Differences in Response Bias Among Younger and Older Preschoolers

Abstract

Groups of children in Japan and Hungary aged 2 to 5 years were asked yes-no questions pertaining to familiar and unfamiliar objects by either strange adults or mothers. In Experiment 1, 3- to 5-year-old Japanese and Hungarian children were interviewed by strange adults, and 3-year-olds in both countries exhibited a yes bias for familiar objects. Japanese 3-year-olds exhibited a yes bias for unfamiliar objects as well. Japanese 4- and 5-year-olds exhibited a yes bias for familiar objects, whereas Hungarian 4- and 5-year-olds exhibited a nay-saying bias for unfamiliar objects. In Experiment 2, Hungarian 2-year-olds exhibited a yes bias for both familiar and unfamiliar objects when asked by strange adults. In Experiment 3, Japanese 3-year-olds exhibited a yes bias for both familiar and unfamiliar objects, whereas 4-year-olds did not exhibit it for familiar objects and exhibited a nay-saying bias for unfamiliar objects when they were asked by mothers. The results suggest that young preschoolers have a common mechanism for exhibiting a yes bias, but older preschoolers may have other reasons to exhibit response biases.

Keywords

language development pragmatics response bias yes bias

Adults question children not only in day-to-day situations but also in formal situations such as forensic interviews and experimental settings in developmental psychology to obtain what children witness, think, or know. Fritzley and Lee (2003) pointed out that developmental psychologists have often used yes-no questions in experiments; however, such questions are not suitable for young preschoolers because they can be misled by these types of questions. That is, children, mostly young preschoolers, often show a tendency to say “yes” (i.e., a yes bias) in response to yes-no questions.

Recent evidence indicates that Canadian 2-year-olds exhibit a consistent yes bias to comprehensible and incomprehensible yes-no questions pertaining to the properties and functions of familiar and unfamiliar objects, while Canadian 3-year-olds exhibit a yes bias in some conditions (Fritzley & Lee, 2003). In their study, the comprehensible questions were “Is this X?” forms of questions, and the incomprehensible questions included nonsense words (e.g., “Is this darh”). Okanda and Itakura (2008) reported that Japanese and Vietnamese 2-year-olds exhibited a yes bias to comprehensible yes-no questions pertaining to familiar and unfamiliar objects, while 3-year-olds exhibited it for familiar objects and tended to exhibit it for unfamiliar objects. Okanda and Itakura (2007) confirmed that Japanese 2- and 3-year-olds exhibited a yes bias to the same type of questions when the questions were asked by the children’s mothers. Moreover, Okanda and Itakura (2010b) reported that 3-year-olds exhibited a yes bias to comprehensible yes-no questions regarding knowledge of familiar objects, preference for familiar and unfamiliar objects, and facial expressions.

Moriguchi, Okanda, and Itakura (2008) found that 3- to 4-year-old children’s performance on a yes bias task that asked children yes-no questions pertaining to objects was correlated to their performance on a Dimensional Card Sorting (DCCS) task and a Picture Vocabulary Test (PVT; Ueno, Utsuo, & Iinaga, 1991). The DCCS examines inhibitory control ability (e.g., Kirkham, Cruess, & Diamond, 2003; Moriguchi, Lee, & Itakura, 2007; Zelazo, Frye, & Rapus, 1996), and the PVT is a Japanese version of a Peabody PVT (Dunn & Dunn, 1981) that correlates significantly with more general language skills (Carvajal, Parks, Logan, & Page, 1992; Hodapp, 1993). The children who did not perform well on the DCCS task and the PVT exhibited a stronger yes bias than the children who performed well on these two tasks. Okanda and Itakura (2010b) pointed out that younger preschoolers may exhibit a yes bias automatically due to underdeveloped inhibitory control abilities and language skills: They may not be able to inhibit a dominant “yes” response because they often say “yes” in response to adults’ questions in their daily life (see also Moriguchi et al., 2008; Okanda & Itakura, 2007). In addition, they may not understand the pragmatic context of questions to give appropriate answers. In fact, there were no noticeable cross-cultural differences on young preschoolers’ yes bias in Canada (Fritzley & Lee, 2003), Japan, and Vietnam (Okanda & Itakura, 2008).

In contrast to the consistent results found in younger preschoolers, older preschoolers exhibited various response tendencies (a yes bias, a nay-saying bias, or no bias) to yes-no questions under different conditions, and there were some cross-cultural differences. We use the term nay-saying bias in this article because the term no bias indicates a lack of response bias (see also Fritzley & Lee, 2003). Okanda and Itakura (2010b) reported that Japanese 6-year-olds tended to exhibit a nay-saying bias to comprehensible yes-no questions pertaining to positive and negative facial expressions (e.g., “Is this happy?”). Canadian 4- and 5-year-olds exhibited a nay-saying bias to incomprehensible yes-no questions with nonsense words pertaining to object knowledge, but they did not exhibit any response biases to comprehensible questions with regular words (Fritzley & Lee, 2003). By contrast, Japanese older preschoolers sometimes exhibited a yes bias to comprehensible yes-no questions pertaining to objects: Japanese and Vietnamese 4-year-olds exhibited a yes bias to questions pertaining to familiar objects (Okanda & Itakura, 2008). Japanese 4-year-olds in Okanda and Itakura’s (2010b) study tended to exhibit it to both familiar and unfamiliar objects in Study 1, but did not exhibit it to familiar objects in Study 2. Japanese 5- and 6-year-olds in Okanda and Itakura’s (2010b) study exhibited a yes bias, but their response bias scores were significantly lower than the scores of 3-year-olds, and these results were only found in Study 1, but not in Study 2.

