Proficiency versus lexical processing efficiency as a measure of L2 lexical quality- Individual differences in word‑frequency effects in L2 visual word recognition

5분 분량

Introduction

Related Papers

Methods

Materials

Quote

  • A list of 30 high-frequency and 30 low-frequency English words was selected based on the English Lexicon Project(Balota et al., 2007). The SUBTLEX frequency for the high-frequency words ranged from 509 (baby) to 4,008 (right) per million, with an average of 929.1 (SD = 714.9). The low-frequency words had SUBTLEX frequency from 4 (crow) to 10 (skirt), with an average of 6.3 (SD = 1.3). The average number of letters per word was 4.7 (SD = 0.4) for both high- and low-frequency groups, and all words were either mono- or di-syllabic. We treated word frequency as a binary factor rather than a continuous one in order to obtain a clearer difference between high- and low-frequency words.

Quote

Since the corpus-based frequency of words may not necessarily apply to L2 learning contexts, we conducted a familiarity rating survey on the words to ensure that their frequency also represented their familiarity to L2 speakers. Fifty-three Korean learners of English, none of whom participated in the main task, were invited to rate how familiar they felt with each word on a 7-point scale (1: not at all familiar – 7: very familiar). Results of the survey indicated that the words with high SUBTLEX frequency received familiarity ratings of 6.7 on average (SD = 0.2), and those with low frequency received ratings of 5.0 on average (SD = 1.2). This difference was statistically significant (t = 7.7, p < .001), confirming that the frequency difference between the two sets of words was applicable to Korean-English L2 speakers’ English lexicon. The semantic concreteness and age of acquisition (AOA) of the target items were obtained from the English Lexicon Project. The average concreteness was 3.64 (SD = 1.19) and the average AOA was 5.72 (SD = 1.87). The two frequency groups did not significantly differ in concreteness (t = 0.33, p = 0.74) or AOA (t = 1.06, p = 0.30). A full list of target items and their frequency is given in the Appendix.

  • control condition propriority:
    • word-length, syllabic
    • Concreteness → 두 빈도 그룹 간 차이가 없게.
    • AoA → 두 빈도 그룹 간 차이가 없게.
  • 고, 저빈도를 나누는 기준으로 빈도수 직접 사용.

Results

Discussion



=== ORIGINAL 변환 ANOVA 결과 요약 ===
 
[HF 그룹]
  🔍 ANOVA 가정 검증
=== 정규성 검정 (Shapiro-Wilk test) ===
RFreq_KF: p-value = 0.0029 ❌ 정규분포 아님
CSAT_RFreq: p-value = 0.0045 ❌ 정규분포 아님
SUBTLRWF: p-value = 0.0000 ❌ 정규분포 아님
RFreq_HAL: p-value = 0.0018 ❌ 정규분포 아님
 
=== 등분산 검정 ===
Levene's test: p-value = 0.0000 ❌ 등분산 아님
Bartlett's test: p-value = 0.0000 ❌ 등분산 아님
 
=== 이상치 검출 ===
RFreq_KF: 2개 이상치 발견
CSAT_RFreq: 이상치 없음
SUBTLRWF: 2개 이상치 발견
RFreq_HAL: 1개 이상치 발견
 
=== 종합 결과 ===
모든 가정 만족: ❌ 아니오
  ANOVA 결과:
    F-statistic: 15.1271
    p-value: 0.0000
    eta-squared: 0.2812
    유의성: 유의함
  Tukey's HSD 결과:
    그룹 쌍별 비교:
      CSAT_RFreq vs RFreq_HAL: diff = 1264.873, CI[645.434,1884.312], p = 0.0000 *** (유의함)
      CSAT_RFreq vs RFreq_KF: diff = -204.162, CI[-823.601,415.277], p = 0.8258  (유의하지 않음)
      CSAT_RFreq vs SUBTLRWF: diff = 212.992, CI[-406.447,832.431], p = 0.8068  (유의하지 않음)
      RFreq_HAL vs RFreq_KF: diff = -1469.035, CI[-2088.474,-849.596], p = 0.0000 *** (유의함)
      RFreq_HAL vs SUBTLRWF: diff = -1051.880, CI[-1671.319,-432.442], p = 0.0001 *** (유의함)
      RFreq_KF vs SUBTLRWF: diff = 417.154, CI[-202.285,1036.593], p = 0.3002  (유의하지 않음)
 
 
[LF 그룹]
  🔍 ANOVA 가정 검증
=== 정규성 검정 (Shapiro-Wilk test) ===
RFreq_KF: p-value = 0.0001 ❌ 정규분포 아님
CSAT_RFreq: p-value = 0.0090 ❌ 정규분포 아님
SUBTLRWF: p-value = 0.0156 ❌ 정규분포 아님
RFreq_HAL: p-value = 0.0000 ❌ 정규분포 아님
 
=== 등분산 검정 ===
Levene's test: p-value = 0.0017 ❌ 등분산 아님
Bartlett's test: p-value = 0.0000 ❌ 등분산 아님
 
=== 이상치 검출 ===
RFreq_KF: 1개 이상치 발견
CSAT_RFreq: 1개 이상치 발견
SUBTLRWF: 1개 이상치 발견
RFreq_HAL: 2개 이상치 발견
 
=== 종합 결과 ===
모든 가정 만족: ❌ 아니오
  ANOVA 결과:
    F-statistic: 9.7054
    p-value: 0.0000
    eta-squared: 0.2078
    유의성: 유의함
  Tukey's HSD 결과:
    그룹 쌍별 비교:
      CSAT_RFreq vs RFreq_HAL: diff = 50.944, CI[18.151,83.737], p = 0.0005 *** (유의함)
      CSAT_RFreq vs RFreq_KF: diff = -1.602, CI[-34.657,31.453], p = 0.9993  (유의하지 않음)
      CSAT_RFreq vs SUBTLRWF: diff = -5.832, CI[-38.625,26.961], p = 0.9668  (유의하지 않음)
      RFreq_HAL vs RFreq_KF: diff = -52.546, CI[-84.418,-20.675], p = 0.0002 *** (유의함)
      RFreq_HAL vs SUBTLRWF: diff = -56.776, CI[-88.377,-25.176], p = 0.0000 *** (유의함)
      RFreq_KF vs SUBTLRWF: diff = -4.230, CI[-36.102,27.642], p = 0.9857  (유의하지 않음)