Trait approaches and applications to psychometric validation Flashcards
What are trait approaches?
They are a person-centred vs. situationism approach, they tend to not do well at accounting for development and focus on the present
What are Catell’s theories?
Personality predicts what a person will do in a situation
Desirable/undesirable, normal/abnormal is irrelevant
Rigorous scientific approach
Science demands measurement
What is Eysench’s theories?
He built on Catells ideas and thought personality to be largely governed by biology and shaped by childhood
What was Cattells, Eyseneck and Costa & MacCrae’s contributions to personality?
Eysenck proposed a hierarchical model emphasizing three dimensions (extraversion, neuroticism, psychoticism), highlighting the biological basis of personality and genetic influences.
Cattell’s key contribution was the 16PF model, defining 16 primary personality traits through factor analysis, enhancing understanding of individual differences.
Costa and McCrae developed the FFM, identifying five broad personality dimensions (openness, conscientiousness, extraversion, agreeableness, neuroticism), improving understanding of personality traits and their impact.
In Eysenecks Hierarchal 3 factor trait model, explain extraversion vs introversion, neuroticism versus emotional stability and psychoticism versus impulse control?
extraversion
Orientated towards outside world
Prefer company
Highly sociable, impulsive, assertive, dominant and adventurous
Neuroticism
Anxious, depressed, tense, irrational, moody
Largely inherited
More activity in sympathetic branch of the ANS
Hypersensitivity and emotionality
Psychoticism versus impulse control (P)
Aggressive, antisocial, tough-minded, cold egocentric versus impulse control (warmth and empathy)
Can be cruel, hostile, insensitive
Large genetic component
What is Costa and McCrae’s NEO personality inventory? (which included some elements of eyesnck’s model but introduced two more) - all factors fall on a continuum.
- Openness to Experience - Fantasy, aesthetics, feelings, actions, ideas, values
- Conscientiousness - awareness of actions and consequences, competence, order, dutifulness, achievement striving, self-discipline, deliberation
- Extroversion - warmth, gregariousness, assertiveness, activity, excitement-seeking, positive emotions
- Agreeableness - Trust, straightforwardness, altruism, compliance, modesty, tender-mindedness
- Neuroticism - anxiety, hostility, depression, self-consciousness, impulsiveness, vulnerability
What is a factor analysis?
A data grouping and data reduction technique
Based on the logic of the correlation coefficient
Measure all surface traits
Develop a correlation matrix (summarises correlations between each measure and all other measures)
Factor - cluster of related behaviour measures
Factor loading - extent to which each measure is related to each factor
How would you apply an EFA to OCEAN?
Each construct is correlated with each of the items - in OCEAN, neuroticism with every item and do that for all subsequent factors having a total of 150 pathways.
You use this when you don’t have much to guide you - no clear ideas of how these items load on these factors so you explore
An unrestricted FA
All latent constructs are correlated with all items
Focus is on establishing an underlying factor structure
Used for scale development, on items that haven’t been tested much
How would you apply a CFA to OCEAN?
5 factors, all the items, rather than correlating neuroticism with all the other factors, you only look at the paths between the factor and items related to it. Therefore, each factor has 6 paths.
You use this model when you have a very specific hypothesis - used with more confidence and structured and a clear hypothesis. Look at massive amounts of data, you need to make sense of it and find patterns in it.
Used to verify a factor structure
Based on theoretical reasons and/or past empirical research
Hypothesis a structure and test how the data fits with that structure
It is much harder to get a measure over the CFA because it is very restrictive
what is an Exploratory factor analysis?
EFA is an analysis technique used to explore and uncover the underlying latent factors or dimensions in a dataset. It helps identify the relationships between observed variables and group them into a smaller number of factors. EFA is often used when researchers do not have a predetermined factor structure or when they want to explore the underlying structure of a new measure. It allows for the discovery of potential patterns and associations among variables, providing insights into the underlying dimensions or constructs.
What is a confirmatory factor analysis?
CFA is a statistical technique used to test a pre-existing theory or hypothesis about the factor structure of a set of observed variables. In CFA, researchers specify a hypothesised factor structure based on theoretical considerations and then examine whether the observed data fit the proposed structure. CFA determines the extent to which the observed variables align with the expected factor structure. It helps assess the validity of the theoretical model and provides quantitative measures to evaluate the fit between the data and the proposed structure, such as goodness-of-fit indices.
What is Chi-Square (χ²) statistic?
The chi-square statistic tests the discrepancy between the observed data and the model-predicted data. A significant chi-square value indicates a lack of perfect fit. However, it is sensitive to sample size, and large samples tend to produce significant results. Therefore, it is often accompanied by other fit indices for a more comprehensive assessment.
A non-significant chi-square test indicates good fit, but it is often influenced by sample size.
what is Root Mean Square Error of Approximation (RMSEA)?
RMSEA represents the average discrepancy between the model and the observed data. It is a sophisticated fit index that tests how far a hypothesised model is from a perfectly fitting model.
Lower RMSEA values indicate better fit. Generally, values below 0.06 indicate good fit, while values between 0.06 and 0.08 indicate reasonable fit, and values above 0.10 suggest poor fit. In psychology go from <0.08
What is the comparative fit index?
The CFI compares the fit of the hypothesised model with a null model where no relationships exist between the variables. CFI values range from 0 to 1, with values closer to 1 indicating better fit. A CFI value above 0.95 is often considered indicative of good fit, although values above 0.90 may also be considered acceptable in some cases.
What is the Standardized Root Mean Square Residual (SRMR)?
The SRMR measures how much the model is different from what we actually observe. A smaller SRMR means the model is closer to what we see, while a bigger SRMR means there is more difference between the model and what we actually observe.
SRMR values below 0.08 suggest good fit.