In addition, there is evidence that 5- to 7-year-olds (Hughes & Grieve, 1980; Pratt, 1990) and adults (Pratt, 1990) try to answer “yes” or “no” in response to bizarre yes-no questions such as “Is milk bigger than water?” and “Is a fork happier than a knife?” Similarly, 5- to 9-year-olds and adults tried to answer unanswerable yes-no questions involving picture stories which they had not been told or had not seen (Waterman, Blades, & Spencer, 2001a).

Generally, 4- to 5-year-old children are able to pass the DCCS task (e.g., Moriguchi et al., 2007; Zelazo et al., 1996), and older preschoolers (4;9–6;0) understand others’ utterances better than younger preschoolers (3;6–4;6) (Siegal, Iozzi, & Surian, 2009). Therefore, older preschoolers should not exhibit a response bias due to the underdeveloped cognitive abilities as younger preschoolers do. There should be different reasons for older preschoolers to make “yes” or “no” decisions (Okanda & Itakura, 2010b).

Okanda and Itakura (2010b) suggested an alternative explanation, namely that older preschoolers may give yes-no answers depending on the social context in which they are being asked the questions. We assume that some social reasons should be partly influenced by cultural factors. For example, Okanda and Itakura (2010b) found that older Japanese children exhibited a tendency to say “yes” to experimenters’ comprehensible yes-no questions pertaining to objects. They indicated that the children probably responded this way because they believed it is a polite response in their culture. We assume that Japanese older preschoolers sometimes exhibit a yes bias to strange adults because of social pressure; therefore, they should not exhibit it to other interviewers such as their own mothers. There is a finding that 2- and 3-year-olds exhibited a yes bias to their mothers (Okanda & Itakura, 2007); however, this has not been studied in older preschoolers.

Moreover, Canadian 4- and 5-year-olds children did not exhibit response biases to comprehensible questions (Fritzley & Lee, 2003). We assume that children in Western cultures are facilitated to say “no” as well as to say “yes.” However, previous studies only investigated a limited number of countries such as Canada, Japan, and Vietnam (Fritzley & Lee, 2003; Okanda & Itakura, 2008, 2010b); therefore, it is still unknown whether older preschoolers in different countries show different response tendencies to yes-no questions.

There are two issues that we investigated in this study. First, we wanted to confirm whether previous results that younger preschoolers in various countries exhibit a yes bias is a robust result. We also wanted to examine response tendencies of older preschoolers in other countries. Second, we wanted to confirm whether older preschoolers change their response tendencies depending on the status of the interviewer. In Experiment 1, we asked a group of 3- to 5-year-old Japanese and Hungarian children yes-no questions pertaining to the properties and functions of certain objects and then compared their performances. The previous studies were conducted in two Eastern countries (i.e., Japan and Vietnam; e.g., Okanda & Itakura, 2008, 2010b), but there was only one study in Western countries (i.e., Canada; Fritzley & Lee, 2003). Therefore, Hungary can be another representative of a Western country. In Experiment 2, we determined whether 2-year-old Hungarian children exhibit a consistent yes bias regardless of conditions (i.e., familiar objects vs. unfamiliar objects), as do children in other countries. It is hypothesized that if younger preschoolers exhibit a yes bias automatically, Hungarian (younger) preschoolers should exhibit a similar yes bias. It is also hypothesized that if older preschoolers make their binary decision based on social (or cultural) reasons, then older Hungarian preschoolers should show response tendencies that differ from children in other countries. In Experiment 3, related to the second hypothesis, 3- and 4-year-old Japanese children were asked the same type of yes-no questions by their mothers. If Japanese children say “yes” to strange adults to show their politeness, they should not say “yes” to their mother, who is the most familiar adult.

Experiment 1

In Experiment 1, 3- to 5-year-old Japanese and Hungarian children were asked yes-no questions pertaining to the properties and functions of six objects. This experiment was conducted to investigate whether a yes bias would be a universal phenomenon that could be observed in young preschoolers across different countries. We also compared developmental patterns for the yes bias in Japanese and Hungarian children.

