United States Census Analyzing Census Data
The elements of good census analysis include thoroughness, source citations, correlating and corroborating sources, testing for relevance, likelihood, and reasonableness, explaining contradictions, and follow up through sharing findings and your continuing education.
The more of these ideas you use, the better your analysis will be.
|“Census records are the most used—yet most under used of all genealogical resources.”|
Confidence that a reasonably thorough search has been made is a cornerstone of good census analysis. Use all the censuses and use them well. And don't forget the value of comparing censuses to other sources.
Find and photocopy every family member in every census. Find every member of an ancestor's family in every available census during each of their lifetimes. Use federal, colonial, territorial, state, and local census. Use all census schedules: population, mortality, veterans, slave, etc.
Continue to hunt hard-to-find ancestors. If some ancestors are difficult to find in a census, work even harder until you do. See Guessing a Name Variation for 20 ideas to help you locate hard-to-find ancestors. If you still cannot locate them in the census try a substitute jurisdiction, or substitute record types such as tax lists, plat maps, directories, or voting registers. It is better to assume they ARE in the census under some garbled name than to give up.
Add census events to the family group record. Use the "Add Custom Event" feature to add to the family group record the data from each census for each family member.
Study the family in community context. Proximity implies a relationship. Research each individual in the same census household with a different surname. Look several census pages before and after an ancestor for people with similar given names, family names, occupations, or places of origin to see if they could be relatives.
Compare with non-census sources. Find as many non-census sources as you can for each family member. Begin a preliminary evaluation by briefly comparing censuses with other documents, especially land records, to identify neighbors and relatives. Study all sources to identify family associates and discover their relationship.
Cite Your Sources
Good record keeping contributes to good analysis. Start with a well documented family group record even before you look at a census. Keep organized and document as you go. Good source footnotes are another cornerstone of good analysis.Citing all your sources makes it easier to correlate and analyze sources against each other. It helps other researchers check your work. Add each census event to the family group record and make at least one source footnote for every census for every member of an ancestor's family.
Footnote other events listed on each census. Most censuses are full of information about non-census events in people's lives. Your analysis will be better if you know (footnote) each and every source that mentions an event. Even if such an event already has a source citation from another document, consider footnoting the following events on your family group record with a census citation (add custom events where needed):
- Born – from age, or birthplace, or parents' birthplaces information on the census
- Married – from (a) wife, (b) if married with the year, or (c) marital status
- Died – from (a) widowed status [regarding spouse], or (b) mortality schedule
- Military Service – veterans, or widows on veterans schedules
- Land – real estate value, or mortgage questions
- Education – attended college question (can read and write may only show home schooling)
- Prison (new defined custom event) – convict question on the census
Footnote the other events found in all the non-census sources for each family member.
Use the comment field of each footnote for a preliminary evaluation of the source. A few words of comment about the source's reliability, or how the new data compares with data in other sources will improve the more formal analysis later.
Correlate, Corroborate, Analyze and Interpret
After you have conducted a thorough search of the sources, and source footnoted them, you are ready to correlate, analyze, and interpret the evidence. Thoughtful correlation and comparison of many sources is what analysis is all about.
Relevance. Determine if a census or other source is about your family by comparing the new data with what you already know. If most of the names, dates, places, relationships, and sources match well, and only a few mismatch, then the new source is about your family.
Census data is rarely completely consistent from year to year. But usually enough remains consistent to recognize a family as it changes over time. Between 1865 and 1875 many African American families changed their surnames. If you cannot find them under the surname you expect, try looking for their given names and disregard family names. Or, look for neighbors from other census years to find them on the 1870 census—then see if your ancestor is nearby.
Primary vs. secondary census information. A census schedule shows a mix of primary and secondary information. Residence on a census is primary information because the census taker stood at the door of the home asking questions. But most of the other data are secondary. Usually other more primary sources would better document names, relationship to head-of-house, age, marital status, education, place of birth, citizenship, occupation, or military service. But it is best to corroborate one piece of data with several independent sources that agree, even if the source's information is secondary census data.
