Skillbuilding: Data Analysis
Stefani Evans, CG, “Data Analysis,” OnBoard 18 (May 2012): 13--14.
Standard 20 of The BCG Genealogical Standards Manual describes the separate but interrelated processes of analyzing items of information and correlating them.1 The Genealogical Proof Standard explicitly calls for genealogists to analyze and correlate data to weigh its quality as evidence.2
Throughout the research process we consciously and unconsciously formulate and discard hypotheses that shape our searches. Each hypothesis is subjected to two processes. The first “to draw out its logical or empirical consequences,” calls for analyzing data within a source and applies at all stages of research.3 The second process, to “test [the hypothesis in] accord with facts that are known or may be determined,” occurs later, when we correlate data across sources.4 Analyzing data within a source is this article’s focus.
The preliminary survey of published and family literature is a starting point—a first step—alerting the researcher to collections and scholarship on place and family that relate to the research question. Sources found at this early stage establish a base knowledge and guide research.5
An example involves a source found during the preliminary survey that questions when Peter DeGarmo—who died in 1840 in Dutchess County, New York—was born.6 The source, a 1957 typescript by DeGarmo descendant Leona Paxton, includes a family record allegedly based on the Bible of Peter DeGarmo and Mary Robinson of Dutchess County.7 Paxton’s DeGarmo family record offers a knowledge base and suggests a research path pertaining to the research question.
Useful derivative sources found through preliminary surveys should be analyzed as thoroughly as original sources found later through “reasonably exhaustive searches.” The acts of citing and transcribing even legible derivative sources enable detection of nuances that might otherwise be overlooked. Citing and transcribing apply the first steps of what the BCG calls “Document Work,” and establish what we “know” from a document.8 This knowledge—correct or not—suggests what we want to know or need to learn in order to answer the research question. A research plan based on what we want to know identifies repositories and sources that might provide the wanted information. These steps allow us to form hypotheses for the purpose of testing.
Paxton’s family record, modified in format only, follows.
“PETER DEGARMO, who was born in January, 1751[,] and his wife, Mary (Robinson) DeGarmo, who was born on November 11, 1763[,] lived in Poughkeepsie, Dutchess County, New York State. Their family:
1. John, born April 6, 1783
2. Rowland, born November 29, 1785 Died June 6, --
3. Elias, born March 2, 1786
4. David, born April 30, 1791 Died in 1845
5. Martha, born June 2, 1793
6. Mary Louise, born November 17, 1795 Died May 24, 1803
7. Peter, born March 4, 1797 Died Aug. 2, 1852
8. Anna (Smith), born March 10, 1800 Died Aug. 2, 1852
9. Henry, born April 22, 1803
10. Elizabeth (Sutton) born November 3, 1805 Died Dec. 4, 1884
Although this record may be incorrect in several instances, it nonetheless proposes a useful guide to Dutchess County research. The data also offers a testable hypothesis relative to the research question: Peter DeGarmo was born January 1751. After developing a research plan and hypothesis for testing, we analyze the document “to draw out its logical or empirical consequences.”9
Creating the citation forces examination of provenance. The source’s form, an interpretation of 200-year-old data from an alleged Bible, is derivative. A derivative source may contain primary information, but when Paxton created this derivative source she erased the clues—publication date, entry scripts and inks, and formatting—that would allow the reader to identify the informant for the 1751 data. Thus, the reader cannot identify whether Bible data relative to DeGarmo’s birth is primary or secondary.10
Despite its flaws as a source, Paxton’s family record is relevant because it offers direct evidence that answers the research question, and it also contains data that may become indirect evidence applicable to the question.
DeGarmo’s hypothetical January 1751 birth is the only data in Paxton’s family record that must be tested in light of England’s adoption of the Gregorian Calendar in 1752. The calendar change moved the beginning of the year to 1 January from 25 March, and it deleted eleven days from September 1752. The calendar change affected Britain and all its colonies, but Paxton does not indicate whether the 1751 date is Old Style or if she adjusted the entry to conform to New Style.
