That's correct. I'm editing it to say "1 of every 25 users" to make this less confusing. The actual sample size was quite large.