Method

Participants

The participants were 138 Japanese and Hungarian children (Table 1). Specifically, 20 three-year-old (M = 42.7 months, SD = 2.74, range = 3;2~3;11; 12 boys, 8 girls), 26 four-year-old (M = 54.2 months, SD = 3.10, range = 4;0~4;11; 15 boys, 11 girls), and 20 five-year-old (M = 68.7 months, SD = 4.49, range = 5;0~6;4; 16 boys, 4 girls) Japanese children participated. Furthermore, 27 three-year-old (M = 41.9 months, SD = 3.51, range = 3;0~3;11; 14 boys, 13 girls), 21 four-year-old (M = 53.1 months, SD = 3.95, range = 4;0~4;11; 8 boys, 13 girls), and 24 five-year-old (M = 67.7 months, SD = 5.21, range = 5;0~6;6; 11 boys, 12 girls) Hungarian children participated. Five additional children (four Hungarian and one Japanese) were tested but were excluded because they refused to complete the experiment.

Table 1.

Sample Descriptions in Experiment 1

	Japanese			Hungarian
	Number	Mean age	Range	Number	Mean age	Range
3-year-olds	20	42.7	3;2–3;11	27	41.9	3;0–3;11
4-year-olds	26	54.2	4;0–4;11	21	53.1	4;0–4;11
5-year-olds	20	68.7	5;0–6;4	24	67.7	5;0–6;6

The children were recruited from kindergarten and nursery schools in Kyoto and Ashiya, Japan, and Budapest, Hungary. Experimenters explained the purpose and method to the directors of the kindergarten and nursery schools and/or to the mothers before the experiments.

Materials and procedure

We followed Fritzley and Lee’s (2003) procedure and used the same materials that were used by Okanda and Itakura (2008). Three familiar and three unfamiliar objects were used. A blue cup, a red apple, and a picture book were the familiar objects, and a plastic coffee filter, a shoehorn, and a computer chip (CPU) were the unfamiliar objects. The experimenter asked ten 3- to 5-year-old Hungarian children about the names and functions of the objects and confirmed that the children could provide the names and functions of the familiar objects but not the unfamiliar ones. Japanese children had already been tested previously by Okanda and Itakura (2008), and similar results were obtained.

The children participated in the study individually. Each child was questioned in his or her native language. Each child was presented with an object and was asked four questions about each object’s properties and functions. Children were not required to be familiar with every object’s name and function to answer all the questions correctly because some questions were easy to answer. For example, they were asked about the colors and shapes of the objects (see Appendix A for exact wording of the questions).

Approximately half of the children were asked about a familiar object first, and the other half were asked about an unfamiliar object first. Two of the four questions were yes questions (the correct answers were “yes”) and the other two were no questions (the correct answers were “no”). The experimenter avoided asking three yes questions or three no questions in sequence.

The children’s responses, including nonverbal responses (nodding and head shaking), were recorded online on an answer sheet by the experimenter. We modified Fritzley and Lee’s (2003) scoring method to reveal children’s response bias scores and calculated the proportion of correct and incorrect responses. First, for the yes questions, each child received a score of 1 for every “yes” response (i.e., correct response) and a score of −1 for every “no” response (i.e., incorrect response), and the sum of these scores was considered as the yes score. Similarly, for the no questions, the children received a score of 1 for all “no” responses (i.e., correct responses) and a score of −1 for all “yes” responses (i.e., incorrect responses), and the sum of these scores was considered as the no score. Second, the yes score was divided by the total number of yes questions to obtain a proportional yes score. A proportional no score was obtained in a similar manner. Third, the proportional no score was subtracted from the proportional yes score, and the mean response bias score of the child was obtained. The maximum response bias score was 1, and the minimum score was −1. A positive response bias score indicated a yes bias, and a negative response bias score indicated a nay-saying bias. A child who did not exhibit any response bias had a zero response bias score. In addition, responses like “I don’t know” and “no answer” were not scored but were counted separately.

Results and Discussions

Figure 1 shows Japanese and Hungarian children’s mean response bias scores for familiar and unfamiliar objects. A 3 (Age) × 2 (Country) × 2 (Familiarity) ANOVA was conducted on the children’s mean response bias scores. The main effect of age was significant, F(2, 132) = 7.50, p < .01, η² = .10, and post hoc Bonferroni tests revealed that the 3-year-olds exhibited a stronger yes bias than the 4- (p < .05) and 5-year-olds (p < .01). The main effect of country was also significant, F(1, 132) = 21.41, p < .01, η² = .14, and post hoc Bonferroni tests revealed that the Japanese children had higher response bias scores than the Hungarian children (p < .01). Furthermore, the main effect of familiarity was significant, F(1, 132) = 52.44, p < .01, η² = .28, and post hoc Bonferroni tests revealed that the children exhibited higher response bias scores for familiar objects than for unfamiliar objects (p < .01). Moreover, the interaction between familiarity and country, F(1, 132) = 16.49, p < .01, η² = .11, was also significant. Figure 2 shows the interaction. Simple main effect analyses revealed that the response bias scores for familiar and unfamiliar objects were significantly different between the Japanese and Hungarian children: Japanese children’s response bias scores were significantly higher than Hungarian children for both familiar (p < .05) and unfamiliar (p < .01) objects. Also, response bias scores for familiar objects were significantly higher than the scores for unfamiliar objects in both Japanese (p < .05) and Hungarian (p < .01) children.

