Effects Of Testing Methods And Testing Formats On Second Language Testing Effect

In recent years, substantial studies have shown that testing is an effective way to promote learning, and testing effect has become a research hotspot for cognitive psychologists and educational psychologists. With the development of the studies of testing effect, researchers have been trying to find out the optimal testing effect under different testing conditions. Prior studies have found that different testing methods, for example, providing feedback and retest after testing are both effective means to promote the effect of testing. Besides, different testing formats, such as multiple-choice(MC) test and gap-filling(GF) test, can lead to different effects on acquisition, and there may be certain interaction effects between testing methods and testing formats. However, there is no consensus on the effects of testing methods and testing formats in prior research, and the research is insufficient, thus more research is needed.The study of testing effect on SLA is just beginning. Which testing method and testing format can bring more favorable acquisition effects to L2 learners remains to be an important issue to be further discussed and addressed. Therefore, the current study aims at exploring the effects of testing methods(feedback & retest) and testing formats(MC & GF) on L2 testing effect. Three research questions are proposed below:1). How do testing methods(feedback & retest) affect the acquisition of L2 non-finite verbs?2). How do testing formats(MC & GF) affect the acquisition of L2 non-finite verbs?3). Are there any interaction effects between testing methods(feedback & retest) and testing formats(MC & GF) on the acquisition of L2 non-finite verbs?An experiment was carried out on a total of 181 subjects who were students in second grade in a junior middle school, and 173 subjects took part in the main experiment. Another eight students, who shared the same language proficiency level with the subjects in the main experiment, participated in think-aloud. L2 non-finite verbs was chosen as the target form in the present study. Subjects were randomly divided into four groups: the FMC group, the FGF group, the RMC group and the RGF group. During the experiment, subjects were first instructed to learn knowledge about non-finite verbs, and then an initial test was maneuvered to examine subjects’ knowledge on selected target form. For the MC groups, the initial test was a MC test, and for the GF groups, it was a GF test. An intervening task was given to all of the subjects after the initial test, and then the feedback groups received feedback and the retest groups received a retest. One week later, all subjects took a final test, which contained both MC items and GF items. The data of the subjects in think-aloud was excluded from the analysis of the experimental results. The results are summarized as follows:1) Testing methods affected the acquisition of L2 non-finite verbs, and the effects of feedback were better than retest. It was chiefly because feedback presented information input to learners for a second time, which would help learners notice the gap between their production and the target form, so as to correct learners’ mistakes, and learners could conduct deep processing of the target form. Contrarily, although retest provided an opportunity for the conduction of elaborative retrieval, yet retest did not generate real reconstruction of the knowledge and acquisition because of the lack of correct information, and then it would lead to a high rate of forgetting and poor acquisition effects. In addition, feedback could enhance learners’ confidence in answering later tests, while repeated testing might make learners become impatient and anxious, which would also influence their performance on the final test.2) Testing formats affected the acquisition of L2 non-finite verbs, and the effects of GF test were superior to MC test. Compared with MC test, GF test was with more retrieval difficulty and required more retrieval efforts, and it would enhance leaners’ storage strength of the target form, and then would lead to better retention of the information tested. Additionally, GF test would bring greater transfer effects to the final test than MC test. Moreover, MC test contained both correct and wrong answers in its alternatives, which might make learners get more familiar with the wrong answer and misguide learners to learn false information, impeding the effects of acquisition.3) There were no statistical interaction effects between testing methods and testing formats, which meant that the difference of testing formats(MC & GF) did not affect the effects of testing methods(feedback & retest), and the difference of testing methods(feedback & retest) did not alter the effects of testing formats(MC & GF). More specifically, the superiority of feedback was found in both MC test and GF test, and it meant that the effect of feedback was better than retest regardless of the testing formats(MC & GF). Additionally, the effect of GF test was detected to be better than MC test under both the feedback condition and the retest condition, which indicated that MC test was inferior to GF test regardless of the testing methods(feedback & retest).It can be inferred from the findings above that feedback was more effective on promoting the testing effect and the effects of GF test were better than MC test on the acquisition of the target form. Meanwhile, it was worth noting that the FGF group outperformed the remainders, which manifested that the testing method as feedback and the testing format as retest could generate the optimal effects on the acquisition of the target form. Thus, L2 instructors should choose proper testing methods and testing formats on the basis of the testing effect, which might make contributions to ideal learning outcomes.