A $25,000 commission, and a committee that ran for thirteen years
The National Research Council's committee began as the School Scale Board in 1919 and was not dissolved until 1932 [4]. Select a numbered point on the chart.
Five names on the title page
The Smithsonian's copy lists the authors alphabetically, Haggerty first; a 1920 notice in The Elementary School Journal put the chairman, Yerkes, first [1]. Select a name.
Chairman of the committee, as he had chaired the Army's Committee on Psychological Classification of Recruits.[1]
Army Alpha (1917) →Two booklets, five tests in each
Each scale was meant to take about thirty minutes [1]. The test names, item counts and sample items below come from the published description of the Estonian adaptation [5], which is the most detailed modern account; the 1920 American forms may differ in item counts.
Sixteen yes-or-no questions, forty seconds
The Vocabulary test asked plain questions about what words mean, answered yes or no [5]. The questions here are ours, written in that format and getting harder as they go. You may skip. When time runs out, see how your score changes under the two ways of counting.
Why guessing matters on a yes/no test
Comparing Estonian pupils of 1933–36 with those of 2006, Must and Must found that later pupils left fewer answers blank and guessed more [6]. Correct answers rose by about 0.79 SD, but wrong answers rose by about 0.15 SD too [9]. Scoring rights minus wrongs cut the gain on Arithmetical Reasoning and Vocabulary by half [6].
Illustration, not study data: a pupil knows 20 of 40 Vocabulary items for certain and guesses on some share of the rest, getting half the guesses right.
Estonia kept the answer sheets
An Estonian adaptation was given to schoolchildren in the 1930s, and the same tests were given again in 1997–98 and 2006 [8][7]. Because item-level answers survived, researchers could see which tests rose, which fell, and how [9]. The chart shows Scale B, 1934 against 2006 (890 and 913 pupils), as effect sizes [7].
References
- [1]Two Psychological Tests with Manual and Key, National Intelligence Tests, Scale A Form 1 and Scale B Form 1 (1920). National Museum of American History, object 1990.0034.094.americanhistory.si.edu
- [2]Psychological Test Documentation, National Intelligence Tests (Manual of Directions, 1921 Revision; key to Scale B, Form 1). National Museum of American History.americanhistory.si.edu
- [3]Teacher, Tester, Soldier, Spy: Psychologists Talk about Teachers in the Intelligence-Testing Movement, 1910s–1930s. History of Education Quarterly. Cambridge University Press.cambridge.org
- [4]Differential Minds: Mass Intelligence Testing and Race (doctoral dissertation). Johns Hopkins University, JScholarship.jscholarship.library.jhu.edu
- [5]Must, O. & Must, A. (2014). Data from "Changes in test-taking patterns over time" concerning the Flynn Effect in Estonia. Journal of Open Psychology Data, 2(1), e2. doi:10.5334/jopd.abopenpsychologydata.metajnl.com
- [6]Must, O. & Must, A. (2013). Changes in test-taking patterns over time. Intelligence, 41(6).sciencedirect.com
- [7]Shiu, W. (2012). An Examination of the Flynn Effect in the National Intelligence Test in Estonia. ProQuest dissertation; ERIC ED551855.eric.ed.gov
- [8]Wicherts, J. M. et al. (2004). Are intelligence tests measurement invariant over time? Intelligence, 32. Study 5: Estonian children, 1934/36 and 1997/98.gwern.net
- [9]Williams, R. L. (2013). Overview of the Flynn effect. Intelligence, 41(6).sciencedirect.com
- [10]Must, O., te Nijenhuis, J., Must, A. & van Vianen, A. E. M. An item-level examination of the Flynn effect on the National Intelligence Test in Estonia (abstract and related records). ResearchGate.researchgate.net
- [11]Yerkes, R. M. (Ed.). (1921). Psychological examining in the United States Army. Memoirs of the National Academy of Sciences, Vol. 15. Internet Archive.archive.org
Related historical tests
The Smithsonian describes the NIT scales as adapted from the Army's wartime tests. Alpha was the written one, and Yerkes chaired both projects.
Otis had already published a group scale for schools by 1918. He also worked in the Army statistical unit Thorndike ran.
Haggerty brought out his own group examination the year after the NIT. It reached higher, into grade 12.
Before group tests, the Stanford-Binet was how a school measured one child at a time. The NIT promised a whole class in about half an hour per scale.
Thorndike's later scale kept two NIT ingredients, sentence completion and arithmetic. It added vocabulary and following directions.
The Army's official report records the wartime work of Yerkes, Terman, Whipple and Thorndike. It is the background to everything the committee built for schools.