Relationships Among Ratio Variables with Common Components: Fact or Artifact

This paper examines relationships among ratio variables with common components. Simulation experiments are reported in which data samples with known relationships between ratios are analyzed. Observed ratio correlations are then compared to the known relationships. The simulation findings tentativel...

Full description

Saved in:
Bibliographic Details
Published inSocial forces Vol. 58; no. 4; pp. 1109 - 1128
Main Authors Macmillan, Alexander, Daft, Richard L.
Format Journal Article
LanguageEnglish
Published Chapel Hill, N.C The University of North Carolina Press 01.06.1980
University of North Carolina Press
Subjects
Component/Components/Componential
Correlation/Correlations/ Correlated/ Correlatives/ Correlational
Correlations
Crime rates
Criminal punishment
Jurisdiction
Mathematical variables
Population education
Population size
Ratio analysis
Ratio/Ratios
Ratios
Statistics/Statistical/ Statistically
Variable/Variables/ Variableness
Zero
Online AccessGet full text

Cover

More Information
Summary:This paper examines relationships among ratio variables with common components. Simulation experiments are reported in which data samples with known relationships between ratios are analyzed. Observed ratio correlations are then compared to the known relationships. The simulation findings tentatively suggest the absence of bias or spurious correlation. The sharp contrast between our findings and the results of simulation experiments carried out by ratio variable critics is examined. Finally, based on a more general mathematical argument, ratio correlations with common components are shown to be a valid form of analysis.
Bibliography:istex:C7FF956C089206D5C0E68B42844F5CBB422EB0EB
ark:/67375/HXZ-NQQZ5DR4-D
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ISSN:0037-7732
1534-7605
DOI:10.1093/sf/58.4.1109