Questions: Quantitative and Digital History: Theory and Practice

Question 1

A historian builds a database of 'worker strikes' across industrial England from 1800–1900 and finds that strike frequency correlates with periods of wage decline. A critic points out that the coding scheme excluded certain forms of collective action common earlier in the period. This criticism highlights which fundamental issue with quantitative historical databases?

• A. The sample size is too small to support statistical inference
• B. Category decisions embedded in database construction are interpretive and political, not merely technical, and shape the findings
• C. Correlation cannot establish causation in historical data
• D. The correlation coefficient is inappropriate for time-series data

Answer: (B)

This is the core epistemological problem the topic addresses: quantitative historical databases appear objective, but their category definitions — what counts as a 'strike,' a 'riot,' a 'famine death' — require interpretive decisions that are embedded in the methodology and hidden behind the objectivity of numbers. The critic is correctly applying the same critical interrogation to quantitative sources that historians routinely apply to qualitative ones. Options C and D are real methodological concerns but not what this criticism targets.

Question 2

The Fogel-Engerman controversy over *Time on the Cross* (1974) became a flashpoint illustrating which tension in quantitative history?

• A. That statistical methods require too much computing power to be used by most historians
• B. That the precision of quantitative outputs can exceed the quality of the underlying data, and that mathematical objectivity can obscure the values embedded in research design
• C. That economic history is too specialized to contribute to mainstream historical questions
• D. That databases from plantation records are too incomplete to support any conclusions

Answer: (B)

Fogel and Engerman used plantation records to argue that American slavery was economically efficient — a finding that appeared to follow from the data but was simultaneously a methodological claim (was the database correct?) and a moral one (what does 'efficient' mean applied to slavery?). The controversy showed that quantitative precision can create an appearance of objectivity that obscures the values baked into the research design. The precision of the output can suggest more certainty than the data quality supports.

Question 3

Quantitative historical methods are especially well suited to questions about the distribution, frequency, and longitudinal change of phenomena across large populations.

• T. True
• F. False

Answer: True

This is the genuine advantage of quantitative history: it enables pattern recognition at a scale impossible through close reading or archival work with individual documents. Questions like 'how did strike frequency vary by region?' or 'what was the correlation between literacy rates and political participation across counties?' require quantitative methods to answer responsibly. The limitation is not that quantitative methods are weak in general — it is that they answer a different kind of question than qualitative methods, and the two are complementary rather than competing.

Question 4

Once a quantitative historical database is published with its methodology, the interpretive decisions embedded in it no longer require critical scrutiny.

• T. True
• F. False

Answer: False

Published methodology makes the decisions transparent, but transparency does not eliminate the need for scrutiny — it enables it. Category definitions, coding rules, and source selection are interpretive acts with historical and political dimensions. Revealing them invites critique. Quantitative outputs require the same critical interrogation as any other historical source. The appearance of objectivity is a presentation effect; the interpretive content is real.

Question 5

Why do quantitative historical findings 'not speak for themselves,' even when the statistical methods are correctly applied?

Model answer: Quantitative outputs — frequencies, correlations, regression coefficients — are produced by a research design that required interpretive decisions: what categories to use, which sources to include, how to handle missing data, what to measure as a proxy for what. These decisions are historical and political, not merely technical. A frequency count of 'riots' is partly a record of what happened and partly a record of how the historian defined 'riot.' Additionally, the meaning of a correlation or trend requires historical interpretation: why is this pattern there? Numbers do not contain their own interpretation.

The phrase 'numbers speak for themselves' implies statistical findings are self-evidently meaningful. But every quantitative result is downstream of contestable choices — as the cliometric controversy demonstrated. Even unimpeachable arithmetic does not produce unimpeachable historical conclusions if the underlying categories and sources are contestable.