Beyond data derived from church sources, the second major type of data comes from direct study of the population through national censuses and independent sociologic surveys. In the absence of meaningful official church statistics regarding member participation and activity, data derived directly from the population is necessary to achieve meaningful indicators of the Church's real growth and progress.
Original sociologic surveys allow a wide range of questions to be asked, including questions about religious belief, church attendance, and other faith-based behaviors. Such surveys can thus offer a broad range of information of great interest and value. Independent surveys, including the Pew American Religious Landscape Study, the CUNY American Religious Identification Survey, and others, collect and report original field data, and thus have far more to offer students of church growth than dependent research which merely analyzes and compares church-reported data. Because each original survey has a specifically defined methodology which is applied to all respondents, regardless of their religious affiliation, such research can offer more valid comparisons between adherents of different religious groups than studies like Glenmary which attempt to compare heterogeneous inter-denominational data without accounting for the wildly disparate membership definitions and reporting methodologies of different faiths.
The greatest drawback of independent surveys is their methodological limitations. With the limited resources available to religious researchers, it is not possible to query the entire population as is done in a national census, and so studies are done based on random sampling. It is particularly difficult to get an accurate picture of LDS membership through random sampling for several reasons. LDS membership concentration demonstrates profound geographic variation, from approximately 75% of the Utah population to less than 1% in the eastern United States and as little as 0.1% in many European countries. The characteristics of Utah, U.S., and international Latter-day Saints are different in salient ways which forbid the extrapolation of data from one group to another, yet the concentration of LDS members outside of the Mountain West is so low that even fairly large studies have relatively wide margins of error for the population of Latter-day Saints among the respondents. Finally, such surveys have often been conducted by researchers with little knowledge of the membership reporting definitions and retention challenges of the LDS Church, leading to errors in analysis and data interpretation when research findings are extrapolated to official church membership statistics.
No survey method is perfect, and some limitations are innate even to well-designed studies. Independent sociological surveys have added a great deal to our knowledge of the LDS Church and its members, although care is needed.