Texting rates and percentage of texting estimations in the lower 31.6%* of each order
| Daily | Weekly | Monthly | ||
|---|---|---|---|---|
| Estimated number of texts sent over different periods† | ||||
| Total N | 248 | 240 | 247 | |
| n (%) who provided a range | 66 (27) | 51 (21) | 55 (22) | |
| Mean of estimated texts sent | 37.04 | 146.90 | 643.44‡ | |
| Percentage of estimated and billed weekly texts in lower 31.6%* of each order of magnitude | ||||
| n | Estimated (%) | n | Billed (%) | |
| First order (0–9) | 40 | 50 | 18 | 72 |
| Second order (10–99) | 71 | 48 | 55 | 33 |
| Third order (100–999) | 74 | 58 | 75 | 64 |
The second order of magnitude is most relevant as there are no outside influences on the distribution. It is not clear whether first-order values are estimated on a mental linear or logarithmic scale, and the third order is influenced by the group who had only 500 texts available monthly: their weekly estimates will fall in the lower half of the order, and are more likely to be less than about 150.
*This represents the half-way point on a logarithmic scale.
†Includes data given as a range with the geometric mean applied.
‡617.60 with top outlier excluded.