Explicit and implicit knowledge of English tenses in primary school efl learners in Bosnia and Herzegovina

Data collection procedures and analysis

Prior to distributing the instruments, we obtained adequate informed con- sent from the school administration and the participants’ parents and the participants themselves upon clarification of anonymity, confidentiality, and the volunteer nature of the participation. Following the procedure of other, similar studies (Ellis, 2005), we first gave the participants the test measuring implicit knowledge (EIT), then, 7 to 10 days later, the test meas- uring explicit analysed knowledge (GJT), and finally, the test measuring ex- plicit metalinguistic knowledge (MLT), the correction part preceding the explanation part. The oral test lasted approximately 7.5 minutes for each participant. The other tests were not time pressured, but the students need- ed approximately 20 minutes for the GJT and 15 minutes for the correction part of the MLT, while all the students finished the explanation part in 20 minutes.
The data analyses were performed in SPSS. Table 1 shows that all the tests were reliable, as the Cronbach’s alpha coefficients ranged from α=.876 (for the GJT) to α=.956 (for the MLT total). Therefore, all the tests were internally consistent. The values for asymmetry and kurtosis were taken into consideration to prove normal distribution, with values between -2 and +2 for asymmetry and kurtosis considered acceptable in order to prove normal distribution (George & Mallery, 2010). Descriptive statistics were calculated first, and then a paired samples T-test to illustrate the difference between the different types of knowledge, and Pearson correlation coeffi- cients to show the relationship between explicit and implicit knowledge, while the main and interaction effects of the aforementioned three factors on the participants’ EFL knowledge were tested by a three-way ANOVA, and on the separate scores by a factorial MANOVA.
Table 1
Reliability coefficients for the tests