Accuracy across Proficiency Levels

Sommaire

General Introduction 19
Part I: The Concept of Error in Previous Error Analysis Work 25
1. Errors in early error analysis studies 27
1.1. Traditional error analysis 27
1.1.1. Gathering error samples 27
1.1.2. Error detection 35
1.1.2.1. Errors of competence vs. errors of performance 35
1.1.2.2. Overt vs. covert errors 37
1.1.2.3. Local vs. global errors 39
1.1.3. Error classification 40
1.1.4. Error counting 44
1.1.4.1. Obligatory occasion analysis 45
1.1.4.2. T-unit analysis 48
1.1.4.3. Using part-of-speech denominators 49
1.1.5. Error explanation 52
1.1.6. Error gravity 56
1.2. Concluding remarks 57
2. A new-look error analysis: the learner corpus approach 59
2.1. Learner corpora in error analysis 59
2.1.1. Treatment of the proficiency level factor 59
2.1.2. New methodological directions 64
2.2. Taking stock of CEA findings 74
2.2.1. Error rankings 75
2.2.2. Grammatical errors 83
2.2.3. Lexical errors 91
2.2.4. Orthographic errors 97
2.3. Concluding remarks 98
Part II: Rating Learner Performance,
Annotating and Counting Errors 101
3. Annotating and rating the International Corpus of
Learner English 103
3.1. Data: The International Corpus of Learner English 103
3.2. Error tagging the learner corpus sample 109
3.2.1. Detecting, correcting and tagging the errors in ICLE 109
3.2.2. The checking procedure 117
3.2.3. Results of the checking procedure 122
3.3. Rating the learner corpus sample 127
3.3.1. The raters and rating guidelines 127
8
Accuracy across Proficiency Levels
3.3.2. Rating results 133
3.3.2.1. Inter-rater reliability scores 134
3.3.2.2. Assigning a final CEFR score to each learner essay 139
3.4. Concluding remarks 142
4. Counting errors with potential occasion analysis 143
4.1. Potential occasion analysis: definition 143
4.2. Creating part-of-speech denominators 144
4.2.1. The part-of-speech tagger 144
4.2.2. The part-of-speech denominators 150
4.3. The error tag-denominator pairings 163
4.4. Interpreting potential occasion analysis results 169
4.5. Concluding remarks 170
Part III: Capturing Corpus-Based Developmental Patterns: Findings
for Second Language Acquisition and Language Testing Research 173
5. Capturing EFL accuracy developmental patterns 175
5.1. Statistical method: ANOVA 175
5.2. Three main error developmental patterns 179
5.2.1. Breaking down the strong developmental pattern 179
5.2.1.1. Strong pattern 1: B1>B2>C1>C2 180
5.2.1.2. Strong pattern 2: B1>[B2/C1/C2] 184
5.2.1.3 Strong pattern 3: B1>B2 and [non-adjacent levels] 189
5.2.1.4. Strong pattern 4: B2>B1 191
5.2.1.5. Strong pattern 5: [B1/B2]>[C1/C2] 193
5.2.2. Breaking down the weak error developmental patterns 194
5.2.2.1. Weak pattern 1: The [B1/C1]>[B2/C2]>[B1/C2] pattern 194
5.2.2.2. Weak pattern 2: The B1>C1 & B1>C2 and B2>C2 &
B1>C2 patterns 195
5.2.2.3. Weak pattern 3: The B1>C2 and B1>C1 patterns 197
5.2.3. The non-progressive error developmental pattern 199
5.3. Adding to findings on developmental second language
acquisition 203
5.3.1. Scope of the L2 features studied 205
5.3.2. L2 proficiency level establishment 206
5.3.3. L2 learning context 209
5.3.4. L2 developmental routes: fragmentary results 209
5.4. Concluding remarks 210
6. Working towards L1- and L2-dependent
proficiency descriptors 215
6.1. What the CEFR critics have to say 215
9
Table of Contents
6.2. Dissecting the CEFR descriptors for linguistic competence:
Layer 1 220
6.2.1. Where are the cannot do's? 221
6.2.2. Implied CEFR developmental patterns
for linguistic competence 225
6.3. Towards L1- and L2-dependent descriptor scales:
Layer 3 231
6.3.1. Grammatical accuracy snapshot 232
6.3.1.1. Article errors 233
6.3.2.2. Verb tense errors 235
6.3.2.3. Noun number agreement errors 238
6.3.2.4. Uncountable noun errors 240
6.3.2.5. Adverb placement errors 241
6.3.2.6. Dependent preposition errors 243
6.3.2.7. Determiner and pronoun errors 244
6.3.2. Vocabulary control snapshot 250
6.3.3. Orthographic control snapshot 258
6.3.4. Punctuation control snapshot 261
6.4. Reconsidering the six-level proficiency scale 265
6.5. Concluding remarks 267
General conclusion 269
References 279
Appendices 303
Subject index
Author index