Look closer. Find your family story.

You collected records for your direct Canadian line, but do they actually prove each connection from one generation to the next? When documenting a parent-child relationship, the number of records matters less than what each record says.

For example, a death record may identify the parents, but an informant provided that information years after the birth and may not have known it firsthand. An online family tree may name the parents without providing reliable evidence.

How can you tell whether you are following the correct ancestral line as you go further back in time?

First, lay out each person's records separately. Start with yourself. Examine your government-issued birth certificate. Does it name your parents correctly? If your name changes through marriage, divorce, or a court order, identify the official record that connects your former and current names.

Next, follow the relevant line backward from yourself to your parent, grandparent, and earlier generations until you reach the Canadian-born or naturalized ancestor. At each step, identify the authority-issued record that connects the child to the parent.

If the record names only one parent or neither parent, you need additional records to prove the relationship. A record that omits a parent may still tell you something useful, but it does not establish that parent-child relationship. Look for other records that might name the parent, then check whether they concern the same child and family. For example, the mother’s later marriage identifies her spouse, but it does not, by itself, identify the child’s father.

Compare the birth date or age, birthplace, and parents’ names, where available, with records from other points in that person’s life. If the details do not agree, or the records do not provide enough information to distinguish people with the same name, flag that parent-child connection for further research.

After reviewing each generation, you will know which parent-child connections the records support and which still need research. A record may name the expected parents but leave the person’s identity uncertain.

As an Accredited Genealogist®, I review the records, identify the unresolved connection, and develop a focused research plan.

Consider this hypothetical example: You have Grandma Jean's birth record. It names Bob and Wilma as her parents, but does not say where either parent was born. A U.S. census places Jean in a household with Bob and Wilma and lists Canada as Bob's birthplace. Is this the right family?

If so, how can you document Bob's Canadian birth when you cannot find his birth or baptism record? First, test the census household against other records. Compare Jean's age, parents' names, residence, and the other children in the household. Birth records for Jean's siblings may help identify Bob and Wilma across several years. A marriage record may connect Wilma's maiden and married names. Check those details before treating the census birthplace as a clue about Jean's father. The census gives you Canada, but it may not name a province or a town. Look for other records that might narrow Bob's birthplace and help distinguish him from another man with the same name.

Depending on when and how he moved, a border crossing record or passenger list might help. Read the original record carefully. A last residence, a birthplace, and a port of arrival describe different places. Once you have a likely birth location, check which birth records existed there during the relevant years and who holds them now.

The FamilySearch Wiki can help you learn about the place and its records. Search the FamilySearch and Ancestry catalogs for collections covering that location and period, rather than relying only on name searches, hints, or other family trees. Then contact the office or archive that holds the original records.

Consider baptismal records as well. Follow the family's religious connections, which may lead beyond the reported birthplace. Relatives or an earlier family residence might point to a congregation elsewhere, including in the United States.

Keep a search log as you work. Record each repository and collection you checked, the place and years it covers, when you searched for it, and the result. Record the result even when you find nothing.

Keep a separate correspondence log of requests to record holders, and save their emails, letters, and replies. If you cannot provide an official document, IRCC asks for a written explanation and proof that you tried to obtain it. Your actual correspondence helps document those efforts.

If the birth or baptism record remains missing, evaluate other records issued by the appropriate authorities to establish the facts you need. Do they identify the same Bob named on Jean's birth record? What do they actually say about his Canadian birth? Your research may document the connection, or it may reveal a specific gap that still needs evidence. Either way, you can explain what the records support, what they do not, and where the search should go next.

As an Accredited Genealogist®, I can review your records, identify the gap, and research the sources that may help document your Canadian ancestor’s birth.

One word in the 1850 census made me stop: Canada.

The Hampton family lived in Wyoming, Lee County, Illinois, but the census listed Canada as the birthplace for the family members. If I wanted to trace this family backward, I needed to answer a much more specific question: Where in Canada did they live?

It can be tempting to see “Canada” and immediately start searching Canadian records. But that is not where I would start.

Before crossing the border, I would fully research the family in the United States. I would follow them through every available census and build a timeline of their lives in Illinois. Later census records may provide additional clues about immigration or naturalization. Marriage and death records, military records, naturalization files, church records, obituaries, land records, and other local sources may also reveal where the family came from.

Each record adds another piece of evidence—or tells me where I still have a gap.

This family also illustrates why migration research does not always follow a straight line. The Hamptons' story did not simply begin in Canada and end in Illinois. Evidence suggests a much larger migration pattern involving Pennsylvania, Upper Canada, and eventually Illinois. Understanding that pattern requires looking beyond one person and one set of records.

