Academic Corpora to Enhance the Revision Process The

Academic Corpora to Enhance the Revision Process: The Google Alternative Randall Sadler Ulugbek Nurmukhamedov

For those who can’t stand the suspense! Introductions (us) Intro of the study Definition Lit Review – Error Correction – Corpus studies – Technology & Feedback Participants Methodology – Data Collection – Tools Data Analysis Results Problems & Future Research

Introduction to the Research Students making heavy use of Google in research……maybe too heavy? Also heavy electronic and paper dictionary use Using MS Word Spell Czech Also using Grammar Check—great for avoiding the horrors of the passive sentences. …. . so why don’t they all have perfect grammar? ? ?

Definitions: Google Wildcard Search GWS refers to the use of an * in a Google search Original term: The results are high significant GWS modified search: "The results are * significant. ” First 20 resulting hits: 1. The results are statistically significant (13) 2. The results are highly significant (3) 3. The results are clinically significant (3) 4. The results are marginally significant (1) 5. The results are not statistically significant (1)

Literature Review Error Correction

Literature Review II Corpus Research

Literature Review, III (last one!) Technology and Feedback

Participants Large Midwestern U. Taking Graduate ESL Writing Course (level 2 of 4 possible) Focus on “the use of academic writing conventions” Placement in course via TOEFL, EPT, and Diagnostic Essay Variety of majors – Mostly science, computers, etc. 2 Courses Parallel Curriculum & Assignments – Control: 16 – Experimental: 13 In lab at least once/week Last paper of the semester: Persuasive Essay on a topic from their field of study

Methodology 1. Electronic versions of papers collected from each student. 2. Samples examined to determine most common error types. 3. All papers coded for error type by color. 4. Researchers cross-checked coding, any differences discussed and resolved. Nouns Verbs Articles Prepositions Adjectives Pronouns Adverbs General

Sample Coding

Procedure: Control Group – – – Intro to the Study/Human Subjects Forms (3 minutes) Discussion of the papers and markings (10 minutes) What do the colors mean? Sample paper & Key Handout Explanation of sections and subsections Intro to the Online Resources (15 minutes) Take ss. to online resources page: http: //randallsadler. googlepages. com Show them the various sections, and explain how they can use it with samples. Have students explore the Tools (10 minutes) Practice (15 minutes) Handout: What does the marking mean? (GIVE HANDOUT) What to do when revising Let’s work on the sample together Demo Cam. Studio (10 minutes) Setup (10 minutes)---(GIVE HANDOUT “WHAT TO DO NOW”) Save a new draft of your paper on the desktop Start Camstudio Open your paper Treatment (45 minutes)

Procedure: Experimental Group Intro to the Study/Human Subjects Forms (3 minutes) Discussion of the papers and markings (10 minutes) Intro to the Google Wildcard Searches (15 minutes) Practice (15 minutes) Demo Cam. Studio (10 minutes) Setup (10 minutes)---(GIVE HANDOUT “WHAT TO DO NOW”) Treatment (45 minutes) – What do the colors mean? Sample paper Key Handout Explanation of sections and subsections – Show sample Google Wildcard Searches – Explain How to use it. Give some sample phrases with errors – Have students explore the Tools (10 minutes) – Handout: What does the marking mean? (GIVE HANDOUT) – What to do when revising – Let’s work on the sample together – Save a new draft of your paper on the desktop – Start Camstudio – Open your paper

Data Analysis Data Sources Pre-revision papers Post-revision papers Researcher observation of revision process Movies of desktop showing revision process Survey

Post-Treatment, items (original errors) coded: + or – change – was there some change made to the identified error segment? + or – effective – Was there an effective change (result in a grammatically correct item)? + or – expected – Was it an expected change?

Results Control 16 students Total Sample: 7874 492 words per student Error Total: 522 Errors per student: 32. 6 Experimental 13 students Total Sample: 6789 522 words per student Error Total: 394 Errors per student: 30. 3

Pre-Treatment Errors by Category

Error Results by Percentage

Post-Treatment Changes Control Experimental 91. 9% of errors changed (480/522) 380 out of 480 items that WERE changed were effective 79. 2% 76. 2% of errors changed (275/361) 86. 9% of changed items were effective (239/275)

Changes by Category (%) Articles 98 76 Nouns 90 87 Verbs 91 73 General 91 59 Prepositions 94 79 Adjectives 80 84 Pronouns 93 71 Adverbs 83 54 0 10 20 30 40 Experimental 50 Control 60 70 80 90 100

% of changed and effective items Articles 88 Nouns 93 88 85 Verbs 73 83 General 57 85 Prepositions 73 Adjectives 89 75 95 Pronouns 92 40 Adverbs 60 0 10 20 30 40 Experimental 50 Control 60 86 70 80 90 100

Control Video 1

Control Video 2

Experimental Video 1

Experimental Video 2

Experimental Video 3

Problems & Future Research

Thanks for coming! Randall Sadler: rsadler@illinois. edu Ulugbek Nurmukhamedov: unurmuk 2@uiuc. edu