3.2 Methods and Materials
3.2.3 Four Tasks
3.2.3.1 Phase 1
The purpose of this phase was to investigate the subjects’ interpretations of the seven frequency adverbs, from the most frequent to the least frequent. An identification task and an ordering task were designed. Two stories were created with seven characters, most of whom had the same habits but exhibited the habits at different frequencies. The first story was about the characters’ habits of being late while the second story described the characters’ habits of forgetting to do their homework. Each character’s habit was described with one of the seven frequency adverbs. The distribution of the frequency adverbs is shown in Table 3-3.
54
Table 3-3 The Distribution of the Test Items Used in the Two Tasks in Phase 1
Type Level Task Number of
In the identification task, the subjects were told the habits of each character and they were asked to report how many days each character exhibited the habit in a week by giving checks in the circles. A circle meant the character did the habit one day in a week. Table 3-4 demonstrates an example.
55
Table 3-4 An Example of the Identification Task in Phase 1
The subject saw The subjects heard
Characters
Zhe qitian yilai, Xiaoming hen guai, jihubu hui chidao, laoshi jeude hen bang.
‘Xiaoming is great. He is rarely late for school for the past seven days. His teacher thinks he is excellent.’
Characters Circles for the subjects to put checks
Xiaopengyou, ni juede yinggai da ji ge gou ne?
‘Kids, how many checks do you need to give?’
After finishing reporting each character’s habits in a week, they were told to help a teacher decide how to deliver presents to praise the characters according to their behaviors. The present of the largest size was the best present because there was much stuff inside, as demonstrated in Table 3-5.
56
Table 3-5 Examples of the Ordering Task in Phase 1
The subject saw The subjects heard
Xiaopengyou, qing genju gougou de shuliang, ba mei ge liwu lian dao gai na dao ta de xuesheng.
‘Kids, please match the presents with the corresponding students according to their habits.’
In this ordering task, the subjects’ ordering of frequency adverbs was revealed. The subjects listened to the first story and finished the two tasks before moving on to the second story. The scripts of the two stories are provided in Appendix A, and the worksheet is in Appendix B.
3.2.3.2 Phase 2
The purpose of this phase was to examine the subjects’ selection of events that can co-occur with frequency adverbs. Contextual effect was also explored in this phase;
therefore, two tasks were included—a context-free acceptability judgment task and a context-provided acceptability judgment task. Typical frequency adverbs—
changchang ‘usually,’ youshihou ‘sometimes’ or ouer ‘occasionally’—were selected to
match with the event types (cf. Zou 2007:31-32).
57
In the context-free acceptability judgment task, only the sentences that included frequency adverbs were read to the subjects. In the context-provided acceptability judgment task, on the other hand, three sentences including an endpoint were added before the sentence that had a frequency adverb to make a clear context.
These two acceptability judgment tasks were designed on a three-point scale.
When doing the questions, the subjects needed to decide which circle they should put a check in. A check in the green light meant that the sentence was perfect/acceptable in Mandarin; a check in the red light indicated that the sentence was not acceptable; a check in the yellow light meant that the sentence was a little bit weird but still acceptable. The three-scale point is exemplified in Table 3-6.
Table 3-6 The Three-point Scale in Phase 2
If the subject thought the sentence of Question 1 was perfect in Mandarin, he/she should put a circle in the green light.
Both tasks included ten questions. The distribution of the questions is shown in Table 3-7. The scripts of the tasks are provided in Appendices C and D, and a worksheet is provided in Appendix E.
58
Table 3-7 The Distribution of the Questions in Phase 2
Type Task Number of
questions Questions
ET1-1 Context-free 2 Q2, Q10
Context-provided 2 Q1, Q8
ET1-2 Context-free 2 Q1, Q6
Context-provided 2 Q4, Q9
ET1-3 Context-free 2 Q4, Q7
Context-provided 2 Q2, Q7
ET2-1 Context-free 2 Q3, Q9
Context-provided 2 Q5, Q6
ET2-2 Context-free 2 Q5, Q8
Context-provided 2 Q3, Q10
Total 20
The free acceptability judgment task was conducted followed by the context-provided acceptability judgment task. Tables 3-8 and 3-9 present example questions.
