UBC Theses and Dissertations
Word discrimination and spelling : an experimental study Mason, Geoffrey Pliny
Introduction In 1922, Gates obtained a correlation of r = . 55 between spelling ability and a speeded test measuring the ability to discriminate small differences between paired words. This study results from Gates' finding and is divided into two parts. Part I is concerned with checking Gates' results and determining if the speed factor is essential for the word discrimination test. Part II examines the feasibility, at the Grade VI level, of employing drill exercises in word discrimination as an aid to spelling. Part I Tests of spelling, mental ability and word discrimination were given to 211 pupils in a Junior High School in Victoria, British Columbia in the third week of Grade VII. The word discrimination test was given speeded with a one minute time limit, and unspeeded with a four minute limit. The correlation between the speeded discrimination test results and spelling was almost identical to that obtained by Gates thirty-three years ago. The correlation between the word discrimination test unspeeded and spelling was low at r[subscript bis] = .18, though still significant at the .05 level of confidence. Mental age correlated with spelling r = .66 and with word discrimination speeded r = .41. A first order correlation between word discrimination speeded and spelling, with mental age held constant, gave r[subscript 12.3] = .40. Separate correlations computed for boys and girls revealed no significant difference between the corresponding coefficients. Part II Thirty-two drill sheets involving exercises in word discrimination were used with an experimental group of 107 pupils in Grade VI. These drills were started-at the end of September and were conducted daily for the first two weeks, three times weekly for four weeks, and twice weekly for five weeks. After completion of the drill exercises it was found that, in both word discrimination and spelling, the experimental group had made gains, significant, at the .01 level of confidence, over a control group of 97 pupils. It should be noted that the drill exercises in word discrimination were done during part of the school time allotted to spelling, so that, in effect, the experimental group received one third less spelling instruction during the eleven week experimental period than did the control group. Conclusions The ability to discriminate small differences between words is significantly related to spelling ability and is measured more effectively by a speeded test. It is suggested that the inadequate discriminatory ability when dealing with word forms shown by the weaker spellers is a result of insufficient emphasis on this skill, possibly when learning to read. It is evident that most of the students examined by this study could have derived benefit from exercises in word discrimination beginning, probably, in the primary grades. Finally, as the initial spelling achievement of the experimental group compared favourably with Provincial norms and norms obtained from the United States, it is possible that the drill exercises have very wide applicability.
Item Citations and Data