Sometimes records can even lead you somewhere you do not expect. A family living in Canada might cross into Vermont or New York for a baptism, marriage, family visit, or another reason. A church record in Troy, New York, for example, does not necessarily mean the family lived in Troy. Relatives, a longstanding church connection, or a familiar priest could have drawn them there.

That is why I do not let an international border define the boundaries of my research.

Follow the family first. Then follow the border.

Consider this research example. You found your ancestor. The name matches, the location matches, even the dates fit. But what if you've found the wrong person?

Suppose you are researching Nellie Hampton, born in Missouri, and family information suggests her father was Joseph Hampton. Youwant to trace Joseph's ancestry, but first you need to establish which Joseph Hampton was Nellie's father.

In the 1900 U.S. Census for Missouri, I found two Nellie Hamptons, both born around the same time, both listed with Joseph as a father.

The two Nellies lived in different Missouri counties. One was born in May 1895 and the other in November 1898. Both had fathers named Joseph Hampton, but which one was the Nellie I was looking for? Their birth years weren't enough to answer that question. I traced both families through earlier and later census records, comparing household members and looking for details that might help me tell them apart.

I followed both Nellies into the 1910 census and found two different family situations. Nellie, born in 1895, lived with her widowed mother, Catherine. Her father had died sometime after the 1900 census. The other Nellie, born in 1898, still lived with both parents in the same Missouri county where I found her in 1900. By 1910, she was one of five children.

I continued tracing the first Nellie and located her 1912 Missouri death certificate. It identified her as Nellie May Hampton, born 18 May 1895, and named her parents as Joseph Hampton and Kate Eichler. The certificate provided additional details to help distinguish her from the other Nellie.

I followed the second Nellie into the 1920 census, where she still lived with her parents. A 1922 marriage record connected Nellie Hampton to Pearl Porter. I then located a Social Security NUMIDENT entry for Nellie Jane Porter, born 18 November 1898, which also identified her as Nellie Jane Hampton. The entry named her parents as Joseph P. Hampton and Mary V. Cox. These records helped connect the Nellie I had followed through the censuses to her married identity.

Joseph's death certificate revealed another important connection. When I located Joseph Preston Hampton's death certificate, I recognized the informant's name: Mrs. Nell Porter. This gave me another connection between Nellie and Joseph.

Nellie may have known the circumstances of Joseph's death firsthand, but what about the information she provided about his birth and parents? She wasn't there when he was born. She may have known his parents personally or learned their names through family stories.

The death certificate named Joseph's parents as Preston Hampton and Sarah Ockerman. Now I had two more names to research, but I couldn't assume the information was correct simply because it appeared on a death certificate. I would need other records to support those relationships.

This research started with two girls named Nellie Hampton, each with a father named Joseph. By following both families through the records, I could tell them apart and identify Joseph Preston Hampton.

It's easy to find someone whose name, age, and location seem to fit. The challenge is making sure you've found the right person.

When I first started researching my family history, I assumed a published family history book contained verified information. I didn't know to check footnotes or endnotes for source information. I spent hours—even days—copying and pasting names and information into my tree. He had done the research; why should I?

Later, I learned to treat information without supporting citations as clues to investigate, not facts to accept without checking. I wasted a lot of time.

These five common mistakes can lead your research in the wrong direction—and leave you retracing your steps.

1. Trusting transcribed information without checking the record image

While researching an individual, I found a record on someone else’s tree. The transcription appeared to match the name and age of the person I had in mind. When I opened the original record image, however, I noticed that the name differed from the transcription.

Takeaway: Check the actual record image whenever possible. Transcriptions and indexes help you find records, but they can contain errors.

2. Searching without a research plan

When I first started, I jumped into databases without a clear research question. Once, I spent more than an hour searching for an individual across several databases before realizing I had followed the wrong family.

Takeaway: Start with a specific question. Identify sources that might help you answer the question and check each lead against what you already know before following it further.

3. Failing to keep a research log

After spending time searching a database, I had a déjà vu moment. Had I already searched here? Eventually, I realized I had searched that same database without finding what I needed. I did not record the search and spent roughly another 30 minutes repeating work I had already done.

Takeaway: Record the databases and record sets you search, the names and search terms you use, and your results—even when you find nothing. A research log helps you avoid repeating searches and frees up time to explore other sources.

4. Accepting a date because two sources agree

Conflicting dates come up often in genealogy. Suppose a death certificate and a tombstone both give a birth year of 1869, while a baptism record lists 1867. Two sources agree, but that agreement alone doesn’t settle the question.

