Wim Gombert

64 CHAPTER 4 Table 8 gives a rough estimate of time spent on the activities speci cally targeting the listening skills and on other supposedly contributing program elements. TABLE 8. Time spent on listening skills (including literature) Program elements specifically targeting listening skills SB DUB Extensive listening (just listening) 10 50 Intensive listening (listening and exercises) 35 0 Listening exam practice 10 0 Listening exam strategies training 5 0 Amount of instruction time for listening activities: 60 50 Percentage: 15% 13% TESTS Reading and listening skills were tested in the nal year on the basis of the national nal exams, which were developed by Cito as the national testing agency. e nal reading exams consisted of di erent kinds of closed-type questions (like true/false, cloze, etc. but mostly multiple choice) and a limited number of open-ended questions. e nal listening tests consisted exclusively of multiple-choice questions. ese tests were o cial nal exams, so they were not the same for all cohorts, but the Cito (Dutch national testing agency) provides grades which are normalized to ensure reliability between years. erefore, the nal grades were taken as the basis for the analysis. STATISTICAL DESIGN As the independent variable, teaching program, only had two levels (SB or DUB), Hotelling’s T2 was selected as an appropriate test to determine the e ect of both teaching programs on the reading and listening skills a er six years of instruction. SPSS for Windows 27.01 (SPSS, Inc. Chicago, IL) was used and basic requirements of Hotelling’s T2 were checked. SPSS was used to test for linearity, multicollinearity, outliers, normality, sample size and homogeneity of variances. A er checking assumptions, Pillai’s Trace (p < .001) was calculated to determine if there was a signi cant di erence between the programs and the e ect size was also calculated using Cohen’s d.

RkJQdWJsaXNoZXIy MTk4NDMw