The Source of beauty ⅹ Patterns Interaction

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

describe people’s appearance by a simple present tense with an adjective. The common sentence pattern usually puts the adjective on a predicative position, such as he is tall, he is thin, and she is pretty. However, it seems that the teaching materials seldom put an emphasis on that adjectives may have different uses with different syntactic roles.

Some adjectives can not only pre-modify a noun (i.e., in an attributive position) but they can also function as a subject complement or an object complement by following copular verbs (i.e., in a predicative position). Specifically, the two target words in this study—pretty and charming can take both attributive and predicative position. Overall, the unfamiliarity of co-occurrence of the adjectives and nouns and the gap between the learning materials and the language use may raise the difficulties for the participants to make judgments on pretty in the sentences.

4.2.2 The Source of beauty ⅹ Patterns Interaction

In this section, in order to explore the effect of ‘source of beauty’ and ‘patterns’ on the participant’s judgements of the sentences in each group, a three-way ANOVA was performed. Again, as explained in the previous section (4.1.1), ‘source of beauty’

represents both ‘inner’ and ‘outer’ beauty; ‘patterns’ refers to the three patterns of collocates, namely, pretty-only, charming-only, and common. They are the controlled variables in the study.

The analysis of Group A. In order to examine Group A’s preferences for

describing person-related nouns, a 2 (‘inner’ and ‘outer’ beauty) ⅹ 3 (‘patterns’, namely pretty-only, charming-only, and common) mix-measured ANOVA was conducted.

From the analysis, the main effects in all the two variables were non-significant (‘source of beauty’ F (1,345) = 0.945, p=0.332, and ‘patterns’ F (2,345) = 2.004, p=0.136).

Whereas, an interaction effect between ‘patterns’ and ‘source of beauty’ was found significantly, F (2,345) = 3.5, p< .05, indicating that in terms of ‘source of beauty’, the Taiwanese EFL learners showed different preference for the three ‘patterns’ used to

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

describe someone’s attractiveness.

To interpret the interaction, it may be easier to understand by looking at the estimated marginal means of acceptability.

Figure 4-1 Group A’s Estimated Marginal Means of Acceptability of Sentences   Describing Person-Stimuli

From Figure 4-1, one could know that the effect of ‘patterns’ on the participant’s acceptability of sentences was dependent on the ‘source of beauty’. Under the ‘pretty-only pattern, the participants showed that using the adjective, pretty, to describe someone’s outer beauty was highly acceptable (M=3.512, SD=0.169), but under the

‘charming-only’ pattern, the effect was opposite, with the participants indicated the lowest degree of acceptability (M=3.318, SD=0.167) of the sentences using charming to describe someone’s outer beauty. On the other hand, ‘charming-only pattern’

(M=3.841, SD=0.167) was more acceptable than ‘pretty-only pattern’ (M=3.591, SD=0.167) when they were used to describe someone’s inner beauty.

However, the interaction effect did not indicate the differences between the six experimental conditions (i.e., ‘source of beauty’ ⅹ ‘patterns’ =2 ⅹ 3). We then

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

conducted a one-way ANOVA analysis³to compare the six experimental conditions by using the Scheffé post hoc test.

The result of Scheffé post hoc comparison indicates that the significant ‘source of beauty’ ⅹ ‘patterns’ interaction was primarily due to the participants having a higher degree of acceptance when someone’s ‘inner beauty’ was described by ‘charming-only’

pattern. The result corresponded to our findings from the corpus analysis. Furthermore, the participants significantly tended to use both pretty and charming to describe a person’s ‘outer beauty’ (i.e., voice in the ‘common pattern’); whereas, the sentences depicting someone’s ‘outer beauty’ under the ‘charming-only’ pattern were significantly less acceptable. These results may suggest that Taiwanese EFL learners (i.e., Group A) had a significant tendency of using the adjective charming to describe a person's inner beauty. However, they may be less familiar with the use of charming to describe a person’s outer beauty. Moreover, when describing a person’s outer beauty, the modified nouns which can co-occur with both pretty and charming were significantly strongly acceptable to the Taiwanese EFL learners.

The analysis of Group B. From the observations of the Group B’s use of the

adjectives, the interaction effect between ‘patterns’ and ‘source of beauty’ was not significant, F (2,352) = 0.292, p=0.747. In terms of the two main effects in the analysis, the ‘source of beauty’ was not significant (F (1,352) =0.854, p=0.356). However, different from Group A, the main effect of ‘patterns’ was significant, F (2,352) =7.693, p<.01, suggesting that the degree of acceptability of the descriptions of someone’s attractiveness may be affected by the ‘patterns’ of collocates. By inquiring the estimated marginal means, one could see that the uses of ‘charming-only’ pattern were

3  As Ho (2006, p. 64) also suggested that in a factorial design, post hoc comparisons between the experimental conditions cannot be conducted directly. The only way to make post hoc comparison is through one-way ANOVA. 

‧

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

significantly more acceptable (M=3.622, SD=0.118) than those of the ‘common’ pattern (M=3.264, SD=0.084), and ‘pretty-only’ pattern (M=2.967, SD=0.118). This indicates that the participants may be more familiar with the collocation use of ‘charming-only’

pattern; that is, charming smile and charming manner shown in the questionnaires. As earlier mentioned, we assumed that this language performance may also be related to the learner’s learning materials. It seems that most of the English teaching materials emphasize on describing a person’s appearance (i.e., ‘outer beauty’) by adopting a simple present tense with a predicative adjective, such as he is tall. This common use may cause a misunderstanding that utilize a predicative adjective to describe someone’s appearance is the most proper way of description.