variables, design and hypotheses Flashcards
1
Q
what is an experimental design
A
- vary an independent variable whilst holding everything else constant
- measure changes in the dependent variable
- changes in the DV should be due to changes in the IV
- we can infer causality
2
Q
what is a quasi experimental design
A
- the IV cannot be manipulated
- examples include non-equivalent groups and pretest-posttest designs
- can be trickier to eliminate all confounding variables
3
Q
what is a correlational design
A
- no manipulations are made
- measure two or more variables and determine the extent to which they are related to each other
- we cannot infer causality
4
Q
what is the independent variable
A
- the variable we manipulate
- an experiment can have 1 or more IVs
- each IV should have 2 or more levels
5
Q
what is the dependent variable
A
- what we measure
- an experiment should have 1 or more DVs
- we should operationalise our DV - specify how we measure it
6
Q
what is nominal data
A
- non-numerical categories
7
Q
what is ordinal data
A
- discrete numbers that are in a certain order e.g. happiness levels
8
Q
what is interval data
A
- values that have a meaningful difference between them e.g. temperature
9
Q
what is ratio data
A
- values that have an absolute zero e.g. height, weight, income - no minus
10
Q
what are confounding variables
A
- a variable that was not manipulated but could have an influence on the results of an experiment
- want to eliminate as much as possible
11
Q
what is a between-subjects design
A
- pps only take part in one level of the IV
- can account for individual differences if we randomly assign pps to one of the groups
- less powerful - we need more pps for a genuine effect
12
Q
what is a within-subject design
A
- all pps do all conditions
- also known as repeated measures
- more powerful - fewer pps needed
- could be effected by order effects - so use randomisation of trails or counterbalancing
13
Q
what is a matched subject design
A
- want to do within subjects design but can’t
- pp matched to somebody else with regards to demographic characteristics
- the ‘pair’ are tested as one individual over two levels of an IV
14
Q
what is a hypotheses
A
- a theory-driven idea as to why a narrow set of phenomenon occur
15
Q
what is an experimental hypothesis
A
- a conceptual idea that tries to explain an observation
- based on our original theories
16
Q
what is a statistical hypothesis
A
- a specific statement that we can use to collect data and test our hypotheses with
- also known as a prediction
17
Q
what is a null hypothesis
A
- no difference
- our observations from our samples imply that they come from the same population
- for parametric statistics all means are equal
- for non-parametric statistics all distributions are equal
18
Q
what is the alternative hypothesis
A
- there will be a significant difference between variables
- can be directional or non-directional
19
Q
what are the properties of null and alternative hypotheses
A
- mutually exclusive - only one statement can be true
- exhaustive - they cover all possible outcomes in an experiment
20
Q
what happens in null hypothesis significance testing
A
- we only reject the null hypothesis when the probability of it being true is lower than a specific criterion
we figure this out by: - generating a test statistic
- setting a specific criterion
- using these values to help determine our probability
21
Q
what is usually the set criterion
A
0.05 - we want to see if p is less or greater than this value
22
Q
if p<.05
A
- less than 5% probability that results are due to chance
- something unique happening between populations
- found a significant result
- we can reject the null hypothesis
23
Q
if p>.05
A
- there is 5% or more probability these events happened by chance
- nothing unique happening between populations
- non-significant result
- failed to reject null hypothesis