Accuracy. Most of the time the census got it right, but . . . Ages on a census and the spelling of names are especially prone to error whether by mistake or deliberate. Old immigration dates on a census may have been remembered incorrectly. A member of a racial minority may have hesitated to reveal his or her race to a census taker. Tired census takers may have taken sloppy notes, or struggled to understand a thick foreign accent. Censuses are an important source but should be compared with all other censuses and other documents before relying too heavily on their information.
Corroboration. The best tools to bring to bear on analyzing censuses are the consistency of the data when compared with other censuses and other sources. Do independent sources created without reference to each other agree on the facts? Are there any contradictory sources? If so, can you explain the contradictions?
Likelihood of events. Logically, could the events described on the census have really happened in the order suggested when compared to other sources? Does census information reflect what really happened?
Milk censuses for what they imply. Think about the implications of what is on the census and what is not on the census. For example what does it mean when someone appears, disappears, or lives nearby on the census. Could a 14 year old girl with a different surname appearing in a household be a niece as well as a maid? Could the disappearance of a lady in her 60s mean she died, or went to live with one of her children? If an ancestor with the surname CRAIG lives next to someone with the surname GREG could they be related with different versions of the same name?
Explain Contradictory Evidence
Do not ignore contradictions. Embrace contradictory evidence in censuses or other sources and try to explain it. This honesty will strengthen the case you are trying to make.
Consider the trustworthiness of any document you find. But, when two sources disagree about an event, it is important to analyze which source is the most trustworthy. Was the person who provided the information in a position to know? Was the information recorded near the time of the event or years later? Is the information primary or was it compiled from other sources? Was it reliably copied? Was care taken in making the record and preserving it? Did the information provider or record keeper have any ulterior motives? Which version of events is most consistent with other facts?
Sometimes, it is best to admit you cannot explain every discrepancy, or tell which contradictory version is most trustworthy.
Check for Reasonable Conclusions
Be wary of far-fetched conclusions. Are your analysis and interpretation well-reasoned and are they reasonable?
Share. One of the final steps of good analysis is to share your findings with others. This allows them to make constructive suggestions and may lead to new evidence. Share and collaborate with others to improve your census analysis.
Continue your education. Another part of analysis follow up is continuing to learn about censuses, analysis, and genealogical research. Read about good research and analysis in books, periodicals, and the Internet. Take classes, go to conferences, and institutes to learn more. Trips to places where ancestors lived are also part of your genealogical education.
Using all the censuses available for all the family members, carefully citing censuses in source footnotes, corroborating censuses and other sources, checking the likelihood of events, milking them for all the possible implications, explaining contradictions, and drawing reasonable conclusions are all part of good census analysis. Share your census findings and analysis with others, and continue to grow through education to sharpen your analytical skills.
- Principles of Family History Research – Use the Information
- Genealogical Proof Standard
- Angela McComas, Heads of Household Only: Analysis of Pre-1850 Federal Census (19 minute online video) FamilySearch Research Classes Online and Mid-Continent Public Library, Midwest Genealogy Center, 2010.
- Tiffany Perkins, Take a Research Trip with Tiff! (25 minute online video) FamilySearch Research Classes Online and Mid-Continent Public Library, Midwest Genealogy Center, 2010.
- This and most of the other headings in this article are based on elements of the Genealogical Proof Standard.
- Elizabeth Shown Mills, "Censuses: Analysis, Interpretation & Correlations," Course 4 Advanced Methodology, Interpretations & Analysis, Samford University Institute of Genealogy & Historical Research, 2005, 4M3.
- Thomas W. Jones, "Proved?: Five Ways to Prove Who Your Ancestor Was" (printed handout for a lecture presented to library staff, 23 October 2003, Family History Library, Salt Lake City), 1.