Sir: A T L Foster-Barnes (letter, 26 February) might be right in asserting that there may be ethnic variation in crime perpetration. But how could such possible differences be "relevant to suspect-search in large communities?"
Perhaps it is being suggested that if, for example, black people were shown to be involved in proportionately more crime than other ethnic groups, this would aid police because it would indicate they should spend more time searching for black suspects.
May I suggest a different approach to identifying likely suspects? The police could collect evidence regarding a crime and on the basis of this, and not preconceived notions, narrow their search. A novel method perhaps?