59
Table 3-8 An Example Question in the Context-free Task
The subjects saw The subjects heard
Mama zhe ji tian changchang bu
xiaoxin da po dongxi.
‘Mom usually broke something into pieces accidentally in the past days.’
Qing wen keyi zheme shuo ma?
‘Can we say a sentence like this one?’
60
Table 3-9 An Example Question in the Context-provided Task
The visual aids the subjects saw The sentences the subjects heard
Mama mei tian dou hui daisao jia li, dan
ta dongzuo hen da.
‘Mom cleans the house every day, but she is very rude.’
Shang libai bu xiaoxin da po huaping han panzi.
‘She broke a vase and a plate accidentally last week.’
Qiantian han zuotian you bu
xiaoxin shuai po beizi.
‘These two days, she broke glasses accidentally.’
61
Suoyi mama zhe ji tian changchang
bu xiaoxin da po dongxi.
‘So, Mom usually broke something into pieces accidentally in the past days.’
Qing wen keyi zheme shuo ma?
‘Can we say a sentence like this one?’
As shown in Tables 3-8 and 3-9, the last sentence with the last slide in the context-provided task was identical with those of the context-free task. The difference between the two tasks was that three sentences along with three slides were added in the context-provided task.
3.3 Procedures
3.3.1 Formal Study
The subjects recruited for the formal study were kindergartners and elementary school students in an elementary school in Taipei and native Chinese adult speakers.
62
Before the experiment, all subjects were given a consent form. The child subjects’
consent forms sought their parents’ agreement on allowing their children to participate in the experiment. The consent form is provided in Appendix F.
Two phases of experiment were conducted. In Phase 1, the identification task was conducted before the ordering task. The subjects finished these two tasks of Story 1 first and then moved on to those of Story 2. In Phase 2, the subjects did the context-free acceptability judgment task and then the context-provided acceptability judgment task.
3.3.2 Data Analysis
In Phase 1, a binary code was conducted as the scoring policy. The subjects’
responses towards the ordering of the seven frequency adverbs were compared to the revised classification of the seven frequency adverbs (i.e., Table 3-1). When a subject’s ordering of a certain frequency adverb was identical with that of the revised classification, it was counted as 1. Otherwise, it was 0. In Phase 2, the three-point scale was used. A check in the red light was counted as 0; the yellow light, 1; the green light, 2. The mean score of each event type was counted group by group. If the mean score was close to 2, it indicated that the event type co-occurring with frequency adverbs was highly acceptable for the subjects. On the contrary, if the mean score was close to 0, it meant that the subjects tended to reject the event type with frequency adverbs. All the data were entered into the computer and processed by R.
63
3.4 Summary of Chapter Three
This chapter introduces the background information of the subjects, the task designs, the procedures and the data analysis of the present study. A total of one hundred participants were recruited: eighty children (the kindergartners, Grades 2, 4 and 6) and twenty adults. The identification and ordering tasks in Phase 1 elicited the subjects’
ordering of the seven frequency adverbs. In Phase 2, the free and context-provided acceptability judgment tasks measured the subjects’ acceptability ratings of frequency adverbs when they modified events. The subjects’ answers were examined by statistical analysis to look into children’s acquisition of Chinese frequency adverbs.
64
Chapter Four Results and Discussion
This chapter reports and discusses the results of the present study. Section 4.1 examines the differences of the subjects’ interpretations of typical and atypical frequency adverbs. Section 4.2 discusses the subjects’ acceptability of frequency adverbs when they were used to modify telic and atelic events. Section 4.3 addresses whether frequency adverbs modifying the subtypes of telic events displayed different degrees of acceptability. Section 4.4 explores the effect affected by contextual clues and Section 4.5 reports the age difference. Finally, a summary of this chapter is recapitulated in Section 4.6.
4.1 Typical and Atypical Frequency Adverbs
The first research question of the study addresses a difficulty order of typical and atypical frequency adverbs. Section 4.1.1 reports the subjects’ interpretations of typical and atypical frequency adverbs. Section 4.1.2 provides a second look at the subjects’
interpretations of typical frequency adverbs. Section 4.1.3 further discusses the difficulty order of typical and atypical frequency adverbs and at what age children have the ability to compare various degrees of a set of related words.
