steve2: Special Pioneers... Can anyone verify that?
(Not dates of changes, just WT 'mentions')
Watchtower February 15, 1957
Page 115: It is a grand service wherein regular pioneers spend at least 100 hours a month preaching and special pioneers a minimum of 150 hours.
Yearbook 1977
Page 22: Special pioneer hours were reduced to 140 per month to help them to continue in the full-time service, to give them time to do shepherding work in the congregations and also to find some means of income, if this is necessary in their case
Who Are Doing Jehovah’s Will Today? (Brochure) 2012
Lesson 13 What Is a Pioneer?: Others are selected to serve as special pioneers in areas where the need for Kingdom proclaimers is greater, devoting 130 hours or more to the ministry each month.
steve2: I acknowledge your need to draw this topic out darkspilver
You asked for my comments. I appreciate you taking the time to read and comment on them.
steve2: You also say, "I believe it is therefore a statistical fallacy to use 'peak publishers' - it is much better to use 'average publishers'... could you tell me exactly which statistical fallacy you are referring to when peak publishers are used?
For the Annual Report (covering 1 September to 31 August), the WT is counting, not individual people, but rather the Report forms they submit each month, from the same group of people (ie not random people), and there is a very small increase/decrease over the year.
The Annual Report, as printed in the Yearbook, is distilled from the Monthly Reports submitted by publishers - but remember.....
- When publishers forget to submit a Report, they give in a 'late' Report the next month and they get double or triple-counted (or more!);
- The 'peak' month is not a random month, it is pre-determined as being the end of the 'Service Year' in August, and it is, by far, significantly higher than ALL the other months, before and after;
- The only reason it is higher is because of the 'late' reporting for the 'Service Year' just finishing, and the extra effort put in to collect those late reports (see '1' above).
I believe it is therefore statistically dishonest for the WT to use the 'Peak Publisher' number in the way they do as, in many cases, it gives a knowingly higher figure than what the reality is.
I also think that even comparing like-for-like, peak-with-peak, year-to-year, is not good statistical practice.
If you want a high 'peak' number, you encourage 'late' reporting (!), which would appear to be working against statistical ethics.
The 13-year stretch in my OP started with the 2003 Service Year - therefore the examples below are of the number of publisher's reports submitted each month for the first two years: 2003 and 2004.
NOTE: Admittedly, publishers in some countries are better at Reporting on-time: In 2003, Austria, with annual growth of 0%, reported an average of 20,148, and a peak of 20,331 - but then this causes a statistical problem when comparing 'peak publishers' country-by-country.
