls individual differences- researchers are more certain that the IV is causing the difference
-higher risk of demand characteristics
-risk of order effects
Matched pairs design INCLUDING STRENGTHS+CONS
Participants are matched with another participant who shares a relevant characteristic for the study and split into different conditions
+redoes the effects of participant variables
+no order effects-only one condition
+lower demand characteristics- only one condition
-cannot match participants identically- individual differences still slight difference
-practical issues- time, cost
Type 1 errors
FALSE POSITIVE
when you wrongly accept the alternative hypothesis as significant when it is not
Error of optimists
Type 2 errors
FALSE NEGATIVE
Wrongly rejecting the alternative hypothesis as not significant when it is
Error of pessimists
Probability
Likelihood of something
Inferential testing
Number and formulae that allows researchers to draw conclusions from researcher to say whether there is a real correlation or difference
