Week 2 Lecture 3 - personality approaches 2 Flashcards
What is the Lexical Hypothesis?
all aspects of individual personality can be described from single words used in language
Allport and Odbert (1936) - drawing on Baumgarten’s work - collected all personality terms from Webster’s Dictionary.
What did they do with these words?
- Total 555,000 words
- selected all terms referring to behavioural differences (18,000)
- removed terms relating to cognitive, physical or transient states (4500)
- Much overlap in meaning of terms –> 4500 terms likely represent smaller number of distinct terms
What are latent (hidden) variables?
psychological variables (individual differences) that cannot be directly observed
How are latent variables measured?
must be inferred/estimated on the basis of behaviour and self-reported experience
What are psychometircs?
- the scientific measurement of psychological variables
- measuring the mind
Explain how the latent variable of extraversion might be measured
measures of if a person:
- enjoys talking to others
- enjoys socialising with people
- makes friends easily
- are the life and soul of the party
- relates to words such as talkative, sociable, friendly and lively –> infer extraversion
What do correlations do in relation to latent variables?
enables us to determine which terms are related to one another and so represent a common “latent factor”
- reduction of terms to latent factors
If there is a strong positive correlation between 2 terms, what does this suggest?
they are related to a common latent factor
If there is a strong negative correlation between 2 terms, what does this suggest?
Related to a common latent factor –> just on different ends of the same trait
i.e., extraverted and not extraverted
If there is no correlation between the two terms, what does this suggest?
not related to a common latent factor
independent
What is a correlation matrix? Is is good for assessing relationships between terms?
- shows the relationship between each pair of terms
- difficult to work out which items belong together –> doesn’t provide a clear picture
What is factor analysis?
- a multivariate data reduction technique
- looks for the set of latent variables (factors) that best account for the pattern of correlation within the dataset
- is a statistical technique that uses patterns of correlations between several variables to identify a smaller number of underlying latent variables
- values show correlations for each items and each underlying factor –> factor loadings
What must a researcher decide when using factor analysis?
- how many factors to extract
- what to name the factors based on the items that load on it
What are cross loadings?
- items should load strongly on only one factor
- important because aim of factor analysis is to have independent factors
How did Cattell construct his 16 personality factor model?
- took the 4500 words and sorted them into clusters of synonyms
- paired clusters with matched antonyms
- selected best example from each pair (171 words)
- 100 people rated an individual on each of these 171 terms and the correlations examined to identify 60 clusters
- Cattell then added a few more terms from psychiatric literature that he determined important
- reduced this to 45 terms based on own judgment and literature guidance
- factor analysed 45 “surface traits” to identify 16 personality factors (source traits) which explained the surface trait variability