Figure 1.

Japanese and Hungarian Children’s Mean Response Bias Scores in Experiment 1

Figure 2.

Interaction Between Country and Familiarity

We then conducted one-sample t tests for each age group to examine the interaction described above and to evaluate whether response bias scores were significantly higher or lower than zero (with a zero score indicating no bias). In the group of 3-year-olds, Japanese children exhibited a consistent yes bias for both the familiar objects, t(19) = 3.94, p < .01, and the unfamiliar objects, t(19) = 2.34, p < .05; the Hungarian children exhibited a yes bias only for the familiar objects, t(26) = 2.81, p < .01. In the group of 4-year-olds, the Japanese children exhibited a yes bias for the familiar objects, t(25) = 2.92, p < .01, and the Hungarian children exhibited a nay-saying bias for the unfamiliar objects, t(20) = −3.67, p < .01. Similarly, for the group of 5-year-olds, the Japanese children exhibited a yes bias for the familiar objects, t(19) = 5.34, p < .01, and the Hungarian children exhibited a nay-saying bias for the unfamiliar objects, t(23) = −5.85, p < .01.

In addition, neither the Japanese nor the Hungarian children in this study exhibited a high frequency of “I don’t know” and “no answer” responses. The children in all age groups in both countries showed these responses in 0% to 2.3 % of all responses (see Table 2 for exact numbers).

Table 2.

Frequencies of Children’s “I Don’t Know” and “No Answer” Responses in Japanese and Hungarian Children in Experiment 1

		Japan			Hungary
Response	Familiarity	3 Years	4 Years	5 Years	3 Years	4 Years	5 Years
I don’t know	Familiar	0	0	1	1	1	0
	Unfamiliar	5	9	11	4	3	7
No answer	Familiar	2	3	0	3	0	2
	Unfamiliar	5	3	0	2	3	1

First, we replicated the previous results for Japanese children, showing that the 3-year-olds exhibited a yes bias in both the familiar and unfamiliar conditions (e.g., Okanda & Itakura, 2008) and showing that the 4- and 5-year-olds exhibited a yes bias for familiar objects; this finding was similar in part to those of previous studies involving Japanese children (Okanda & Itakura, 2008, 2010b). Second, the 3-year-old Hungarian children showed inconsistent results in that they exhibited a yes bias for familiar objects but not for unfamiliar objects. Fritzley and Lee (2003), who investigated Canadian children, suggested that a developmental transition might occur at 3 years of age because the Canadian 3-year-olds exhibited a yes bias in some conditions but not in other conditions. Japanese 3-year-olds’ results were also not as robust as the results of 2-year-olds: 3-year-olds showed a yes bias for familiar objects (Okanda & Itakura, 2008; 2010b), but they showed a tendency to exhibit a yes bias (Okanda & Itakura, 2008) or did not exhibit a yes bias (Okanda & Itakura, 2010b) for unfamiliar objects. As Fritzley and Lee (2003) indicated, children might begin to say “yes” and “no” appropriately from the age of 3. In Experiment 2, we investigate whether Hungarian 2-year-olds show a yes bias for both familiar and unfamiliar objects similar to children in other countries.

Third, there was a cross-cultural difference between the Japanese and Hungarian older preschoolers. The Hungarian 4- and 5-year-olds exhibited a nay-saying bias for the questions pertaining to unfamiliar objects, whereas the Japanese 4- and 5-year-olds exhibited a yes bias for familiar objects. On the other hand, Canadian children in the same age groups did not exhibit any response biases when the questions were comprehensible (Fritzley & Lee, 2003). Children older than 4 years in different countries therefore exhibited different response biases. We assume that older preschoolers are more sensitive to this kind of interviewing situation than younger preschoolers, and they make binary decisions for various reasons.

Fourth, both the Japanese and Hungarian children exhibited a stronger yes bias for familiar objects than for unfamiliar objects. This tendency was also observed in previous studies involving Japanese and Vietnamese children (Okanda & Itakura, 2008, 2010b). However, Canadian children exhibited a stronger yes bias for unfamiliar objects than for familiar objects (Fritzley & Lee, 2003). Also, there was no familiarity effect when Japanese 2- and 3-year-olds were interviewed by their mothers (Okanda & Itakura, 2007). Japanese-French bilingual children did not show this effect when interviewed by strange adults (Okanda & Itakura, 2010a). The effect of object familiarity is mixed at this point, and further multiple investigations in a single country should be done to investigate this issue further.

