Self and Observer Reports in Personality Assessment

Self vs. Observed Reports on Personality Inventory Scales

This section delves into combining self and observer reports in personality assessment, highlighting the complexity and validity of personality measurement.

Combining Self and Observer Reports

• Self-reports are obtained from target individuals who complete questionnaires like the Big Five Inventory.
• Observer reports are gathered from others who describe the same target person using the same questionnaires.
• High agreement between self and observer reports supports constructive validity, indicating the personality scales measure what they intend to measure.

Validity of Personality Assessments:

• The correlation between self-reports and spouse's observer reports on the Five-Factor Inventory is around $0.6$, suggesting spouses know individuals well.
• Correlations with friends or neighbors as observers are around $0.4$ to $0.5$, lower than those with spouses, indicating familiarity influences agreement.
• Similar correlations (between $0.4$ and $0.6$) are found in HEXACO-PIR studies with college students.

Convergent Validity:

• High agreement between self and observer reports demonstrates convergent validity, suggesting the scale measures the intended trait.
• However, agreement could also arise from shared inaccurate opinions about the target's personality.
• To differentiate between accurate and inaccurate agreement, compare observer reports from multiple, unacquainted observers from different contexts.

Studies on Target-Observer Agreement

Ostander & Fander Study

• Examined agreement on Big Five traits (Neuroticism, Extraversion, Openness, Conscientiousness, Agreeableness) between targets and parents, college friends, and hometown friends.
• Agreement was high between self and parents (correlations from $0.34$ to $0.5$) and self and college friends (correlations from $0.34$ to $0.53$).
• Agreement with hometown friends was lower, especially when averaging correlations.
• College friends know individuals almost as well as parents, whereas hometown friends know them less well, potentially due to changes after leaving home.

Fund et al. Study

• Focused on inter-observer agreement: how much observers agree among themselves about the target person's personality.
• Compared agreement among observers from different contexts (parents and college friends, parents and hometown friends, college and hometown friends) and the same context (two college friends, two parents, two hometown friends).
• Observers from different contexts agreed less about the target's personality than those from the same context.
• Higher correlations among observers from the same context suggest shared biases influencing judgments.
• Observer assessments are heavily influenced by the context in which they know the target.

Who Knows You Best?

The accuracy of self-reports versus observer reports is debated. Self-reports benefit from access to internal thoughts and feelings, but may suffer from biases. Observer reports offer external perspectives, but may lack comprehensive knowledge.

Kola et al. Study

• Compared self-reports, observer reports, and aggregated reports in predicting behaviors.
• Observer reports sometimes predicted behaviors better than self-reports, indicating potential gaps in self-knowledge or reporting biases.
• Combining self and observer reports improved prediction accuracy.
• People with depression may exhibit negative biases in self-reports, leading to underreporting of positive behaviors.

The Self-Other Knowledge Asymmetry (SOKA) Model

This model explains when observer reports are better predictors of behavior than self-reports, based on trait observability and evaluativeness.

Key Concepts:

• Observability: Whether a trait is easily observed by others (external) or more internal (thoughts, feelings).
• Evaluativeness: Whether a trait has a strong positive or negative evaluation associated with it.

Trait Types

• Internal, Low-Observable Traits: Include thoughts, feelings, anxieties, and self-esteem, which are not easily observed by others.
• External, High-Observable Traits: Primarily overt behaviors, such as extroversion or talkativeness, easily observed.
• High Evaluative Traits: Such as laziness or intelligence, have considerable biases in self-reporting due to social desirability or impression management.

Model Predictions

• Accuracy of self and friend ratings depends on the trait's observability and evaluativeness.

Empirical Support

An examination of traits with varying degrees of observability and evaluativeness confirmed the model's predictions:

• Low observability and low evaluativeness (e.g., anxious): Self-reports are more accurate.
• High observability and low evaluativeness (e.g., talkative): Observer reports are slightly more accurate.
• Low observability and high evaluativeness (e.g., intelligence): Observer reports are significantly more accurate.

Conclusions

• Self and observer reports show fairly high levels of agreement.
• Individuals provide reasonably accurate descriptions of their own and others' personalities.
• Both self and observer reports can predict behavior, particularly when behaviors are aggregated.
• Self and observer reports capture different aspects of personality, with varying accuracy depending on the trait.

Limitations of Self and Observer Reports

• Biases: Socially desirable responding, self-deception, and halo effects can distort self and observer reports.
• Parents may overestimate positive aspects of their children, while teachers may underestimate positive aspects of unruly students.
• The more sources of information, the less bias and the more accurate the assessment.

Key Considerations

• Studying personality traits requires understanding the different information provided by self and observer reports.
• The SOKA model clarifies these differences based on observability and evaluativeness.
• Differences between observer reports matter, depending on whether observers know the individual from the same or different settings.
• Observability and evaluativeness of traits influence the accuracy of self versus observer reports.
• Traits with low observability are better reported in self-reports. Traits high in evaluativeness are better reported in observer reports.