CHAPTER 6. Chunks in writing 95 TABLE 13. Summary of CAF measures Measure Description Complexity Tense use Guiraud Index Sentence length % of tenses other than the present tense / 10 # types / √ # tokens # of words in full sentence Accuracy Subject-Verb agreement Determiner-Noun agreement % of accurate SV forms % of accurate DN forms Fluency Text length total # of tokens in a text ANALYSIS Chunk coverage was calculated on an overall level, and for each individual chunk type. Data were inserted in R (R Studio Team, 2018; version 3.6.1) using the “readxl” package (Wickham & Bryan, 2019). Assumptions of normal distribution and homogeneity of variance were checked in R, using packages “pastecs” (Grosjean & Ibanez, 2018) and “car” (Fox & Weisberg, 2019), respectively. Due to a lack of normal distribution, a nonparametric version of an Independent Samples t-test, the Wilcox Signed Rank test, was used. Cohen’s (1988) d was calculated as an e ect size measure. In order to compare how each group used partially schematic and fully xed chunks, a Multivariate Analysis of Variance (MANOVA) was performed, p<.05. Partial eta squared was calculated as a measure of e ect size. To see if there were di erences in the seven types of chunks, assumptions were checked (Grosjean & Ibanez, 2018; Fox & Weisberg, 2019), and Independent Samples t-tests were conducted, p<.05. E ect sizes were also calculated, using “rstatix” (Kassambara, 2020) and “coin” (Hothorn et al., 2008), for cases where non-parametric t-tests were required. A er checking assumptions, several Pearson r correlation analyses were conducted, p<.05, in R to explore the relationship between chunk coverage on the one hand and Text Length, Average Sentence Length, Guiraud Index, Determiner-Noun Agreement, and Subject-Verb Agreement on the other hand. For each correlation, r was calculated as an e ect size, also in R.
RkJQdWJsaXNoZXIy MTk4NDMw