Introduction

The most useful figures we can state in describing how people with the Cree surname are spread around the world are the total numbers of Cree individuals in each area where they are found and the frequency of those individuals relative to the total population of the area.

The frequency of a surname in a particular area is calculated by dividing the number of people bearing that name in that area by the total population in the same area at the same time. This is normally presented either as a percentage or as a number per million of the total population.

The data used for the calculation may relate to the present day or to some time in history. It is important that we always state what time our population data relate to.

So the frequency of a surname represents how common it is now, or was at some stated time in the past. It means something only when the time and area to which it relates are stated.

Sources of data on surname frequencies

Official census data are the most accurate figures available for the calculation of surname frequencies. While there are some systematic errors in the figures, they are minimal compared to other sources and barely affect the use of Census data for our purposes.

Most availabe census data relate to historical times. For example UK Census data is not available until 100 years after the Census. However the US Census Bureau website lists Frequently Occurring Surnames from Census 2000 and gives both total numbers and frequencies of many thousands of surnames across the whole of the USA in the year 2000.

The most commonly used UK census for surname frequency in the UK is that of 1881 because it was the first from which data was indexed (by a joint project of the Federation of Family History Societies and the Chuch of Latter Day Saints). Most useful was the aggregation of 1881 Census data in Stephen Archer's Surname Atlas, produced as a CD which enabled surname frequencies and distribution maps of any specified surname to be obtained.

Other organisations have used electoral registers and telephone directories to estimate numbers of people with each surname. Inaccuracies arise because these sources do not reflect total populations. Children are not shown in either type of source, and many people elect not to have their names published in either the publicly available electoral registers or in telephone directories. Nevertheless some research organisations claim to have developed methods of making accurate corrections for these errors.

A leader in this field is the Centre for Advanced Spatial Analysis (CASA) of University College, London, which created the Surname Profiler website. Spin-offs from this research include Public Profiler, from which much of the Cree data in these web pages has been drawn, the formerNational Trust Names site and Mark Goodge's British Surnames website, although the latter now uses data from other sources. American Surnames is a sister site to British Surnames.

The Public Profiler site will create distribution maps for most surnames for some of the main English-speaking countries but the frequencies represented by the different colours on the maps are not specified, which reduces considerably the usefulness of the maps.

Cree surname frequencies around the world

Table 1 shows the frequencies of the surname CREE in about 2005 for those countries where it occurs. These figures are supplied by the Pubic Profiler web site. They are not very accurate as they are estimated from data abstracted from "publicly available telephone directories or national electoral registers" over the period 2000-2005. These are relatively inaccurate sources of population figures, even when, for example, corrections are applied to allow for the absence of children in the source data.

Public Profiler also shows the "regions" where the surname Cree has its greatest frequencies. They are shown in Table 2.

Public Profiler also gives the top ten "cities" for Cree numbers, though whether this is based on actual numbers or on frequencies is not made clear. I suspect the former as large cities like Glasgow and Newcastle would would struggle to be in the top ten for frequency. The top ten are shown in Table 3.

1	Kilmarnock, Scotland
2	Belfast, N Ireland
3	Newcastle upon Tyne, England
4	Glasgow, Scotland
5	Nottingham, England
6	Ayr, Scotland
7	Bangor, N Ireland
8	Leeds, England
9	Newark, England
10	Edinburgh, Scotland

Table 3.

Country	CREE Frequency per million	*Guild benchmark
United Kingdom	33.46	Low freq.
New Zealand	21.87	Low freq.
Australia	20.97	Low freq.
Canada	10.09	Low freq.
United States	6.86	Rare
Ireland	2.4	Rare

Table 1.

Area Name	CREE Frequency per million	*Guild benchmark
Scotland, UK	131.97	Medium freq.
N Ireland, UK	127.91	Medium freq.
Waipa District, NZ	126.76	Medium freq.
Thames-Coromandel District, NZ	92.34	Low freq.
Manawatu District, NZ	91.59	Low freq.
Horowhenua District, NZ	88.05	Low freq.
Porirua City, NZ	83.33	Low freq.
North Shire City, NZ	58.95	Low freq.
Kapiti Coast District, NZ	57.18	Low freq.
North of England, UK	55.15	Low freq.

Table 2. "Top Regions" for CREE frequency c2005, from Public Profiler web site.

*Guild benchmarks of surname frequencies

Guild benchmarks are standard terms for different ranges of frequency. They were proposed by the Guild of One-Name Studies in 1998 and are shown in this table.

(Source: The Journal of One-Name Studies, Vol 6, no. 6, April 1998, p 98.)