📄 Abstract
Likert-type scales are extensively used in academic research to measure attitudes, perceptions, behaviours, and other latent constructs; however, methodological decisions concerning the neutral response option and aggregation of multiple items remain insufficiently addressed in inferential statistical analysis. This analytical paper examines the role of the neutral option and composite aggregation of Likert-type responses in strengthening the validity, interpretability, and theoretical alignment of inferential analysis. Using a thematic content analysis and synthetic review of more than 20 recent journal articles and established statistics and research-methodology textbooks, the paper examines response-category design, neutral responses, measurement error, construct validity, reliability, composite scoring, multicollinearity, and regression modelling. The synthesis indicates that a neutral category can preserve meaningful respondent information where neutrality represents a legitimate position, although it should be distinguished from uncertainty, lack of knowledge, and non-applicability. The analysis further establishes that multiple statements measuring a theoretically coherent construct should, after appropriate reliability, dimensionality, and validity assessment, generally be aggregated into a composite score for construct-level inferential analysis. Composite aggregation improves parsimony, reduces unnecessary item-level redundancy, strengthens interpretability, and facilitates alignment between theoretical constructs and statistical models. Nevertheless, aggregation should not be applied mechanically because item-level predictors may provide advantages in prediction-oriented research. The paper concludes that valid neutral responses preserve the authenticity of measurement, while valid composite aggregation preserves the integrity of construct-level inference. Researchers are therefore encouraged to make response-category and aggregation decisions on theoretical, psychometric, and analytical grounds rather than convenience.
🏷️ Keywords
📚 How to Cite:
Bashir M. Maalim , George M. Wanjobi , THE NEUTRAL OPTION AND COMPOSITE AGGREGATION OF LIKERT-TYPE RESPONSES: IMPLICATIONS FOR INFERENTIAL STATISTICAL ANALYSIS IN SOCIAL SCIENCE RESEARCH , Volume 13 , Issue 8, August 2026, EPRA International Journal of Economics, Business and Management Studies (EBMS) , Pages: 139 - 146 ,