Finally, the Japanese and Hungarian children did not frequently give “I don’t know” and “no answer” responses. This result for Japanese children was inconsistent with the results of previous studies, which reported that Asian children gave a high number of these responses (Okanda & Itakura, 2008) and that Japanese children showed this tendency when they were asked questions by their mother (Okanda & Itakura, 2007). Fritzley and Lee (2003) argued that children in Eastern cultures may say “I don’t know” more often than children in Western cultures because the former respects modesty and the latter respects self-esteem and self-confidence. Fritzley and Lee (2003) used incomprehensible questions that included nonsense words; however, none of the studies in Japan and other countries used this type of questioning, so further studies are needed to investigate this issue. This further investigation may also give us a better understanding of the issue of object familiarity. Okanda and Itakura (2008) indicated that Japanese and Vietnamese children showed “I don’t know” and “no answer” responses more for unfamiliar objects than for familiar objects probably because Asian children were likely to avoid a “yes” or “no” decision when they had less information about objects. It is hypothesized that “I don’t know” responses in Asian children would increase when yes-no questions include nonsense words rather than regular words.

Experiment 2

In Experiment 1, we found that the 3-year-old Hungarian children exhibited a yes bias in answering questions about familiar objects. However, the tendency toward a yes bias was not confirmed with respect to questions about unfamiliar objects. The results of previous studies involving 3-year-olds were inconsistent (Fritzley & Lee, 2003; Okanda & Itakura, 2008, 2010b), but results for 2-year-olds were clearly consistent across countries and conditions (Fritzley & Lee, 2003; Okanda & Itakura, 2007, 2008). Given these results, we predicted that 2-year-old Hungarian children would show a consistent yes bias, similar to children in other countries. Because there have been several reports of 2-year-old Japanese children consistently showing a strong yes bias (Okanda & Itakura, 2007, 2008), we did not include them in this investigation.

Method

Participants

Thirteen 2-year-old Hungarian children participated in this study (M = 29.5 months, SD = 2.60, range = 2;1~2;10; 7 boys and 6 girls). The children were recruited from nursery schools in Budapest, Hungary, and from a list of families that had volunteered to participate in psychological experiments. Some of the 2-year-olds were tested while the mother held the child on her knee, but the mothers were instructed not to give their children any hints. Before the experiment, the experimenters explained the purpose and method of the study to the mothers.

Materials and procedures

The procedures and materials were exactly the same as those in Experiment 1.

Results and Discussion

The Hungarian 2-year-old children’s mean response bias score for familiar objects was 0.58 (SD =.42) and for unfamiliar objects 0.50 (SD = .65). A within-subjects t test was conducted on the children’s response bias scores for familiar and unfamiliar objects. No main effect was revealed. Next, the children’s response bias scores were compared to the score of zero by means of one-sample t tests to examine whether they exhibited a response bias. The 2-year-old Hungarian children exhibited a yes bias to yes-no questions pertaining to both familiar, t(12) = 4.61, p < .01, and unfamiliar, t(12) = 2.78, p < .05, objects.

The Hungarian 2-year-olds consistently exhibited a yes bias for both familiar and unfamiliar objects. This result, along with the results of previous studies (e.g., Fritzley & Lee, 2003; Okanda & Itakura, 2008), indicates that a yes bias for 2-year-olds may be a universal phenomenon. Younger preschoolers may exhibit a yes bias possibly due to underdeveloped inhibitory control and language skills.

Experiment 3

In Experiment 1, Japanese children exhibited a yes bias at the age of 5. We hypothesized that Japanese older preschoolers exhibit a yes bias to yes-no questions that were asked by strange adults because they believed it was the way to show their politeness. We, therefore, investigated whether 3-year-old and 4-year-old Japanese children show a yes bias to their own mothers’ yes-no questions. If Japanese 4-year-olds do not exhibit a yes bias to their mothers’ questions, we can assume that they say “yes” to strange adults for social reasons.

Method

Participants

Thirty-three Japanese children who had previously participated in Okanda and Itakura’s (2007) study participated again in this study. The mean age of 3-year-olds (16 children; 8 boys, 8 girls) was 35.5 months (SD = 3.83, range = 28 to 42), and the mean age of 4-year-olds (17 children; 11 boys, 6 girls) was 47.9 months (SD = 4.36, range = 42 to 54 months). The children were recruited from a waiting list of families who wished to participate in our study in Kyoto, Japan.

Materials and procedure

A banana, a colored ink pen, and a plastic fork were used as familiar objects, and a small plastic part for hanging laundry, a paper filter for green tea, and a tie wrap were used as unfamiliar objects. As in Experiments 1 and 2, every object had four questions: two yes-questions for which the correct answers were “yes” and two no-questions for which the correct answers were “no.” The lists of the questions asked are provided in Appendix B.

To confirm whether these objects were familiar or unfamiliar to preschoolers, they were asked each object’s name and function at the end of the experiments. The results indicated that they knew the familiar objects but not the unfamiliar objects.

