Prior spelling test lists (Ayres, Buckingham) are too easy for better pupils, with more than half of eighth-graders spelling all Buckingham list words correctly¶

Claim · All claims
Evidence · 1 study · design r2 · q1
Currency · newest study 1910 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q1 i? The author argues that Ayres' ten-word-per-grade lists are unreliable for individual pupils and too easy for better pupils, and that in the eighth grade more than half of pupils can spell all words in Buckingham's lists correctly. → Starch 1910

Evidence¶

Starch 1910¶

Starch, D. (1910). The Measurement of Efficiency in Spelling, and the Overlapping of Grades in Combined Measurements of Reading, Writing and Spelling. The Journal of Educational Psychology. https://eric.ed.gov/?q=The+Measurement+of+Efficiency+in+Spelling

q1 · i? · design · r2

Methodological critique of earlier test construction, citing Buckingham's published distribution curves as evidence that the words are too easy for better pupils even as far down as the fourth grade. The author also notes the small Ayres lists make individual scores unreliable and prevent measuring grade-to-grade progress.

"In the upper grades more and more of the pupils can spell correctly all the words in the list until in the eighth grade more than half of the pupils can spell all of the words correctly."

Discussion¶

Related Claims¶

• Grades overlap massively in spelling: on average 20.8% of pupils reach or exceed the median of the next grade above — related
• Two 100-word dictionary-sampled spelling lists administered to the same pupil yield scores agreeing within a mean variation of 2.2 points — related
• The six 100-word spelling lists are practically identical in difficulty, differing from list 1 by .9 to 3.4 points — related
• Three factors make for overlapping between grades — test unreliability, unimportance of the function tested, and improper classification — and improper classification can be established only when the first two are ruled out — related