SEI Research Methods and Evidence 2 — Questions and Answers
Question 1: In SEI research, what does a Cronbach's alpha of 0.85 indicate about a subscale?
- Poor reliability requiring scale revision
- Acceptable internal consistency among items (Correct answer)
- The scale measures too many unrelated constructs
- Items are redundant and should be removed
Correct answer: Acceptable internal consistency among items
A Cronbach's alpha of 0.85 indicates good internal consistency, meaning the items within a subscale reliably measure the same underlying construct.
Question 2: Which research design would best establish that EQ training causes improved workplace performance rather than merely correlating with it?
- Cross-sectional survey of high performers
- Randomized controlled trial with pre/post measurement (Correct answer)
- Retrospective case study analysis
- Meta-analysis of existing EQ studies
Correct answer: Randomized controlled trial with pre/post measurement
A randomized controlled trial with random assignment and pre/post measurement is the gold standard for establishing causal relationships.
Question 3: The SEI uses norm-referenced scoring. What does this mean for interpreting results?
- Scores are compared against a fixed absolute standard
- Scores are interpreted relative to a comparison population (Correct answer)
- Every score above 50 is considered emotionally intelligent
- Scores are based solely on behavioral observations
Correct answer: Scores are interpreted relative to a comparison population
Norm-referenced scoring means an individual's scores are compared to the performance of a normative reference group, not an absolute standard.
Question 4: A researcher finds that SEI scores differ significantly by industry sector. This finding is best described as evidence of:
- Construct validity
- Predictive validity
- Discriminant validity
- Group-level differences in the construct (Correct answer)
Correct answer: Group-level differences in the construct
Significant mean differences across groups indicate that the construct being measured varies systematically across those populations.
Question 5: Which statistical method is most appropriate for examining whether SEI competencies form distinct but related clusters?
- Multiple linear regression
- Confirmatory factor analysis (Correct answer)
- Independent samples t-test
- Pearson correlation only
Correct answer: Confirmatory factor analysis
Confirmatory factor analysis tests whether a hypothesized factor structure fits the observed data, making it ideal for validating competency clusters.
Question 6: In an SEI validation study, convergent validity would be demonstrated by:
- Low correlations between SEI and IQ measures
- High correlations between SEI and other validated EI tools (Correct answer)
- SEI scores predicting performance better than personality tests
- The absence of gender differences in SEI scores
Correct answer: High correlations between SEI and other validated EI tools
Convergent validity is shown when a measure correlates highly with other instruments designed to assess the same or similar constructs.
Question 7: A study reports that coaching using SEI feedback led to a 0.4 standard deviation improvement in scores. How should this effect size be interpreted?
- Trivially small, not worth reporting
- Small to medium, practically meaningful in applied settings (Correct answer)
- Large, indicating dramatic transformation
- Cannot be interpreted without a p-value
Correct answer: Small to medium, practically meaningful in applied settings
Cohen's d of 0.4 is considered a small-to-medium effect size, which is often practically significant in organizational and coaching interventions.
SEI Research Methods and Evidence 2
In SEI research, what does a Cronbach's alpha of 0.85 indicate about a subscale?