Early baptismal registers represent one of the most substantial surviving records of the PMC's first half-century, and their principal use to date has been as a gauge of paternal employments. In an early sample analysis of them, Gilbert concluded that while there was a gradation of prosperity from Congregationalism to the PMC, the pattern was one of essential commonality, in that all dissenting denominations had a majority following among the poorer fractions of society; by contrast, some comparative local studies have claimed a more pronounced prosperity gap between the PMC and others, and the fact of local variability is amply confirmed in Watts’ analysis of 1995. Yet the methodology rests upon some questionable if sometimes implicit assumptions: that data extraction can be done consistently in a non-skewing manner; that baptising fathers are a reliable cross-section of the membership; and that paternal employment is a sufficient and reliable analogue for the family's socio-economic circumstances. At each stage, researcher-induced differences occur.

The principal problems of data extraction relate to poor orthography and the treatment of repeat presenters. The former may be more of an issue in early PM registers than elsewhere, because the sacrament was conducted by so many people. For example, the 601 baptisms recorded in the PM circuit register for Tunstall in North Staffordshire for the period 1820 to 1837 were conducted by ninety-three officials, whose literacy and orthography were uncertain. At least eighteen parent sets appeared on two or more occasions with materially different surname spellings, and three of these (Clews-Clowes, Lea-Lear and Lees-Leese) indicate that there was the added issue of regionally distinctive pronunciations. These cases were found by matching not only the paternal Christian and surnames but also spouse Christian and maiden names, plus male employment, of original entries often many pages apart. In a pre-Excel world, or where the researcher did not use a spreadsheet approach, many would have gone undetected. One particularly testing example may serve as an indication of the impact, too, of the researcher's judgement. Potters named John Alcock or Allcock, married to women named either Eileen or Ellin Lees, or Ellen Leigh, each offered a child for baptism between 1824 and 1829. Even those researchers who discerned the potential commonalities here might reasonably conclude that these were three, or only two, couples.