1) Extra case (division) - this is detected by looking at the
totals (descriptive statistics), the numbers don't make sense because
of a different aggregation level.
2) Unlikely and Missing values in the following variables: Area, Population density and HH size.
3) Problems with ACPRVF and ACPRVM- within range on one variable, but unlikely compared with the other.
4) Column shift - look at variables after ACPRM in last case (?)
5) VACM variables - a combination of problems which is first
detected by summing VACMs into a new variable (VACTOT)
which should equal 100%. Then look at individual VACMs to see problems (e.g. VACM0)