The effects of output task types on noticing and learning of the English past counterfactual conditional

Focus of the study: This study too tests the Output Hypothesis; however, it also looks at the interaction of pushed output with task type by comparing two kinds of tasks which are commonly used in the instructed SLA research. By using reconstruction and picture-cued writing task, the article aims to compare differential effects of these two types of tasks on learning of English past counterfactual conditional. This study questions whether there is a relationship between output task type and noticing and subsequently learning of a target structure. This study examines the effect of output and type of output tasks on noticing and learning of English past counterfactual conditional. 
Instructional settings and participants:  The study took place in a required-English course at a Korean university in Seoul. The participants were 52 intermediate-level undergraduate students, whose proficiency was determined by the Seoul National English Proficiency Test. The study was conducted over a period of approximately four weeks. 10 intact classes were randomly assigned as one of these three groups: experimental group 1= reconstruction (EG1), experimental group 2= picture-cued writing task (EG2), non-output or control group (CG).     
Instrumentation and procedure: All groups in the study received first input and then were asked to work on a task. Whie this task was production-based for output groups, it was comprehension-based for the non-output group. The input was given through a reading passage including many instances of English past counterfactual conditonal. After reading the passage, the reconsrtuction group participants was asked to reconstruct the reading passage without looking the passage as correctly as possible. They were also told to underline important words or phrases to use the following task. The picture-cued writing task group was given 8 pictures and prompts under each picture. The participants were intructed to write a past conditional sentence combining two pictures. The nonoutput group worked on comprehension questions after reading passage and they wer not allowed to look back at the passage, either. To measure learning of the conditional forms, all groups were given a pretest and a posttests of recognition and production tests. In the recognition test, students listened and saw sentences and judged if the sentences were correct or not and if not, underline the mistake and correct it. In the production test, the students were given 3 contexts and they were asked to write sentences containing past conditionals. 
Results and discussion: In terms of noticing, the three experimental groups did not differ significantly. However, the output groups triggered noticing significantly more than non-output group. For noticing measure, underlining method was used as in Izumi. This method should be taken into consideration for evaluating the effect of noticing on learning. Moreover, the output groups outperformed the non-output groups on the production post-test. The two output groups did not differ in terms of task type on the production test. Overall, this study shows confirming evidence for the noticing function of the Output Hypothesis and how it benefits learning of L2 structures and how output-taks promote learning in comparison to non-output tasks.