The procedure was a replication of the study of Okanda and Itakura (2007). The participants’ mothers received five objects (the banana was brought by mothers from their own stock), namely, an instruction sheet, a question and answer sheet, an example answer sheet, an envelope for sending back the answer sheet to the experimenter, and a consent document. The method of the experiment was thoroughly explained to mothers by providing them with instruction sheets and, if necessary, by giving additional instructions by phone. The mothers showed an object to their children and asked four questions about each object. They recorded both the verbal and nonverbal responses of their children. The order in which the children were presented with the familiar objects or the unfamiliar objects was counterbalanced. Approximately, half of the children were asked about a familiar object first, and the other half were asked about an unfamiliar object first. An experimenter arranged/counterbalanced the order of conditions and questions prior to the experiments and provided the mothers with the question and answer sheets with different orders. Mothers were asked to conduct the experiment when alone with the child at home and were told not to give their children any hint of the questions in advance. In addition, if the children said “I don’t know” or did not respond to the questions, the mothers were asked to record these responses as well.

Results and Discussion

The scoring method was identical to that in Experiment 1. A 2 (Age) × 2 (Familiarity) mixed ANOVA was conducted on the children’s mean response bias scores. The result indicated that the main effect of age was significant, F(1, 31) = 18.09, p < .01, η² = .37 (Figure 3), and post hoc Bonferroni tests indicated that 3-year-olds exhibited a stronger yes bias than 4-year-olds (p < .01). In addition, the main effect of familiarity was significant, F(1, 31) = 18.90, p< .01, η² = .38, and post hoc Bonferroni tests indicated that the children’s mean response bias scores were higher for the familiar objects than for the unfamiliar objects (p < .01). The interaction between age and familiarity effects was also significant, F(1, 31) = 13.88, p < .01, η² = .31. Simple main effect analyses revealed that older children’s response bias scores for familiar objects were significantly higher than that for unfamiliar objects (p < .01).

Figure 3.

Japanese Children’s Mean Response Bias Scores in Experiment 1

Next, we conducted one-sample t tests to examine the interaction above and whether the children’s response bias scores were significantly above or below zero (zero score indicates no bias). Mean response bias scores of children in two age groups for the familiar and the unfamiliar objects were compared to the score of zero. Three-year-olds exhibited a significant yes bias for both the familiar and unfamiliar objects, t(15) = 3.78, p < .01, and t(15) = 2.55, p < .05. Four-year-olds did not exhibit any response bias for the familiar objects, t(16) = 1.00, p = .33, and they exhibited a nay-saying bias for the unfamiliar objects, t(16) = −5.639, p < .01.

The children in this study rarely showed “no answer” response (3 times total) out of all questions. The children said they did not know the answers more often when they were asked about questions pertaining to unfamiliar objects (16 times total) than when they were asked about questions pertaining to familiar objects (6 times total). Again, these responses were less than 2.1% of all responses.

The previous studies reported that Japanese children aged 4 or older sometimes exhibited a yes bias to yes-no questions pertaining to objects when they were asked by strange adults, especially when they were asked questions about familiar objects (Okanda & Itakura, 2008, 2010b); however, the results we presented here revealed that Japanese children around the age of 4 did not exhibit a yes bias to the same type of questions when they were asked by their mothers. Surprisingly, the mean age of the 4-year-olds who did not exhibit a yes bias in Experiment 3 was about 48 months, while the mean age of 4-year-olds who exhibited a yes bias was around 54 months in the previous studies (Okanda & Itakura, 2008, 2010b) and in Experiment 1 in this study. The results in Experiment 3 indicate that Japanese 4-year-olds were able to answer yes-no questions appropriately: hence, they were likely to choose “yes” to yes-no questions pertaining to objects which were asked by strange adults. One plausible explanation is that this behavior is due to a cultural factor (Okanda & Itakura, 2010b). Japanese children may be more inclined to say “yes” than to say “no” to strangers because the former response shows much more respect to others, especially to older people in Japanese culture (see also Okanda & Itakura, 2010b).

In addition, the Japanese children in Experiment 3 exhibited a stronger yes bias for familiar objects than for unfamiliar objects. The previous study that used mothers as interviewers did not show this effect (Okanda & Itakura, 2007). The effects of object familiarity seem inconsistent across studies and may require further investigation.

General Discussion

In this article, we investigated whether Hungarian preschoolers exhibit a yes bias to yes-no questions pertaining to familiar and unfamiliar objects and compared their response tendencies with those of Japanese preschoolers. Also, Japanese 3- and 4-year-olds were asked the same questions by their mothers. We hypothesized that a yes bias would be found universally in younger preschoolers if they exhibit this bias, owing to a common mechanism (i.e., underdeveloped cognitive abilities). In addition, we hypothesized that older Japanese preschoolers would not exhibit a yes bias to their mothers, as they may be likely to say “yes” to strange adults for reasons of social pressure. We also investigated whether older preschoolers in different courtiers exhibit different response tendencies.

