Table 1.
List of tables provided as input into the VML, subject to minimum stated counts.
Fig 1.
Proportion of forenames in different ethnic groups estimated from cluster analysis of names in the 2011 Census (left) and from predictions by Onomap (right).
Fig 2.
Proportion of surnames in different ethnic groups estimated from cluster analysis of names in the 2011 Census (left) and from predictions by Onomap (right).
Table 2.
Frequencies of aggregate categories of self-assigned ethnicity in 2011 Census (England and Wales) rounded to the nearest 100 from calculations made using VML.
Fig 3.
Unweighted prediction accuracy for different age bands.
Fig 4.
Prediction accuracy for women (solid line) and men (dashed line) by age in each self-ascribed ethnic category (the small Arab/Mixed/Other category is not shown).
Fig 5.
Prediction accuracy for women (black bar) and men (grey bar) by marital status in each self-ascribed ethnic category (the small Arab/Mixed/Other category is not shown).
Fig 6.
Prediction accuracy according to region of residence (North East NE; North West NW; East Midlands EM; West Midlands WM; Yorkshire and Humber YH; East England EE; Greater London GL; South East SE; South West SW; Wales WA).
Fig 7.
Prediction accuracy for women (black bar) and men (grey bar) by region in each ethnic group (codes for regions as in Fig 6).
Table 3.
First set of follow-up algorithms and their main characteristics.
Table 4.
The Census-based surname lookup table for Ethnicity Estimator.
Table 5.
Example of assigning ethnic class using Ethnicity Estimator.
Table 6.
Aggregate prediction accuracy of all algorithms for each ethnic group.
Fig 8.
Prediction accuracy of OM-GA for women (solid line) and men (dashed line) by age in each ethnic group.
Table 7.
Instances of incompleteness of forename and surname pairs, and types of inconsistency between forenames and surnames.
Fig 9.
Prediction accuracy of EE-A by format of names in each ethnic group.
Table 8.
Re-weighted modifications of algorithms to mitigate the dominance of the ‘White British’ category for EE-A and EE-RS.
Table 9.
Aggregate prediction accuracy of re-weighted algorithms for each ethnic group.