65
4.1.1 Overall Findings
As shown in Table 4-1, the subjects interpreted atypical frequency adverbs (M = 0.81) better than typical ones (M = 0.50), and the overall scores of the two types of frequency adverbs were significantly different (F (1,198) = 73.76, p < .001).
Table 4-1 Subjects’ Overall Performance on the Interpretations of Typical and Atypical Frequency Adverbs
Type M SD F p-value
Typical 0.50 0.26 73.76 2.61e-15 ***
Atypical 0.81 0.25
Note: *** indicates that the p-value is smaller than .001.
The mean scores of each group for the typical and atypical frequency adverbs are provided in Figure 4-1.
Figure 4-1 Overall Performance of Each Group on the Interpretations of Typical and Atypical Frequency Adverbs
66
One-way ANOVA indicated that there was a significant difference among the five groups in interpreting typical and atypical frequency adverbs (Typical: F (4,95) = 16.25, p < .001; Atypical: F (4,95) = 40.17, p < .001). All the groups interpreted atypical frequency adverbs better than typical ones (Kindergartner: M = 0.46 > 0.33; Grade 2:
M = 0.71 > 0.36; Grade 4: M = 0.93 > 0.47; Grade 6: M = 0.94 > 0.59; Control: M = 1.00 > 0.77). However, the result that even the control group received a low grade in the typical type of frequency adverbs deserves our attention. To explore underlying reason for this result, we examined the ordering of the frequency adverbs in typical types.
4.1.2 A Second Look at Typical Frequency Adverbs
On account of the low grade (M = 0.77) the control group received, a second look at typical frequency adverbs was conducted. A closer examination revealed that some adult controls put youshihou ‘sometimes’ and ouer ‘occasionally’ in a reverse order.
According to Wu and Ding (2005), youshi(hou) ‘sometimes’ indicates more frequent than ouer ‘occasionally,’ so youshihou ‘sometimes’ is placed higher on the frequency scale than ouer ‘occasionally.’ However, the dictionary of the fifth version provided by the Ministry of Education, R.O.C. (http://dict.revised.moe.edu.tw/cbdic/index.html) shows that youshi(hou) ‘sometimes’ and ouer ‘occasionally’ are synonyms. Since they
67
are synonyms in the modern era, they indicate the same degree of frequency. As a result, it is improper to conclude that the control group failed to put youshihou ‘sometimes’ in a higher degree than ouer ‘occasionally.’ Rather, the classification of frequency adverbs in Wu and Ding (2005) should be modified to fit the modern language use, as presented in Table 4-2.
Table 4-2 A Second Revised Classification of Frequency Adverbs Original Classification Revised Classification Examples
L1 L1 zongshi ‘always’
L2 L2 changchang ‘usually’
L3
L3
youshihou ‘sometimes’
L4 ouer ‘occasionally’
L5 L4 henshao ‘seldom’
L6 L5 jihubu ‘rarely’
L7 L6 congbu ‘never’
The subjects’ overall performance on the interpretations of the two types of frequency adverbs is revised accordingly, as presented in Table 4-3 and Figure 4-2.
68
Table 4-3 Revised Subjects’ Overall Performance on the Interpretations of Typical and Atypical Frequency Adverbs
Type M SD F p-value
Typical 0.70 0.22 9.59 0.00224 **
Atypical 0.81 0.25
Note: ** indicates that the p-value is smaller than .01.
Figure 4-2 A Revised Comparison between Typical and Atypical Frequency Adverbs
The revised overall scores for the two types of frequency adverbs were still significantly different (F (1,198) = 9.59, p < .01). There was also a significant difference among the five groups in interpreting typical frequency adverbs (F (4,95) = 18.64, p < .001).
Furthermore, as indicated in Figures 4-1 and 4-2, each group’s mean score for typical frequency adverbs was raised. As for each age group’s interpretations of the two types of frequency adverbs, the within-group differences are presented in Table 4-4.
69
Table 4-4 The p-values for the within-group Differences between the Interpretations of Typical and Atypical Frequency Adverbs
Kindergartner Grade 2 Grade 4 Grade 6 Control
F 1.17 3.51 20.29 16.62 19
p .29 .07 .000*** .000*** .000***
Note: *** indicates that the p-value is smaller than .001.
