Date of Graduation


Document Type



This experiment studied the effect of size and format of instructions given to college students enrolled in an introductory statistics course on the number of errors committed and the amount of time required in order to complete a problem. A 3 (size) by 2 (format) factorial design was used. The three sizes of instruction given were propositional structures, subgoals, and complete algorithm. The two formats used were word and formula. The dependent variables used to study errors were Completed Errors, Total Errors, and Accuracy. The dependent variables used to study time were Solution Time, Steps Attempted, Errors per Minute, and Steps per Minute. Using a series of two-way analysis of variance procedures and multiple comparison tests as appropriate, this study found that students committed significantly more errors when the problem was given in either the Formula-Subgoal or the Formula-Complete condition than when it was given in the Formula-Proposition, Word-Proposition, Word-Subgoal, or Word-Complete format. Formula conditions required significantly more time to complete than word conditions, and the propositional conditions required significantly more time than did either the subgoal conditions or the complete conditions. Students in the word conditions attempted to complete significantly more steps than did students in the formula conditions. The Formula-Subgoal and FormuIa-Complete conditions resulted in significantly more Errors per Minute than did any of the three word conditions. Students attempted more steps per minute in the word conditions than in the formula conditions. They also attempted more steps per minute in the complete conditions than in the propositional structure conditions. Students reported perceiving the formula conditions as more difficult than the word conditions. There was a –.493 correlation between the perceived problem difficulty and the Total Errors committed. This means that students who perceived the problem as more difficult had a greater number of Total Errors than those who perceived it as easy.