Abstract

Theoretically, the persistence of surface error in student writing may be understood, at least in part, as a normal side effect of development in writing skill. Language tactics newly attempted by a writer increase the likelihood that new mistakes will be made, or old mistakes made anew. This theory, that the context of writing improvement helps explain writing error, is tested by comparing the impromptu essay performance of college freshmen, sophomores, and juniors, and of postcollege employees. Eight surface errors were measured: misinformation of possessives, faulty predication, faulty pronoun reference, faulty syntactic parallelism, mispunctuation of final free modifiers, sentence fragments, comma splices, and misspellings. For each, four error rates were constructed in order to compare different ways of visualizing the relation of error to other aspects of writing. Generally, the findings support the theory: The college students here do measurably improve their writing and do continue making mistakes at about the same rate, but mistakes allied to the improvement. An implication is that undue efforts by teachers to prevent the mistakes may hinder the improvement.