A significant difference was found in Grade 4 (F (1,38) = 20.29, p < .001), Grade 6 (F (1,38) = 16.62, p < .001) and the adult controls (F (1,38) = 19, p < .001). The result indicated that typical frequency adverbs were more difficult for the older children (Grades 4 and 6) and the adult controls to interpret than atypical ones. For the younger children (the kindergartners and Grade 2), differentiating the degrees in both types was a challenging task.
Regarding the between-group differences, Table 4-5 shows the p-values for the differences in the interpretations of the two types of frequency adverbs.
Table 4-5 The p-values for the between-group Differences in Typical and Atypical Frequency Adverbs
Grade 2 Grade 4 Grade 6 Control
Typical Atypical Typical Atypical Typical Atypical Typical Atypical
Kindergartner .57 *** ** *** *** *** *** ***
Grade 2 - - .07 *** ** *** *** ***
Grade 4 - - - - .57 .80 *** .40
Grade 6 - - - ** .42
Note: ** indicates that the p-value is smaller than .01 and *** is smaller than .001.
70
In the typical type of frequency adverbs, Grades 4 and 6 significantly outperformed the kindergartners (p < .001, p < .01, respectively), but Grade 2 did not perform significantly differently from the kindergartners (p > .05). Grade 6 significantly outperformed Grade 2 (p < .01), but the performance of Grade 6 was not significantly different from that of Grade 4 (p > .05). Grade 4 did not significantly outperformed Grade 2 (p > .05), either. However, the adult controls performed significantly better than the kindergartners, Grade 2, Grade 4 (p < .001) and Grade 6 (p < .01). In the atypical type of frequency adverbs, Grade 2 significantly performed better than the kindergartners (p < .001), and Grade 4 significantly performed better than Grade 2 (p
<. 001) and the kindergartners (p < .001). Grade 6 significantly performed better than Grade 2 and the kindergartners (p < .001) but not significantly differently from Grade 4 (p > .05). The adult controls significantly outperformed the kindergartners and Grade 2 (p < .001), but not Grades 4 and 6 (p > .05).
4.1.3 General Discussion
The major findings revealed that acquiring typical frequency adverbs was more challenging than atypical ones. Figure 4-2 shows the subjects’ outperformance on the atypical type over the typical one across all age groups except for the kindergartners.
Atypical frequency adverbs show less vagueness than typical frequency adverbs, and
71
the meanings of atypical frequency adverbs were more precise (Bocklisch et al. 2012).
For instance, zongshi ‘always’ indicates that the event happens all the time and congbu
‘never’ indicates that the event does not happen. On the other hand, typical frequency adverbs share the notable feature of vagueness (Zou 2006, Piao 2011). Due to the fact that their meanings are indefinite and unclear, they posed greater difficulties for children. This result corresponds to the argument of Hoffner et al. (1990), who argue that children acquire absolutely positives and negatives before they understand the meanings of inner members in an ordered semantic set. In addition, the kindergartners’
performance on both typical and atypical frequency adverbs was significantly different from that of the adult controls. The results indicated that the two types of frequency adverbs were challenging for the kindergartners.
Regarding the between-group comparison, the older children (Grades 4 and 6) interpreted atypical frequency adverbs similarly with the adult controls, and the older children performed significantly better than the younger children (the kindergartners and Grade 2). The findings indicated that by the time the children reached the fourth grade, they were capable of interpreting atypical frequency adverbs like the adults did.
However, for typical frequency adverbs, even Grade 6 had not yet acquired adult-like interpretations. This finding demonstrated the feature of ambiguity of typical frequency adverbs (Zou 2006, Piao 2011).
72
An age-related increase in children’s ability to discriminate among these frequency adverbs was found, which is consistent with previous findings (e.g., Cohen and Hansel 1956, Kuczaj 1975). Moreover, the results showed that the acquisition denoting various degrees in a dimension starts at around the age of ten (Grade 4), which is in accordance with Hoffner et al. (1990). According to their argument, children do not acquire the meanings of words until they have developed the necessary cognitive skills to understand the underlying concepts (e.g., Clark and Clark 1977, Kuczaj 1982).
