Summary of findings for changes in six categories of measurement error by four level-of-effort measures†
| . | Results for the following levels of recruitment effort paradata: . | |||
|---|---|---|---|---|
| Follow-up attempts . | Refusal conversion . | Date of interview . | Combination . | |
| Question-specific item non-response | Strong evidence for higher item non-response rates (16 studies) | Strong evidence for higher item non-response rates (7 studies) | No clear difference (2 studies) | Weak evidence for higher item non-response rates (2 studies) |
| Aggregate item non-response | Weak evidence for higher item non-response rates (15 studies) | Strong evidence for higher item non-response rates (11 studies) | No clear difference (2 studies) | No clear difference (2 studies) |
| Accuracy | Weak evidence for less accurate responses (6 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Scale reliability | No clear difference (2 studies) | No available evidence (0 studies) | No available evidence (0 studies) | No available evidence (0 studies) |
| Variability in answers | Clear evidence of difference, but not on direction (4 studies) | No clear difference (2 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Measurement error—attitudinal questions | No clear difference (5 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| . | Results for the following levels of recruitment effort paradata: . | |||
|---|---|---|---|---|
| Follow-up attempts . | Refusal conversion . | Date of interview . | Combination . | |
| Question-specific item non-response | Strong evidence for higher item non-response rates (16 studies) | Strong evidence for higher item non-response rates (7 studies) | No clear difference (2 studies) | Weak evidence for higher item non-response rates (2 studies) |
| Aggregate item non-response | Weak evidence for higher item non-response rates (15 studies) | Strong evidence for higher item non-response rates (11 studies) | No clear difference (2 studies) | No clear difference (2 studies) |
| Accuracy | Weak evidence for less accurate responses (6 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Scale reliability | No clear difference (2 studies) | No available evidence (0 studies) | No available evidence (0 studies) | No available evidence (0 studies) |
| Variability in answers | Clear evidence of difference, but not on direction (4 studies) | No clear difference (2 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Measurement error—attitudinal questions | No clear difference (5 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
‘No available evidence’ indicates that, at the time of writing, no empirical studies have examined this combination of level of effort and the measurement error indicator. ‘No clear difference’ indicates that there are empirical studies, but there is no clear evidence of a difference between high and low effort respondents.
Summary of findings for changes in six categories of measurement error by four level-of-effort measures†
| . | Results for the following levels of recruitment effort paradata: . | |||
|---|---|---|---|---|
| Follow-up attempts . | Refusal conversion . | Date of interview . | Combination . | |
| Question-specific item non-response | Strong evidence for higher item non-response rates (16 studies) | Strong evidence for higher item non-response rates (7 studies) | No clear difference (2 studies) | Weak evidence for higher item non-response rates (2 studies) |
| Aggregate item non-response | Weak evidence for higher item non-response rates (15 studies) | Strong evidence for higher item non-response rates (11 studies) | No clear difference (2 studies) | No clear difference (2 studies) |
| Accuracy | Weak evidence for less accurate responses (6 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Scale reliability | No clear difference (2 studies) | No available evidence (0 studies) | No available evidence (0 studies) | No available evidence (0 studies) |
| Variability in answers | Clear evidence of difference, but not on direction (4 studies) | No clear difference (2 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Measurement error—attitudinal questions | No clear difference (5 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| . | Results for the following levels of recruitment effort paradata: . | |||
|---|---|---|---|---|
| Follow-up attempts . | Refusal conversion . | Date of interview . | Combination . | |
| Question-specific item non-response | Strong evidence for higher item non-response rates (16 studies) | Strong evidence for higher item non-response rates (7 studies) | No clear difference (2 studies) | Weak evidence for higher item non-response rates (2 studies) |
| Aggregate item non-response | Weak evidence for higher item non-response rates (15 studies) | Strong evidence for higher item non-response rates (11 studies) | No clear difference (2 studies) | No clear difference (2 studies) |
| Accuracy | Weak evidence for less accurate responses (6 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Scale reliability | No clear difference (2 studies) | No available evidence (0 studies) | No available evidence (0 studies) | No available evidence (0 studies) |
| Variability in answers | Clear evidence of difference, but not on direction (4 studies) | No clear difference (2 studies) | No clear difference (1 study) | No available evidence (0 studies) |
| Measurement error—attitudinal questions | No clear difference (5 studies) | No clear difference (3 studies) | No clear difference (1 study) | No available evidence (0 studies) |
‘No available evidence’ indicates that, at the time of writing, no empirical studies have examined this combination of level of effort and the measurement error indicator. ‘No clear difference’ indicates that there are empirical studies, but there is no clear evidence of a difference between high and low effort respondents.
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
