© J.H.Mathieson
Map Styles: The most common map styles used in the study of surname distributions are the proportional symbol and area fill maps. Each  form has its strengths and weaknesses. Area fill maps have the advantage that it is easy to interpret values as each colour represents a range of values. What it doesn't do as well is give a sense of the variation in the intensity of a distribution. By contrast, proportional symbol maps portray distributions well, but the values are often more difficult to interpret. Compound symbol maps can be very valuable when sorting out variants. Are two surnames related or did they evolve independently? Which one was the original which one was the variant? When two surnames where the spelling appears similar can be found spatially associated, there is a very good chance they are related. As an example we can illustrate the point with the Halkyard and Hawkyard surnames. The spellings are very similar. An  “l” has been substituted for a “w”, or visa versa. This could be the result of a transcription error or a misinterpretation due to dialect. If the errant spelling becomes fossilized a new surname variant is born. In the case of the Halkyard/Hawkyard question a Halkyard baptism took place in 1598 in Almondbury, and the earlies Hawkyard baptism took place in adjacent Huddersfield in 1693. It may be Halkyard is the original surname form. Standardized Data: Surname distributions are displayed by aggregating the frequency of surnames at specific locations. The most basic approach displays the distribution using raw numbers. While this is a perfectly acceptable methodology it is often useful to standardize data using “ratios”. For example surnames with a high frequency and widespread distribution, will reflect the impact of urbanization when using raw numbers . To eliminate this bias, a distribution can be “standardized” to take into account the relative size of the population within a given area. A surname could be expressed as X/100,000 population. The examples to the right illustrates the principle using the Bradshaw surname. Using raw numbers the surname distribution in the Liverpool Manchester region appears to be much higher when compared with a standardized distribution. The origin of the Bradshaw surname can likely be attributed to the numerous places called Bradshaw most of which are found some distance from the two urban centrers. The Banwell Index: The late Guild of One Name Studies member Eric Banwell introduced the “Banwell Index” to the study of surnames. Intensity indexes are widely used in geographic studies. The Banwell Index (BI) compares the frequency of a surname within an areal sub region, to the frequency within a larger area. If a surname’s incidence was calculated to be 1.2% of the national population, and the incidence within a Registration District was 4.8%, the BI would be 4.8% / 1.2% = 4.4. By plotting Banwell Indexes it is possible to identify significant regional concentrations of a surname. The Ashley surname can be used to illustrate the application of the Banwell Index. With an 1881 frequency of 5,005 the surname is concentrated in the English Midlands. Derived from a place name or a wood, there are likely several origins. We can filter out the distributions “noise” by plotting the BI with a value >= 5. The Registration Districts of Wem (31.7), and Market Drayton (23.6), constitute the core of the Ashley distribution in the West Midlands. The  place names of Ashley and Ashley Court are found nearby.