Tech Forum

I never understood why, when submitting the quarterly report, do I only report the last month of the quarter?

i.e. when reporting sacrament meeting attendance, why am I only sending the last month of the quarter? Why not average all 3 months?

Same goes for all other statistics reported (HT/VT)... it doesn't seem to make sense to me. If anything, one could be tempted to only collect data for the last month of each quarter

billybob00 wrote:I never understood why, when submitting the quarterly report, do I only report the last month of the quarter?

My understanding is that it is a snapshot, and that is where we have been asked to point the camera.

Other than that, I know not. I suppose this one is like Moses 5:6.

Seems to me that one month each quarter would catch any trends of concerns, while simplifying the effort needed to create the reports. Statistically, any trend would be more pronounced then if it had been averaged over the whole quarter.

RussellHltn wrote:Seems to me that one month each quarter would catch any trends of concerns, while simplifying the effort needed to create the reports. Statistically, any trend would be more pronounced then if it had been averaged over the whole quarter.

Huh? I'll agree that the method is simpler and cheaper. But I can't see that it is statistically advantageous at all.

To the contrary, just sampling a single month and calling it a "trend" that is "more pronounced" can obviously be misleading. For example, the Sunday that Sacrament Meeting attendance is counted might happen to be the day of a big snowstorm. That's why economists, scientists and statisticians generally prefer more datapoints, not fewer -- to smooth out the data and eliminate undue influence by outliers.

And it's the snowstorm effect that has me wondering why, if we're reporting sample months anyway, why we don't report median sacrament meeting attendance for the month in question instead of mean.

JTaber wrote:And it's the snowstorm effect that has me wondering why, if we're reporting sample months anyway, why we don't report median sacrament meeting attendance for the month in question instead of mean.

Well, I don't know much about statistics, mean, median, or mellow. I just know that everyone around here is gone during the summer and the June numbers are always an anomaly. Minimally statistically speaking, of course.

boomerbubba wrote:For example, the Sunday that Sacrament Meeting attendance is counted might happen to be the day of a big snowstorm.

It's been a problem for student wards as well. Many of them are out for summer during June and not all are in full session at the first Sunday of September.

While quarterly report data may indeed provide a snapshot in time, I feel it is too limited for ward or organization analysis.

boomerbubba wrote:For example, the Sunday that Sacrament Meeting attendance is counted might happen to be the day of a big snowstorm.

Exactly correct, and proves my point. We had two very snowy sundays in December. We have many older members in the ward who just don't come out in the snow. Our sacrament numbers took a noticeable dip on the year-end quarterlies from the the quarter before.

Why, the bishop asked? Was it the snow that dropped the December numbers? What about the fact that in September (again, a counting month) we had several baby blessings and the primary program? Did that "skew" the 3rd quarter numbers to the high side?

Probably September's numbers were abnormally high and December's numbers were lower then usual, giving the impression of a noticeable downward trend. So, the "snapshot" of two months doesn't reflect the 6 months it represents.

Only by tracking attendance every week while documenting occuring events would you see the dips and spikes that occur weekly. You could attribute some of these anomalies to events, like holidays, snow storms, baby blessings, etc. As you graphed a whole year, you would see the true average for your attendance. If you did this every week for years, I assume you should see noticeable seasonal patterns as well.

Your bishop would love you.

As for organizations, some of ours stopped taking role on the months it's not required, and then noticed a definite drop in the leaders knowledge of the attendance habits of their members. I think most have returned to weekly attendance roles.

Just to give a qualifier for my comments, I have an MS in statistics and make a career out of things like this.

The quarterly report offers a systematic sample of church activity and member progress. Systematic samples have fairly good statistical qualities. Certainly, they aren't as good as random samples, but when applied over cycles (such as years), they do have the potential to show cyclical trends, something that random sample have the potential to miss. Statistically speaking, a better sample would be obtained by randomly selecting a month from each quarter to report, but that introduces obvious organizational headaches that probably aren't worth the improved sample.

For the most part, however, the numbers on the quarterly report are worthless at the ward level. The numbers are too subject to random variation to provide much of a picture of what is going on and to develop a true understanding of trends and changes and what might be causing them. Personally, I doubt the actual numbers have much value even at the stake level (except to provide a starting point from which to guide leadership interviews). By the time an Area aggregates the numbers, they might be able to see trends and investigate their causes.

What's of more value to the wards is the lists that should be reported every month. Recall that the handbooks instruct each quorum and auxiliary to report, in addition to the numbers, lists of members who did not attend their meetings that month. And they're supposed to report this every month. In fact, to encourage this, I've stopped collecting numbers in my ward. I have the secretaries submit a list of those who haven't attended and I use that list to get my numbers for the quarterly report.

With regard to the sacrament meeting attendance, I can see arguments for reporting the average of all three months and for reporting the average of just the reporting month. Both have their advantages and disadvantages. While I don't have the authority to say anything definitive, my suggestion is that if you have huge variations (either increases or decreases) due to family events or weather, you should average your entire quarter. To me, that number should reflect what you typically see at Sacrament as that is the number that will be most useful in helping you plan other ward activities. I would love to see the Church adopt the full quarter average.

For the other quarters, as long as you're collecting the names that go along with the numbers, just the last month of each quarter is plenty sufficient.

nutterb wrote:Just to give a qualifier for my comments, I have an MS in statistics and make a career out of things like this.

thanks for your insights. We also use "members not attending" monthly reports both for organization and bishopric uses, and I find that If organizations give me a number of those not attending for the quarterly, they don't stress about about the numbers of their individuals that never match between their rolls and the MLS counts.