Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Missing data in Likert ratings: A comparison of replacement methods

R G Downey1, C King

  • 1Institutional Research andAnalysis, 226 Anderson Hall, Kansas State University, Manhattan66506. downey@ksu.edu.

The Journal of General Psychology
|February 6, 1999
PubMed
Summary

When replacing missing Likert scale data, both item mean and person mean methods work well for up to 20% missing data. However, person mean substitution can inflate reliability estimates with higher missing data percentages.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

MATERNAL HEALTH LITERACY AND QUALITY OF CARE OFFERED TO PREGNANT WOMEN ATTENDING ANTENATAL CARE SERVICES IN LAGOS, NIGERIA.

Annals of Ibadan postgraduate medicine·2025
Same author

ASSESSMENT OF CAREGIVERS' KNOWLEDGE, RETENTION AND UTILIZATION OF CHILD HEALTH CARDS IN OYO STATE, NIGERIA: A MIXED METHOD STUDY WITH A CONVERGENT DESIGN.

Annals of Ibadan postgraduate medicine·2025
Same author

Derivation of a simple risk calculator for predicting clinical worsening in patients with pulmonary hypertension due to interstitial lung disease.

JHLT open·2025
Same author

Food insecurity and periodontitis in US adults.

Community dental health·2024
Same author

Laparoscopic low anterior resection for deeply infiltrating endometriosis - a video vignette.

Colorectal disease : the official journal of the Association of Coloproctology of Great Britain and Ireland·2024
Same author

Concussion injury management, perception, and knowledge in amateur field hockey.

South African journal of sports medicine·2024

Area of Science:

  • Psychometrics
  • Statistical Analysis
  • Survey Methodology

Background:

  • Missing data is a common issue in Likert scale surveys.
  • Improper handling of missing data can bias results.
  • Two common imputation methods are item mean and person mean substitution.

Purpose of the Study:

  • To evaluate the impact of item mean and person mean imputation on Likert scale data.
  • To determine the acceptable thresholds for missing data when using these methods.
  • To assess the effects on scale reliability and data representation.

Main Methods:

  • Simulated Likert scale data with varying percentages of missing values.
  • Applied item mean imputation technique.
  • Applied person mean imputation technique.

Related Experiment Videos

  • Analyzed changes in inter-item correlations and reliability estimates.

    • Main Results:

    • Both methods accurately represented original data when missingness was 20% or less.
    • Person mean substitution led to inflated inter-item correlations and reliability with increased missing data.
    • Item mean substitution consistently reduced reliability estimates.

    Conclusions:

    • Caution is advised when using person mean imputation for Likert scales with >20% missing data.
    • Item mean imputation may underestimate scale reliability.
    • The choice of imputation method depends on the extent and pattern of missing data.