Answers, Correct Responses Marked, and Detailed Explanations for Each Question to Support Mastery
of Exercise Science Research Terminology.
Questions 1–150 (Version 3)
1. A construct is…
a) A directly observable behavior
b) An abstract, theoretical concept (e.g., motivation, fatigue)
c) A statistical test
d) A type of error
Correct Answer: b
Explanation: Constructs are not directly measurable; they require operational definitions.
2. An operational definition…
a) Defines a construct by how it is measured or manipulated
b) Is the same as a dictionary definition
c) Cannot be used in experiments
d) Only applies to qualitative research
Correct Answer: a
Explanation: Operational definitions make constructs measurable (e.g., “VO2max” as operational
definition of aerobic fitness).
3. A latent variable is…
a) Directly observed in data
b) An unobserved variable inferred from observed indicators
c) Always categorical
d) The same as an outlier
Correct Answer: b
Explanation: Latent variables (e.g., depression, self-efficacy) are measured via multiple items or tests.
4. A manifest variable is…
a) Unobserved
b) Directly measured or observed
,c) Theoretical
d) A latent construct
Correct Answer: b
Explanation: Manifest = observed variable (e.g., heart rate, test score).
5. Factor loading is…
a) Correlation between manifest and latent variable
b) Difference between two means
c) p-value
d) Sample size
Correct Answer: a
Explanation: High factor loading (>0.70) indicates item strongly measures the latent factor.
6. Eigenvalue represents…
a) Variance explained by a factor
b) Correlation coefficient
c) Alpha level
d) Standard error
Correct Answer: a
Explanation: In factor analysis, eigenvalues >1 are often retained.
7. Varimax rotation in factor analysis…
a) Correlates factors
b) Orthogonal rotation simplifying factor structure
c) Oblique rotation allowing correlated factors
d) Removes factors
Correct Answer: b
Explanation: Varimax = uncorrelated factors, easier interpretation.
8. Oblique rotation (e.g., Promax)…
a) Forces factors to be uncorrelated
b) Allows factors to correlate
c) Removes all cross-loadings
d) Reduces eigenvalues
,Correct Answer: b
Explanation: More realistic for psychological/behavioral constructs.
9. Communality (h²) is…
a) Proportion of variance in an item explained by factors
b) Unique variance of item
c) Error variance
d) Factor loading squared summed across factors — yes
Correct Answer: a
Explanation: Low communality (<0.40) suggests item poorly represents factors.
10. A scree plot helps determine…
a) Sample size
b) Number of factors to retain (elbow point)
c) Reliability
d) Statistical significance
Correct Answer: b
Explanation: Scree plot eigenvalues: retain factors before leveling off.
11. Parallel analysis compares eigenvalues to…
a) Random data eigenvalues
b) Zero
c) Alpha level
d) Factor loadings
Correct Answer: a
Explanation: More accurate than eigenvalue >1 rule.
12. Structural Equation Modeling (SEM) combines…
a) ANOVA and t-test
b) Factor analysis and regression
c) Chi-square and correlation
d) Descriptive statistics only
Correct Answer: b
Explanation: SEM tests complex causal models with latent variables.
, 13. In SEM, a path coefficient is…
a) Standardized regression weight
b) Factor loading
c) Reliability coefficient
d) p-value
Correct Answer: a
Explanation: Path coefficients show strength and direction of relationships.
14. Model fit indices in SEM include…
a) CFI, RMSEA, SRMR
b) Cronbach’s alpha
c) Levene’s test
d) t-test
Correct Answer: a
Explanation: CFI ≥0.95, RMSEA ≤0.06, SRMR ≤0.08 indicate good fit.
15. RMSEA (Root Mean Square Error of Approximation)…
a) Measures parsimony-adjusted fit
b) Measures internal consistency
c) Is a test of normality
d) Measures power
Correct Answer: a
Explanation: Lower RMSEA (<0.06) = better fit.
16. CFI (Comparative Fit Index) ranges…
a) 0 to 1 (≥0.95 good)
b) -1 to +1
c) 0 to infinity
d) 0 to 100
Correct Answer: a
Explanation: CFI compares model to null model.