TABLE 1.
Participant responses to a range of questions sub-divided by their initial response to altering the witness’s accuracy.
| Reduce | No change | Increase | |
| Total N | 28 (21.4%) | 73 (55.7%) | 30 (22.9%) |
| Cab/witness statistics provided | |||
| (Mean) Cab blue estimate | 9.5% (0.3%) | 16.8% (2.0%) | 18.2% (3.6%) |
| (Mean) Witness accuracy estimate | 76.0% (1.9%) | 73.3% (2.1%) | 71.6% (2.6%) |
| (Mean) Confidence | 67.1% (2.9%) | 68.8% (2.7%) | 69.9% (3.3%) |
| Witness reports cab is blue | |||
| (Proportion) Increasing estimate cab blue | 67.9% (9.0%) | 56.2% (5.8%) | 83.3% (6.9%) |
| (Proportion self-reported assuming): | |||
| Witness correct | 25.0% (8.3%) | 60.3% (5.8%) | 80.0% (7.4%) |
| Cab green | 46.4% (9.6%) | 15.1% (4.2%) | 3.3% (3.3%) |
| Neither/other | 28.6% (8.7%) | 24.7% (5.1%) | 16.7% (6.9%) |
| Told cab actually green | |||
| (Proportion) Reducing witness accuracy | 78.6% (7.9%) | 52.1% (5.9%) | 73.3% (8.2%) |
In brackets on the left it is indicated whether the figure provides the mean response, or the proportion providing a particular response (in the case of non-continuous outcomes). Questions are in chronological order and the delivery of key pieces of information is indicated. Standard errors for estimates are included in brackets.