Table 2.
Characteristics of warning manipulations in studies in the meta-analysis
| Variable | Pictorial (k=37) | Text (k=37) | ||
|---|---|---|---|---|
| k | Per cent | k | Per cent | |
| Number of different warnings viewed | ||||
| 1 warning | 14 | 38 | 15 | 41 |
| 2–64 warnings | 22 | 59 | 20 | 54 |
| Not reported | 1 | 3 | 2 | 5 |
| Number of times viewed each warning | ||||
| 1 time | 32 | 86 | 32 | 86 |
| 2–5 times | 5 | 14 | 5 | 14 |
| Number of exposure sessions | ||||
| 1 session | 36 | 97 | 36 | 97 |
| 2–4 sessions | 1 | 3 | 1 | 3 |
| Days from exposure to assessment | ||||
| 0 days (immediate assessment) | 36 | 97 | 36 | 97 |
| 1–28 days | 1 | 3 | 1 | 3 |
| Exposure medium | ||||
| Just warning | 4 | 11 | 6 | 16 |
| Warning on a 2D pack | 21 | 57 | 20 | 54 |
| Warning on a 3D pack | 8 | 22 | 8 | 22 |
| Not reported | 4 | 11 | 3 | 8 |
| Exposure channel | ||||
| Digital | 21 | 57 | 21 | 57 |
| Printed or paper | 4 | 11 | 4 | 11 |
| Cigarette pack | 8 | 22 | 8 | 22 |
| Not reported | 4 | 11 | 4 | 11 |
| Label order | ||||
| Random | 10 | 27 | 9 | 24 |
| Non-random | 5 | 14 | 5 | 14 |
| Not reported | 6 | 16 | 5 | 14 |
| NA (1 label or all shown at once) | 16 | 43 | 18 | 49 |
| Warning exposure controlled by… | ||||
| Researcher | 9 | 24 | 9 | 24 |
| Participant | 21 | 57 | 21 | 57 |
| Both | 1 | 3 | 1 | 3 |
| Not reported | 6 | 16 | 6 | 16 |
| Nature of pictorial warnings | ||||
| Image only | 3 | 8 | – | – |
| Image with text | 33 | 89 | – | – |
| Not reported | 1 | 3 | – | – |
| Pictorial text vs comparison text | ||||
| Matched completely | 16 | 43 | – | – |
| Did not match | 16 | 43 | – | – |
| NA (pictorial condition had no text) | 3 | 8 | – | – |
| Not reported | 2 | 6 | – | – |
All but a single study69 assessed individuals only directly after exposure.
2D, 2-dimensional; 3D, 3-dimensional; k, number of effect sizes; NA, not applicable.