The death certificate and tombstone may draw on information from the same informant, who may not know the deceased person’s correct birth year. If the baptism took place shortly after birth, that entry offers information closer to the event. Consider when each record originated, who supplied the information, and how the records connect to the person you’re researching. How might the informant have known the birth date?

Takeaway: Two records can repeat the same mistake. Compare the details, consider when each record originated and who supplied the information, and investigate conflicting evidence before choosing a date.

5. Assuming a family relationship without evidence

I initially assumed two people shared the same parents because they shared a surname and lived in the same town. Land records pointed to different family lines and challenged that assumption.

Takeaway: A shared surname and location do not prove kinship. Look for evidence in land, probate, and other records that connect people to their families.

What I Learned

I can’t recover the hours I spent following the wrong leads, but I can use those lessons to guide my research now. Checking record images, keeping a research log, comparing evidence, verifying relationships, and working from a clear question all help me avoid repeating those mistakes.

Have you made a genealogy mistake that changed how you research?

What happens when two historical records tell completely different stories about the same family?

As a professional genealogist, I frequently encounter conflicting information. Sometimes the discrepancy results from an enumerator’s mistake, an informant’s faulty memory, or someone misunderstanding a question.

But occasionally, that inconsistency leads to an unexpected discovery.

While researching Philipp Henry Siegel and his wife, Catharine Schlemmer, I encountered a discrepancy between two census records that raised an intriguing question: Did this family have a child whose existence had gone undocumented in the records I had found?

What followed illustrates how genealogists use indirect evidence to investigate missing family members, question existing conclusions, and piece together information that no single record provides.

A Census Mystery

Philipp and Catharine married on 23 August 1888 in Monroe Township, Allen County, Indiana. By 1900, the couple lived in Madison Township with four children: Charles, George, Matilda, and Henry.

Their birth years placed their approximate ages at the time of the 1900 census as follows:

Charles, born in 1889 — approximately 10 or 11 years old

George, born in 1893 — approximately 6 or 7 years old

Matilda, born in 1895 — approximately 4 or 5 years old

Henry, born in 1899 — approximately 1 year old

1900 U.S. census, Madison Township, Allen County, Indiana, Ancestry.com

All four children were born after Philipp and Catharine married, supporting the possibility that they were children of the marriage. Their presence in the household and their ages provided important context, although neither independently established Catharine as their mother.

The 1900 U.S. census included two questions specifically for married women: How many children had they given birth to, and how many of those children remained alive?

According to the census, Catharine had given birth to five children, but only one survived. Yet four children appeared in the household. Something clearly did not add up.

My initial thought was that the enumerator might have recorded the information incorrectly. Perhaps someone other than Philipp or Catharine answered the questions and did not know the correct information. The informant might also have misunderstood the question and reversed the numbers.

I turned to the 1910 census for another perspective.

This time, the census reported that Catharine had given birth to six children, with five still living.

1910 U.S. census, Monroe Township, Allen County, Indiana, Ancestry.com

Now I had two conflicting accounts:

Census	Children born	Children living
1900	5	1
1910	6	5

Both could not accurately describe the family’s recorded history. If the information from 1910 was more reliable, it suggested that Philipp and Catharine had experienced the death of a child. But who was this child, and what records might identify them?

Building a Timeline Before Searching

Rather than immediately search for an unidentified child, I began with the children I already knew. The four children recorded in the 1900 household were born in 1889, 1893, 1895, and 1899.

Examining the intervals between their births revealed several possible periods when another child could have been born: 1890–1892, 1894, and 1896–1898.

These intervals did not prove another birth occurred. They narrowed my search and helped me develop a working hypothesis. This distinction matters in genealogical research. We cannot assume an event occurred simply because it fits a timeline. We need evidence. With these possible dates established, I began searching for birth and death records.

When the Expected Records Do Not Exist

I searched the available Allen County and Indiana birth and death indexes covering 1888 through 1905, including variations of the Siegel surname. The searches produced no matching record for an unidentified child.

Unfortunately, nineteenth-century vital record coverage was inconsistent. Allen County began recording births within Fort Wayne city limits in 1882 and outside those limits in 1887. Death records began earlier within the city, but records outside city limits did not begin until 1907.

Even where registers existed, recording every birth and death was not compulsory. Consequently, the absence of a record did not establish that the child never existed. I needed another approach.

A Clue in a Family Cemetery

My research led me to cemetery records in Monroe Township, where the Siegel and Schlemmer families had lived.

An 1880 township plat map showed the Siegel and Schlemmer families owned land near one another. The intersection also included a school, German Lutheran church, and cemetery.

