Chapter Text
N.B. The above percentages refer to % of total responses to this question, and sum to 100%. Percentages shown below refer to the % of total survey participants unless otherwise specified.
Question
This question was phrased as “How long have you been using AO3?” Only one response could be selected.
Responses
The responses were selected with the following frequencies:
- Less than 6 months - 727 respondents. 7.3%
- 6 months - 1 year - 2298 respondents. 23.0%
- 1 - 2 years - 3474 respondents. 34.7%
- 2 - 3 years - 1898 respondents. 19.0%
- 3 - 4 years - 899 respondents. 9.0%
- 4 - 5 years - 404 respondents. 4.0%
- More than 5 years - 264 respondents. 2.6%
41 respondents (0.4%) did not respond to this question.
Taking the values for each range to be at the central point of the group, with the outermost values being 3 months and 5 years respectively, the mean length of use was found to be 1.82 years (22 months) with a standard deviation of 1.18 years (14 months). The median length of use was in the range 1 - 2 years.
Reliability
Archive of Our Own’s open beta began on 14 November 2009, a little less than 4 years before this survey took place. However, some respondents may have been involved in the closed beta which began on 3 October 2008, just under 5 years before this survey. This was probably the earliest point in time that AO3 was available for use. (ETA: Some people have contacted me saying that developers and OTW members were using the site before this point, but that this group was definitely smaller than 264 people.)
The survey’s original closing date was the end of October, which would have allowed the very earliest users to have been using the site for more than 5 years. However, the early close of the survey means that the 264 people who entered this category were misremembering their length of time on the site. It is also possible that some of the earlier users of the site selected the longest length of time without checking its accuracy.
This question, and all others in this section, are subject to recall bias of this kind, as the responses given may not accurately reflect reality.
