week 6 PTTR (validity and random assignment) Flashcards
what is external validity?
the extent to which we can generalise the results of the research to people, settings, times, measures and characteristics other than those of the study
what are three types of generalisation?
generalisation from a sample to the general population
generalisation from one research study to another
generalisation from a research study to a real world situation
what is internal validity?
if a study produces a single, unambiguous explanation for a relationship between two variables
what threatens internal validity?
any factor that offers an alternative explanation for a relationship between two variables
what threatens external validity?
any characteristics of a study that limits the ability to generalise the results
what are 5 threats when considering generalising from a sample to the general population
- selection bias (was the researcher biased when selecting the sample)
- is the sample college students (most behavioural studies are). SEARS argues that they are likely to have less formulated sense of self, stronger tendency to comply with authority, less stable peer relationships and higher intelligence
- volunteer bias
- participant characteristics- e.g social demographics
- cross species generalisation
what are 3 things to consider when generalising across features of a procedure
- NOVELTY EFFECT
- in the novel situation of participating in a study individuals may perceive and respond differently than they would in the real world - MULTIPLE TREATMENT INTERFERENCE
- if a person is participating in multiple conditions, effects can carry over that impacts performance in the next treatment
- for example: fatigue and practice
- this means the participation in previous treatment can be a threat to external validity as the results obtained from those who have may not be generalisable to those who havent already participated - EXPERIMENTER CHARACTERISTICS
- demographics and personality characteristics of the experimenter can limit the generalisability of the results
what are 3 things to consider when generalising across features of the measures
- SENSITISATION
- can repeated experience of the measure lead to sensitisation of the measure
- for example, when a pretest measure followed by a posttest measure is involved, pretest sensitization can occur
- this may be the case if the first measure leads to an increased awareness of the topic
- for example, the process of self monitering of mood has been shown by itself to improve depression - GENERALITY ACROSS RESPONSE MEASURES
- some variables can be measured in different ways
- for example fear could be measured by heart rate or mood questionaire
- if only one measure is used, the results may not be generalisable to other ways of measuring the variable - TIME OF MEASUREMENT
- effect of treatment/experiment may decrease or increase over time
- if only measured at one point in time, it may not be generalisable to if it was measured at other points in time
what’s the difference between a within subjects design and a between subjects design
- within subjects designs are when the different groups of scores are all obtained from the same group of participants
- between subject designs are when the different groups of scores are from different groups of participants
what are advantages or disadvantages of between subjects designs?
ADV
- the scores are uncontaminated from other treatment factors
- e.g practice, fatigue or contrast effects
DISADV.
- requires large number or participants
- individual differences between participants in different groups
why are individual differences a disadvantage of between subjects designs?
- individual differences can become confounding variables
- individual differences can lead to high variability in scores, making it difficult to determine whether the treatment has any effect
what are two major sources of confounding variables in between subjects design?
- confounding from individual differences
- confounding from environmental differences
e. g if one group is in a bigger room or smaller room etc.
what are 3 rules for different groups in between subjects design
the seperate groups must be:
- created equally
- treated equally
- composed of equivalent individuals
what are 3 ways of assigning groups in between subjects design?
- randomisation
- matched assingment
e. g matching no. males and females - restricting range of variability
e. g deciding females only
what are 4 ways to minimise variance within treatments?
- standardise procedure and treatment setting
- limit individual differences
- dramatically increase sample size