Trade-off Between Data Collection Methods, Bias and Costs
| Face-to-face
Interview Survey |
|
Can be high with call-backs.
Needs pre-information and publicity to improve involvement.
Contact rate can then be c70% |
Can be Low |
| Telephone Survey |
|
Can be high, using call-backs
and pre-information helps contact rate. |
Selection bias
likely to be high (given reducing use of land-lines, TPS etc.
Some response bias likely |
| Postal Survey |
|
Response rates seem to be
declining (40% is a good rate at present) |
Response bias can
be high. Certain groups less likely to respond. |
| Online survey |
|
Response is usually fairly
low (less than half of postal). Incentives can increase
response. |
Usually high –
from selection bias and response bias. |
|
