Abstract
We investigate whether survey answers of trained
respondents
differ systematically from answers of novice respondents.
Focusing on difficult attitudinal or preference questions, we find that
novice respondents answer “don’t know” significantly more often.
Importantly, however, there is no systematic evidence for a conditioning
effect on measured preferences (substantial answers). Overall, our
results suggest that panel conditioning does not pose a serious problem,
even in the case of rather difficult questions.
respondents
differ systematically from answers of novice respondents.
Focusing on difficult attitudinal or preference questions, we find that
novice respondents answer “don’t know” significantly more often.
Importantly, however, there is no systematic evidence for a conditioning
effect on measured preferences (substantial answers). Overall, our
results suggest that panel conditioning does not pose a serious problem,
even in the case of rather difficult questions.
| Original language | English |
|---|---|
| Pages (from-to) | 783-797 |
| Journal | Public Opinion Quarterly |
| Volume | 77 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 2013 |
Keywords
- panel conditioning