The ?Southwestern States Mission? series (homepage) examines mission life in (mostly) Texas around 1900.

[1] I have excluded Mission President Duffin, Sister Cluff, and Sister Carling as their city-based, semi-permanent-resident work seems to have followed different patterns than those of the travelling Elders.

[2] Below I describe how I arrived at this graph. The procedure below attempts to control for variability among missionaries? singing and recording habits. Each missionary both sang different amounts and used different criteria for deciding when to write about singing. Thus, if I just added up the total number of music events on a given day, the experiences of Elders who either sang more or wrote about it more would occlude those who did less. In order, I:

1. Made a list of the 375 instances when the five travelling Elders recorded making or hearing music. (The totals per Elder were: Brooks, 35; Clark, 114; Folkman, 76; Forsha, 44; Jones, 106. I excluded the few months that Duffin was a travelling Elder. In a few cases there were more than one instance per journal entry or an event was recorded on a different day than it happened; each instance was counted separately and assigned to the appropriate weekday.)
2. Tabulated the instances per day for each Elder and then,
3. Normalized the totals for each Elder. That is, for a given Elder, I divided the count for each day by the largest count for that Elder. For example, Elder Jones?s 106 events were divided thus: Sun 22, Mon 16, Tue 14, Wed 16, Thu 14, Fri 15, Sat 9. The largest count for Jones was Sunday, 22; I divided each day?s count by 22, yielding normalized values: Sun 1.00, Mon 0.73, Tue 0.64, Wed 0.73, Thu 0.64, Fri 0.68, Sat 0.41.
4. Averaged the normalized values for each day from each respective Elder.
5. Converted the average, normalized values into percentages of the total amount of singing done per week, which are shown in the graph.

[3] ?3wk RA? is the three-week running average. The procedure for developing the ?Music Events per Week, Normalized? graph was similar to ?Percentage of Music Events per Weekday? graph, with a few exceptions. Besides controlling for variability among Elders in amount of singing and of writing about singing, this procedure attempts to uncover seasonal patterns. Instead of looking at how often missionaries sang in the summer of one year, I am looking for patterns that endure despite variability between years. The procedure:

1. Made a list of the 375 instances when the five travelling Elders recorded making or hearing music.
2. Tabulated the instances per week for each Elder using Microsoft Excel?s ?WEEKNUM? function.
3. Averaged the instances per week for each Elder. For example, in Week 51 (mid-December), Elder Folkman had 1 musical event in 1899, 0 in 1900, and 4 in 1901. Elder Folkman?s average for Week 51 is, therefore, 1.67 events.
4. Normalized the average per week for each Elder. Since the highest average number of events Elder Folkman had was 2.0, the normalized value for his Week 51 becomes 0.83 (ie, 1.67/2.0).
5. Averaged the normalized values for each week from each respective Elder, which, along with a three-week running average, is shown in the graph. Note that the running average ?wraps around?: Weeks 52, 53, and 1 are averaged together, as are Weeks 53, 1, and 2.