browser icon
You are using an insecure version of your web browser. Please update your browser!
Using an outdated browser makes your computer unsafe. For a safer, faster, more enjoyable user experience, please update your browser today or try a newer browser.


To make understanding and discussion possible, we have to agree on what evidence can support a claim.

Before the 17th century, people relied on authority and tradition for evidence.

After then, some people began relying on observation and experiment for evidence.

When writing a research essay or report, you should emulate a good criminal investigator: Take no one’s word for anything. Test your claims with evidence.

For inexperienced, unconvincing writers, evidence is like an ornament on a Christmas tree. For experienced, effective writers, the evidence is the trunk and branches of that tree. Without evidence, claims collapse. The essay becomes a rant, an unloaded gun getting waved around.

Four general types of evidence to use in the essays for this course

Representations of reality in quantitative data

City Treasurer’s account book (Rekeningen van de thesaurier Dirk van der Dusse) from 1658, Delft, the Dutch Republic (today’s Netherlands)

Numerical/quantified data, especially statistics, that is, descriptions of facts

Analogies from other times and places based on quantitative data, especially statistics

In the absence of quantitative data:

Qualitative data, for example, interviews

Expert testimony

Anecdotes; hypotheticals

How evidence is expressed

Words: ink-on-paper and online

Images: still and moving (aka video)

Numbers: statistics, quantified reality

Where evidence is found

Primary sources

  • original materials on which other research is based
  • from a relevant time period
  • not filtered through interpretation or evaluation

Primary sources are usually the first formal appearance of results in physical, print or electronic format. The raw data. For purposes of this course, primary courses present original thinking about new information, especially based on observation, measurement, or surveys.

Cite primary sources whenever possible. You do not need to cite the search service (like Google) or the secondary or tertiary source that you first encountered and that led you to the primary source.

Videos and images can be primary sources, especially for anecdotes and analogies. For an historian like me, anything from my time period (The Dutch Republic from 1650 to 1750) is a primary source.

The farther along you get in research, either academic or corporate, the primary data is more likely to be uniquely yours. For an animal scientist in veterinary medicine, it’s the observations and test results for that animal on that day. For a forensic scientist in law enforcement it’s the physical objects recovered at the crime scene. For my historical research, I find primary documents like notary records or municipal financial ledgers (see image on right) that no other historian has found or written about before.

Meta-primary sources

(Note that some research guides call these tertiary sources.)

Raw data can be difficult to understand. Many organizations have taken it upon themselves to compile primary data in lists, graphs, and charts. They are not secondary sources because they are not interpreting and evaluating. However, they are helping the reader visualize the data. In that sense, they are making an original contribution that can be considered a primary source. You can cite these meta-primary sources.

Examples would be the Gapminder Foundation’s visualization project at

For another, the Wikipedia has data compilations that cannot be found elsewhere. It has a sortable table of the world’s countries that includes the percentage of each country’s GDP per capita that would be earned by someone working full time at that country’s official minimum wage. It is based on data from the World Bank, but as far as I can tell, this table is original to Wikipedia and shows a computation and comparison that is not shown elsewhere. In this case, I would cite the Wikipedia page and note that the information came from the World Bank.

Credible meta-primary sources always include their data sources, so you should make the reader aware of that. For example, much of Gapminder’s data comes from the World Bank. Much of Wikipedia’s global terrorism data comes from the National Consortium for the Study of Terrorism and Responses to Terrorism (START) at the University of Maryland.

Secondary sources

  • accounts written after the primary evidence and with the benefit of hindsight
  • reports, interpretations, and evaluations of primary sources

Secondary sources like magazine and website articles are seldom evidence, but rather commentary on and discussion of evidence. Avoid quoting from them, especially if the quote itself cites primary evidence.

Videos can also be secondary sources, for example, a newscast.

Tertiary sources

  • information which collects and distills primary and secondary sources.

Tertiary sources can be helpful to get your started on your research, so you would not stop at that level and cite such a source.

The above is adapted from a Yale University library guide, where you can learn more about sources.

Criteria for evidence

Use only valid and reliable evidence that is relevant to your claim. Use some evidence for every claim. Use enough evidence and cite the primary source whenever possible so that the reader can learn more.

Validity – Does it measure what it says it does?

Reliability – Is it evidence that further testing would confirm? Can it be generalized? If evidence is valid, it is also reliable. If evidence is reliable, it is not necessarily valid.

Relevance – Does it have a logical connection to the claim it is offered in support of?

Sufficiency – Is it enough to support the claim?