First, we found that young preschoolers from Japan and Hungary exhibited a yes bias to yes-no questions pertaining to objects as we hypothesized. More specifically, the Japanese 3-year-olds exhibited a yes bias to yes-no questions in both the familiar and unfamiliar conditions in Experiment 1, as reported in a previous study (Okanda & Itakura, 2008). The Hungarian 3-year-olds exhibited a yes bias only in the familiar condition in Experiment 1, but the Hungarian 2-year-olds exhibited a yes bias consistently in both conditions in Experiment 2. Results of Japanese 3-year-olds’ response tendency to unfamiliar objects were inconsistent: They tended to exhibit a yes bias in Okanda & Itakura (2008), but they did not exhibit any biases in Okanda and Itakura (2010b), but exhibited a yes bias in Experiment 1 in this study. These results suggested that young preschoolers might exhibit a yes bias due to a common mechanism (i.e., underdeveloped cognitive abilities), and children may begin to say “no” to yes-no questions from the age of 3.

Second, we found cross-cultural differences on older preschoolers’ response tendencies. The Japanese 4- and 5-year-olds exhibited a yes bias for familiar objects, whereas the Hungarian 4- and 5-year-olds exhibited a nay-saying bias for unfamiliar objects. Fritzley and Lee (2003) reported that Canadian children did not show any response biases to the same type of comprehensible questions. Only Japanese children seemed to sustain a yes bias for a longer period of time than the children in other countries; however, this result cannot be explained by a lack of cognitive abilities in Japanese older children. In fact, Japanese 4-year-olds did not exhibit a yes bias when they were asked yes-no questions by their mothers in Experiment 3: they even exhibited a nay-saying bias for unfamiliar objects. This result suggests that Japanese 4-year-olds were in principle able to say “no” to yes-no questions. Some studies provide further evidence to support this result: the performance of Japanese 4- and 5-year-olds on a DCCS task (Kirkham, et al., 2003; Zelazo et al., 1996) did not differ from that of North American children (e.g., Moriguchi et al., 2007), and generally, older preschoolers understood others’ utterances better than younger preschoolers (Siegal et al., 2009). Older Japanese preschoolers might not have exhibited a yes bias as automatically as younger preschoolers did. There could be other reasons for older Japanese preschoolers to say “yes” to strange adult interviewers.

Okanda and Itakura (2010a) alternatively indicated that cultural factors might influence older children’s response biases: Japanese older preschoolers might be likely to say “yes” because saying “yes” is much more polite than saying “no” in their culture. In fact, Tannen (1984) pointed out that polite Japanese do not say “no” in commercial communication. This idea could be expanded further by adopting the theory from cultural psychology. Japanese culture has been described as an interdependent-self culture (Markus & Kitayama, 1998), and a “yes” response may be considered more important than a “no” response because it may help to maintain harmony with others. In contrast, Canada has been described as having an independent-self culture (Markus & Kitayama, 1998), and saying “no” may be more acceptable.

However, this cultural difference between East and West does not explain why Hungarian children’s response tendency was different from that of Canadian children. Alternatively, language differences could be one possible factor that could account for cross-cultural differences in older preschoolers’ response bias. Hungarian, which belongs to the Uralic language family, is different from other European languages (e.g., English, French), which mainly belong to the Indo-European language family. It is also different from Japanese (Japonic language family) and Vietnamese (Austro-Asiatic language family). People need to understand others’ utterances pragmatically in a question-exchange situation, and some of the researchers have pointed out that pragmatic rules, or Gricean maxims, could be different among cultures (Bowe & Martin, 2007; Clyne, 1994; Hymes, 1986; Loveday, 1983). For example, maxims of manner are required to avoid obscurity, ambiguity, and to be brief and orderly (Grice, 1975); however, Japanese culture prefers indirect and tentative discussion about matters, and Japanese avoid plain statements (Loveday, 1983). Moreover, the maxim of quantity requires that one avoids being both less and overly informative (Grice, 1975): the former maxims of quantity seems universal, but Continental European and East/South-East Asian cultures prefer to be provided with more information, and the latter maxims of quantity may not be true for these cultures (Clyne, 1994). Therefore, pragmatic rules, for example, that define how a person should answer others’ questions, can be different in countries that speak different languages.

This issue needs further investigation; for example, we need to examine adults’ communicational styles in question-exchange situations and compare it to children’s response tendencies. Another possible investigation is asking children to elaborate their “yes” or “no” answers. There might be several meanings in “yes” or “no.” For example, Waterman, Blades, and Spencer (2001b) suggested a possibility of why children say “no” to unanswerable yes-no questions such as “Is a jumper angrier than a tree?” Children may either think that a tree is angrier than a jumper or that a jumper cannot be angry at all. However, the children do not elaborate their answers, and thus, they appear to say only “no” to such a silly question. Therefore, it is important to ask older preschoolers why they said “yes” or “no” to the questions to examine their reasons correctly.

In addition, it may be important for developmental psychologists to understand which yes-no questions are suitable for children in their own countries, as developmental researchers conduct experiments in different places around the world and yes-no questions are often used in experiments (Fritzley & Lee, 2003). The present study, along with previous studies, clarified that yes-no questions might not be suitable for young preschoolers, particularly 2- and 3-year-olds, in laboratories in Canada (Fritzley & Lee, 2003), Japan, Vietnam (Okanda & Itakura, 2008), and Hungary; however, there is a question remaining as to whether yes-no questions are suitable for older preschoolers in Japan and Hungary. Despite Fritzley and Lee’s (2003) conclusion that yes-no questions might be useful methodology for older preschoolers, our study indicates that using yes-no questions pertaining to children’s knowledge can result in response biases in Japanese and Hungarian children. Further investigation is needed to understand how we should obtain appropriate information from preschoolers avoiding such response biases in developmental experiments.