Searching local burial information brought me to Schlemmer Cemetery, also known as Hoffman Cemetery. The Mary Penrose Wayne Chapter of the Daughters of the American Revolution had documented the cemetery’s gravestones.

Among the entries, I discovered a child named Mary E. Siegel. A gravestone photograph and transcription recorded her birth as 29 June 189[?] and death as 17 July 18[?]. Portions of the dates were difficult to decipher. Mary’s surname, possible dates, and burial location made her a candidate worth investigating.

Mary E. Seigel tombstone, FindAGrave.com (broken tombstone)

Other Siegel relatives rested in the same cemetery, including Richard and Katharina Siegel, Philipp’s parents, and Richard’s sister Anna.

Unfortunately, Schlemmer Cemetery had no surviving burial records. Available transcriptions listed the individuals alphabetically, so I could not determine whether Mary’s grave stood alongside other family members.

Burial proximity might have provided another clue, but I could not establish it. Nevertheless, Mary offered a possible connection to the missing child.

An Online Memorial Presents Another Problem

When I located Mary’s Find a Grave memorial, I discovered that someone had already identified her parents. The memorial named Reichard Siegel and Wilhelmina “Minnie” D. Germann.

At first glance, this identification could have ended my investigation. But an online memorial’s statement requires the same scrutiny as information found in other compiled genealogical sources. I needed to determine whether the available records supported the claimed relationship.

Reichard and Minnie married on 5 June 1898 in Van Wert County, Ohio. The 1900 census placed them in the household of Reichard’s father, Richard Siegel, in Monroe Township.

According to that census, Minnie had one child, who was still alive. Their son Roland, born in April 1899, appeared in the household. The 1910 census reported that Minnie had two children, both living. The second child, Elanora, was born in 1901. These records raised an important question.

If Minnie had only two children and both survived, how could Mary, whose gravestone indicated that she died during childhood, also be her daughter?

The census entries did not establish that the memorial was incorrect. After all, I had already encountered conflicting birth and survival counts in Philipp and Catharine’s records. However, they certainly challenged the identification and required further investigation.

Following the Indirect Evidence

At this point, I had no birth certificate naming Mary’s parents, no death record identifying them, and no surviving burial register establishing the relationship. Instead, I had several pieces of evidence to evaluate together.

The 1910 census suggested that Philipp and Catharine had lost a child. The gaps between their known children’s birth dates provided potential time frames. A child named Mary E. Siegel appeared in a cemetery associated with their extended family. Meanwhile, the census records for Reichard and Minnie did not readily support the parental identification given on Mary’s online memorial.

Each record answered one question while raising another. None independently proved that Mary was Philipp and Catharine’s daughter. Together, however, they supported a reasonable hypothesis that Mary might have been the unidentified child.

This is where indirect evidence becomes particularly valuable.

Direct evidence answers a specific research question without requiring an additional inference. Indirect evidence contributes information that must be combined with other evidence to answer that question.

A census entry recording the number of children a woman had given birth to does not identify an unnamed deceased child. A gravestone does not necessarily identify the child’s parents. And a family’s association with a cemetery does not establish the parentage of everyone buried there.

When correlating these records, examining their reliability, and considering alternative explanations, we can begin developing a documented conclusion.

What Did the Research Establish?

The investigation uncovered a child whose name had not appeared among Philipp and Catharine’s known children and revealed weaknesses in an existing identification of her parents. It also showed that the 1900 and 1910 censuses required careful evaluation, not unquestioning acceptance.

The evidence developed a plausible case for Mary E. Siegel as Philipp and Catharine’s missing child, although the records examined did not conclusively establish her parentage.

That distinction matters.

A responsible genealogist must recognize the difference between a promising hypothesis and a proven conclusion. We should never present an inference as fact simply because the explanation seems reasonable.

The research also demonstrates that unsuccessful searches can contribute to an investigation. Knowing which records were examined, understanding their limitations, and documenting negative findings all help establish what the evidence does—and does not—tell us.

The Lesson: Follow the Evidence, Not the Assumption

This investigation began with two census records that contradicted one another. Rather than choosing the more convenient entry, I questioned both, built a timeline, searched for additional records, and examined a competing identification.

That process eventually led me to Mary E. Siegel and a possible explanation for the family’s conflicting census information. When researching your own family history, pay attention to discrepancies. A disagreement between records does not always mean your research has reached a dead end.

It may reveal a question you have not yet considered.

Sometimes the most meaningful discoveries begin not with what a record tells us, but with what does not make sense.

At Ancestors of Yesteryear, I help families discover, understand, and document their family history through thorough, evidence-based genealogical research.