To investigate the concept of ethnic density one would have to draw a sample representative of the rich diversity of ethnicities in England. Modood & Berthoud ^{Reference Modood and Berthoud1} define ethnic group as:

The study by Das-Munshi et al ^{Reference Das-Munshi, Bécares, Boydell, Dewey, Morgan and Stansfeld2} included White British, Irish, Black Caribbean, Bangladeshi, Indian and Pakistani respondents. In the EMPIRIC study, ethnicity was defined by self-assessment using the same categories as the 1991 Census. ^{Reference Weich, Nazroo, Sproston, McManus, Blanchard and Erens3} The 1991 Census collected data on nine ethnicity groups: White, Black Caribbean, Black African, Black other, Indian, Pakistani, Bangladeshi, Chinese and any other ethnic group. ⁴ Das-Munshi et al's study has not looked at four of the nine ethnic subgroups listed and is therefore not representative of the ethnicities in England.

Das-Munshi et al have not explained the significant nonparticipation of candidates (37.8%) in the study, making it difficult to draw conclusions from the data. Owing to various factors – including that of participants having moved out of the survey area or to an unknown new address, or the participant having died or reached the age of 75 after the sample was drawn – 11% of the sample was no longer eligible for interview. ⁵

Of eligible individuals, refusals were received in person directly from the selected participant in 18% of cases, a further 2% refusing by post and 2% being proxy refusals. Most non-contacts resulted from the interviewer being unable to make contact with the participant, although there were also cases where no contact was made with anyone at the household after four or more telephone call attempts. The most common reason for any other unproductive outcome was that the participant was away on holiday or in hospital throughout the survey period. ⁵