Data about membership can come from one of two sources: the church itself or from individuals in the population as queried in surveys. Both types of data come in a variety of forms, each with their advantages and limitations.
Church-reported data include (1) official statistics released with annual reports, (2) dependent studies by outside groups which rely upon data obtained from official church sources rather than independent research, and (3) quasi-official data consisting of statements of church leaders or church scholars with access to internal church data (such as the "Vital Statistics section" in Encyclopedia of Mormonism).
Official Church Statistics
The Church reports certain statistical indicators for the prior year in the opening General Conference session every April. Relevant statistics include total membership, convert baptisms, increase of children of record, number of wards and branches (congregations), number of stakes and districts (mid-level administrative divisions, similar to a Catholic diocese), and the number of full-time missionaries serving. These statistics can be found each year in the May conference edition of the Ensign. The Church also publishes these data itemized by country annually in the Deseret News LDS Church Almanac, although the information in the Almanac is nearly two years old by the time it is released.
These are the only data published by the Church regarding its growth or membership. Unfortunately, none of these data except for the number of congregations and the number of full-time missionaries give any indication of member activity or convert retention. Yet both of these measures are exceedingly insensitive, as the size of congregations is primarily based upon administrative considerations and varies from branches attended by ten members up to Utah wards with several hundred, and full-time missionary service involves less than 0.4% of the LDS population in any given year.
Before 1984, the Church released a wider range of data regarding membership, including the number of deacons, teachers, priests, elders, and high priests, the number of living endowments performed, and demographic information, including birth and death rates. Although these statistics did not include data on church attendance, they offered insight into measures more specifically associated with member activity. After 1984, this ancillary data was dropped from annual reports.
Because official LDS membership statistics have no obligatory relationship to member activity or participation, official membership data offer relatively little insight into the growth and strength of the Church. A 10% increase in reported membership in a nation does not imply that 10% more people are attending LDS church services.