Word Length and Complexity Variables in Spelling Difficulty

Abstract

THE COMPILERS of spelling lists at the present time seem to have two major variables in mind when they make up their lists. They are concerned with the frequency of occurrence of a word, because frequent words are those which children are most likely to use. For much the same reason these authors are interested in the meaningfulness of words. Children do not learn to spell, nor are they likely to use those words which they do not understand. Some of the; emphasis on frequency and meaningfulness operates, no doubt, at the expense of other vari ables which could serve as the basis of estimating the difficulty of the word. It is reasonable to assume that the longer a word is, the greater the probability of error in spelling it. In addition to this, the hard words (the spelling demons) always involve some complexity, some two or more letters which are easily confused. It is the purpose of this paper to de termine the relationships between spelling difficulty and frequency of oc currence, the average grade placement of words on spelling lists, mean ingfulness, word length, and two measures of word complexity. The in vestigation of these relationships should enable us to make more s o und predictions of spelling difficulty than is possible by use of the two con ventional variables. For the present paper, spelling difficulty will be defined as the per cent of error per occurrence and will be determined from words select ed from data published by Brittain and Fitzgerald (1). These data are a list of words used by second grade children in writing compositions. The frequency of occurrence and the frequency of spelling error are available for each word. A random sample of 75 words was selected from the Brittain and Fitzgerald list. The spelling difficulty, or per cent of spelling error per occurrence for each word, was calculated from these data. Frequency of occurrence of the sample of words selected was taken from two sources; Rinsland (3) and Brittain and Fitzgerald (1). Theav erage grade placement of the words on spelling lists was found in Gates (2). Word length is measured by the number of letters in the words. Word complexity was measured in two ways. First, the com plexity of the word as a sound pattern is the sum of the judged s i m i lar ity in sound between adjacent letters. This is termed the Sound Dis criminability Index. Second, the complexity of the word as a shape pat tern is the sum of the judged similarities in shape between the letters. This is named the Shape Discriminability Index.