Table 3.
Numbers and Proportions Preferring Verbal or Numeric Probabilities
| Study * | Sample size | n (%) | |||
|---|---|---|---|---|---|
| Preferred verbal | Preferred numeric | Preferred combination | No preference | ||
| Woloshin et al. 198425 | 307 | 91 (29.6) | 135 (43.9) | 81 (26.3) | NA |
| Shaw and Dear 199023 | 81 | 43 (53.1) | 30 (37.0) | NA | 8 (9.9) |
| Freeman and Bass 19926 | 208 | 89 (42.7) | 119 (57.2) | NA | NA |
| Hallowell et al. 1997 33 | 43 | 3 (7) | 9 (21) | 22 (52) | 8 (19) |
| Franic and Pathak 200026 | 74 | 4 (5.4) | 70 (94.6) | NA | NA |
| Lobb et al. 2003 41 | 109 (unaffected by condition) | 24 (22.1) | 55 (50) | 20 (18.3) | 10 (9.6) |
| 84 (affected by condition) | 15 (17.9) | 16 (19.2) | 45 (53.8) | 8 (9) | |
| Graham et al. 2009 49 | 262 | 136 (52) | 125 (47.7) | NA | 1 (0.3) |
| Nagle et al. 200938 | 294 | 85 (28.9) | 132 (44.9) | 76 (25.8) | NA |
| Cheung et al. 201039 | 240 | 60 (25.0) | 180 (75.0) | NA | NA |
| Vahabi 201040 | 180 | 61 (33.9) | 119 (66.1) | NA | NA |
| Carey et al. 2018 57 | 210 | 59 (28) | 33 (16) | 79 (38) | 39 (18) |
NA indicates that this option was not presented to respondents
*Kaplowitz et al. (2002)36 also evaluated format preference among cancer patients and survivors in a hypothetical choice between a verbal probability and a quantitative estimate of survival. However, the findings are not integrated into this table because the options were not mutually exclusive, and the authors do not clarify how many patients chose both. (Table 1 of that paper shows that 80% endorsed verbal probabilities and 53% endorsed quantitative information, suggesting that some subset must have chosen both)