Two facts support Paxton’s assertion that she referenced the DeGarmo Bible. First, Paxton omitted place references as do most family Bible records. Second, deaths other than Elizabeth’s were in or before 1852, which may represent the last correspondence before a migration that finally broke communication links with Dutchess County.
It is implausible that two adult children of Peter DeGarmo died the same date. Ironically, this improbability lends authenticity, suggesting that Paxton copied from original handwritten entries squeezed onto a page, likely between two children. To avoid assigning the death to the wrong child Paxton possibly chose to apply it to both.
If these are natural children of Peter DeGarmo and Mary Robinson, it is biologically impossible that Rowland was born 27 November 1785 and Elias 2 March 1786. Thus, either Paxton mistyped or referenced an erroneous record. Nonetheless, Paxton evidenced great care. She substituted a hyphen for Rowland’s death year, signifying she did not overlook the year; rather, she did not have it.
The BCG’s call for analysis in Standard 20 and the Genealogical Proof Standard support the first part of the hypothesis test, “to draw out its logical or empirical consequences.”11 Analyzing information items within a source allows a researcher to weigh the quality of the source and the credibility and applicability of the information it contains at all stages of the research process.
Endnotes
The author thanks Donn Devine, CG, CGL, and Thomas W. Jones, PhD, CG, CGL, FASG, FNGS, FUGA, for editorial comments and Laura DeGrazia, CG, CGL, Donn Devine, Kay Freilich, CG, CGL, Alison Hare, CG, and Thomas W. Jones for discussions that shaped this article.
1 Board for Certification of Genealogists, The BCG Genealogical Standards Manual (Orem, Utah: Ancestry, 2000), 8, for Standard 20.
2 Ibid., 1, for bullet (c), Genealogical Proof Standard.
3 Quote from definition of a hypothesis in Webster’s Third New International Dictionary, Unabridged (n.p.: Merriam-Webster, 2002), “hypothesis”; database, Merriam-Webster (http://unabridged.merriam-webster.com : accessed 17 March 2012).
4 Ibid.
5 Val D. Greenwood, Researcher’s Guide to American Genealogy (Baltimore: Genealogical Publishing Co., 1983), 2, 38.
6 Dutchess County, New York, probate files, no. 3697, Peter DeGarmo (1840), petition for probate, 17 August 1840; Surrogate Court Clerk’s Office, Poughkeepsie.
7 Leona Paxton, “The Henry C. Sutton Family: Henry C. Sutton (1804-1901), Elizabeth DeGarmo Sutton (1805-1884),” unpaginated typescript, c1957; photocopy of carbon copy January 2000 to author from Harriet and Doris Sutton, Lincoln, Nebraska. Elizabeth DeGarmo, youngest child of Peter DeGarmo and Mary Robinson, allegedly carried the Bible of her parents when she migrated from New York to Michigan with her husband, Henry C. Sutton. Paxton, a granddaughter of Henry C. Sutton and Elizabeth DeGarmo through their youngest child, Isabelle, sent a copy of her work in the early 1960s to Harriet and Doris Sutton, unmarried great-granddaughters of Henry C. Sutton and Elizabeth DeGarmo through their son Smith.
8 Board for Certification of Genealogists, The BCG Application Guide (Washington, D.C.: Board for Certification of Genealogists, 2011), 4--5, for item 3, Document Work.
9 Webster’s Third New International Dictionary, “hypothesis.”
10 Christine Rose, Genealogical Proof Standard: Building a Solid Case, 3rd Edition Revised (San Jose, California: CR Publications, 2009), 8.
11 Board for Certification of Genealogists, The BCG Genealogical Standards Manual, 1, 8. Also, Webster’s Third New International Dictionary, “hypothesis.”
Stefani Evans, CG
This article was originally published in OnBoard, BCG’s educational newsletter and is protected by copyright. Individuals may download and print copies for their personal study. Educators are granted permission to provide copies to their students as long as BCG, OnBoard, and the appropriate author are credited as the source of the material. Republication elsewhere is not permitted.