Statistically evaluating psychology lecture 2a & 2b test Flashcards
What do quantitative methods measure?
They measure aspects of human psychology
What is quantification
Taking human behaviour and coding it numerically
What are numeric codes
They are variables
E.g. we would give aggression a numeric code
Example of no variance
If human behaviour was constant both within and across individuals then there would be no variance
Variance (differences)
It is the difference that we measure and try to explain in psychology
Assess variance within our sample
Variables
The things that change
Variables can be:
Categorical (belonging on a scale)
Continuous (along a scale)
There are many ways variables can be defined, this can affect the conclusion of the study
Need to operationalise variables (define how we will quantify these constructs)
Psychological variables
Psychological variables are often hypothetical constructs - treating these as real is called reification
They assign an intangible?abstract label to human behaviours
E.g. Is measuring memory the same as measuring height
What should operationalised variables be?
- Reliable: test-retest reliability
2. Valid
Cause and effect
The change in one variable causes the change in another variable
In an experiment we are seeking evidence for a casual relationship between variables
E.g. Is Donald trump really causing my aggression
Larrick et al. 2011
Assessed 57,293 baseball games
Pitcher more likely to hit the batsman in hotter temperatures
Increases in hostility and reduced inhibition of retaliation
Aforementioned design
Aforementioned designs
NOT experiments due to:
Do not control whether hot or not
Do not control who is in the hot group
Leads to alternative explanations for the findings
Confounding variables
Uncontrolled variables that may provide an alternative explanation
Rule et al. 1987
Randomly assigned people to a normal or hot room
Participants instructed to complete an ambiguous story that could be interpreted aggressively or not aggressively
Those in hot room used more frustrations, negative emotions and aggression in story completion task
True experiment - Rule had control over whether it was hot or not
Rule randomly allocated people to hot or temperate room conditions
True experiment
We have control over IV
Randomly allocate people to different conditions of the IV
Everything else is held constant
Measure the effects of this IV manipulation on the DV
Attempt to isolate cause and effect
IV
Variable being manipulated
Conceptualised in terms of levels - number of conditions/groups for the IV
How we administer the IV determines the type of design
DV
Variable being measured as an outcome
Between-participant design
If there are different people in each level of IV it is an independent sample (between-participants design) Advantages: P's don't take part in both conditions - less likely to guess manipulation Related - no practice/fatigue Dont need to produce more stimuli Disadvantages: Potentially non-equivalent samples Need more p's
Within-participants design
If the same people are completing each level of the IV it is an repeated measures (within-participants design)
Advantages:
Remove group differences as same people in all conditions
Need less p’s
Disadvantages:
Order effects
Non-equivalent sample
Counterbalancing (within-participant design)
Can limit effects of p’s doing condition A followed by condition B by counterbalancing
Involves:
Half of p’s undertaking condition A followed by condition B
Half of p’s undertaking condition B followed by condition A
Known as ‘ABBA’
Order effects spread equally across 2 conditions
Asymmetric transfer - counterbalancing can sometimes result in asymmetric effects
Matched pairs
Different people in each group
‘Match up’ similar people to create a de facto repeated measures/within participant design
Advantages:
No order effects
Non-equivalent groups issues are reduced/removed
No issues of non-equivalent stimuli
Disadvantages:
Cant ever truly ‘match’ p’s - issues with non-equivalent samples
Matching is time-consuming and potentially difficult