Therefore, children’s acquisition of frequency adverbs relates to children’s ability to distinguish among various degrees. Piaget and Inhelder (1975) concluded that children do not have the cognitive ability to compare different degrees of probability until they become adolescents. Similarly, frequency adverbs denote different degrees in a dimension, so children may not fully interpret frequency adverbs until they become adolescents. Moreover, in Piaget’s theory of cognitive development, eleven-year-olds enter the formal operation stage and start to develop the ability to think about abstract concepts (Piaget 1964). Frequency adverbs are abstract concepts so children older than eleven years old may start to have the ability to interpret frequency adverbs. Also, this later acquisition may attribute to the need of more than one intelligent ability. Gardner (1983, 1999, 2011) identifies several intelligences, including linguistic, logic-mathematical, musical, spatial, bodily/kinesthetic, interpersonal, intrapersonal
73
naturalistic and spiritual intelligences in his theory of multiple intelligences. Frequency adverbs are linguistic terms so people need to rely on their linguistic intelligence to interpret frequency adverbs. But when people are asked to compare different degrees of a set of related words, logic-mathematical intelligence is required as well because the comparison of various degrees is related to mathematics. Therefore, they should have developed these two intelligences before they have the ability to interpret frequency adverbs accurately. That is the reason that the development of comparing different degrees of a set of related words starts at such a late age.
4.2 Telic and Atelic Events
The second research question concerns children’s acceptability of frequency adverbs when they are used to modify telic and atelic events. Section 4.2.1 reports the subjects’ overall acceptability rates of frequency adverbs used to modify telic and atelic events. Section 4.2.2 further discusses the subjects’ acceptability of frequency adverbs in telic and atelic events.
4.2.1 Overall Findings
The overall scores of telic and atelic events are presented in Table 4-61. The results
1 The scores for the red light were translated into 0; the yellow light into 1; the green light into 2.
74
indicated that frequency adverbs modifying telic events (M = 1.35) were more acceptable than those modifying atelic events (M = 0.66), and frequency adverbs modifying the two event types exhibited a significant difference (F (1,1998) = 323.1, p
< .001).
Table 4-6 Subjects’ Overall Acceptability of Frequency Adverbs Modifying Telic and Atelic Events
Event Type M SD F p-value
Telic (ET1) 1.35 0.84 323.1 2e-16 ***
Atelic (ET2) 0.66 0.84
Note: *** indicates that the p-value is smaller than .001.
Figure 4-3 presents each group’s acceptability rates of frequency adverbs used to modify the two event types.
Figure 4-3 Each Group’s Acceptability of Frequency Adverbs Modifying Telic and Atelic Events
Kindergartner Grade 2 Grade 4 Grade 6 Control
Telic 1.08 1.32 1.38 1.41 1.55
Telic SD 0.92 0.82 0.77 0.86 0.75
Atelic 1.15 0.97 0.41 0.36 0.28
Atelic SD 0.83 0.87 0.76 0.72 0.61
0
75
As shown above, the older the subjects were, the higher ratings they gave to frequency adverbs when they modified the telic events (Control: 1.55 > Grade 6: 1.41 > Grade 4:
1.38 > Grade 2: 1.32 > Kindergartner: 1.08). Similarly, the older subjects gave lower ratings to frequency adverbs when they modified atelic events (Control: 0.28 < Grade 6: 0.36 < Grade 4: 0.41 < Grade 2: 0.97 < Kindergartner: 1.15). In other words, the subjects’ acceptability of frequency adverbs modifying the two event types varied and increased with age. Regarding each age group’s acceptability of frequency adverbs when they modified different event types, the within-group differences are presented in Table 4-7.
Table 4-7 The p-values for the within-group Differences between the Acceptability of Frequency Adverbs Modifying Telic and Atelic Events
Kindergartner Grade 2 Grade 4 Grade 6 Control
F .61 16.59 157.7 161.4 414.9
p .44 .000*** .000*** .000*** .000***
Note: *** indicates that the p-value is smaller than .001.
Note: *** indicates that the p-value is smaller than .001.