The purpose of this experimental study was to determine which of three methods of vocabulary instruction, including the common approach method, a combination of the common approach method and keyword method, and a combination of the common approach method and the keyword/illustration method of instruction was most effective in teaching mathematics vocabulary over the course of two mathematical units.
Grade eight students (N=60) from two rural middle schools who scored at the basic level on the 2006 Maryland School Assessment (MSA) in mathematics and reading participated in this experimental study. Students with individualized education plans (n=24) and regular education students (n=36) received one of the three types of computer-assisted vocabulary instruction. Stratified random sampling based on MSA scores in mathematics was used to determine the method of instruction a participant received.
The study utilized a pretest-posttest experimental design. The dependent variable was pretest and posttest scores. The independent variable was method of vocabulary instruction. There was one pretest and one immediate-recall posttest for each of two mathematical units. Several previous studies of vocabulary acquisition (Nagy, Andersen & Herman, 1987; Purkel & Bornstein, 1980; Rosenbeheck, Levin & Levin, 1989; Wang, Thomas & Ouelette, 1992) assessed participants immediately after instruction. Participants were provided individual vocabulary instruction between the pretest and posttest. Participants were exposed to each vocabulary term for one minute. Previous vocabulary studies used similar amounts of time per exposure (Mastropieri, Scruggs & Mushinski-Fulk, 1990; Wang, Thomas & Ouellette, 1992). Total instruction time per unit for this study was 8 minutes. There were eight multiple choice questions on each pretest and posttest. Therefore, there was a total range of continuous scores of 0–16 on the combined pretests and posttests scores. A high score indicated more vocabulary knowledge and a low score meant less vocabulary knowledge. The pretest, instruction, and posttest were presented via PowerPoint.
An analysis of raw scores indicated that participants identified with special needs did not have scores equal to their non-disabled peers on the vocabulary pretest and the immediate recall vocabulary posttest, regardless of method of vocabulary instruction. As a whole, participants with special needs scored lower on the pretest and on the posttest than their regular education peers. However, an Analysis of Variance - Repeated Measures revealed there was a significant overall change from pretest to posttest scores (p=.001) indicating that all three methods of vocabulary instruction were effective. This converges with the findings of Nagy, Anderson, and Herman (1987) who found that if exposed to a word once for a limited time, individuals can learn. Nevertheless, this analysis also revealed no vocabulary group, regardless of method of instruction, performed better than another (p=.355). This does not converge with previous studies using the keyword/mnemonic method of vocabulary instruction (Kleinheksel, 2005; Mastropieri, Scruggs & Levin, 1986; Mastropieri & Scruggs, 1990; Nolan & Polloway, 1992; Uberti, Scruggs & Mastropieri, 2003). Finally, Pearson correlations among pretest scores, posttest scores, MSA scores in mathematics, and MSA scores in reading revealed positive correlations. The strongest correlation was between MSA scores in reading and mathematics (r=.451). This finding converges with the finding of Kipplinger, Haug, and Abedi (2000) that there is a positive correlation between one’s ability to solve word problems and one’s proficiency in reading the English language.