Footnotes

Appendix

Appendix B

Objects Used and Test Questions in Experiment 2

Familiar Object Condition		Unfamiliar Object Condition
Banana	Is this yellow?	Laundry catcher	Is this made of plastic?
	Is this for eating?	(plastic)	Is this blue?
	Is this for tapping shoulder?		Is this for wearing on the head?
	Is this green?		Is this for cutting?
Colored ink pen	Is this for writing?	Paper filter for	Is this square?
	Is this hard?	green tea	Is this used for making green tea?
	Is this made of cloth?		Is this hard?
	Is this for cooking?		Is this round?
Plastic fork	Is this for stabbing?	Tie wrap	Is this blue?
	Is this for eating?	(plastic and white)	Is this for binding?
	Is this made of paper?		Is this red?
	Is this soft?		Is this made of paper?

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a research fellowship from the Japan Society for the Promotion of Science (JSPS) for young scientists to Mako Okanda, by the Global COE Program “Revitalizing Education for Dynamic Hearts and Minds,” MEXT, Japan, and by a grant from the Japan Society for the Promotion of Science (JSPS) (No: 20330150, 202220004, 21220005) to Shoji Itakura. We are grateful to the parents and children who participated in this study.

References

Bowe

H. J.

Martin

(2007). Communication across cultures: Mutual understanding in a global world. Cambridge, UK: Cambridge University Press.

Carvajal

H. H.

Parks

J. P.

Logan

R. A.

Page

G. L.

(1992). Comparisons of the IQ and vocabulary scores on Wechsler Preschool and Primary Scale of Intelligence–Revised and Peabody Picture Vocabulary Test–Revised. Psychology in the Schools, 29, 22-24.

Clyne

M. G.

(1994). Inter-cultural communication at work: Discourse structures across cultures. Cambridge, UK: Cambridge University Press.

Dunn

L. M.

Dunn

L. M.

(1981). Peabody Picture Vocabulary Test–Revised. Circle Pines, MN: American Guidance Service.

Fritzley

V. H.

Lee

(2003). Do young children always say yes to yes-no question? A metadevelopmental study of the affirmation bias. Child Development, 74(5), 1297-1313.

Grice

(1975). Logic and conversation. In Cole

Morgan

(Eds.), Syntax and semantics, Vol. 3: Speech acts (pp. 41-58). New York: Academic Press.

Hodapp

A. F.

(1993). Correlation between Stanford–Binet IV and PPVT-R scores for young children. Psychological Reports, 73, 1152-1154.

Hughes

Grieve

(1980). On asking children bizarre questions. First Language, 1(2), 149-160.

Hymes

D. H.

(1986). Discourse: Scope without depth. International Journal of the Sociology of Language, 57, 49-89.

10.

Kirkham

N. Z.

Cruess

Diamond

(2003). Helping children apply their knowledge to their behavior on a dimension-switching task. Developmental Science, 6(5), 449-467.

11.

Loveday

(1983). Rhetoric patterns in conflict: The sociocultural relativity of discourse organizing processes. Journal of Pragmatics, 7, 169-190.

12.

Markus

H. R.

Kitayama

(1998). The cultural psychology of personality. Journal of Cross-Cultural Psychology, 29(1), 63-87.

13.

Moriguchi

Lee

Itakura

(2007). Social transmission of disinhibition in young children. Developmental Science, 10(4), 481-491.

14.

Moriguchi

Okanda

Itakura

(2008). Individual differences in children’s yes bias and the relationship to verbal ability and inhibitory control. First Language, 28, 431-442.

15.

Okanda

Itakura

(2007). Do Japanese children say “yes” to their mothers? A naturalistic study of response bias in parent-toddler conversations. First Language, 27(4), 421-429.

16.

Okanda

Itakura

(2008). Children in Asian cultures say yes to yes-no questions: Common and cultural differences between Vietnamese and Japanese children. International Journal of Behavioral Development, 32(2), 131-136.

17.

Okanda

Itakura

(2010a). Do bilingual children exhibit a yes bias to yes-no questions? Relationship between children’s yes bias and verbal ability. International Journal of Bilingualism, 14, 1-9.

18.

Okanda

Itakura

(2010b). When do children exhibit a yes bias? Child Development, 81, 568-580.

19.

Pratt

(1990). On asking children—and adults—bizarre questions. First Language, 10(29), 167-175.

20.

Siegal

Iozzi

Surian

(2009). Bilingualism and conversational understanding in young children. Cognition, 110(1), 115-122.

21.

Tannen

(1984). The pragmatics of cross-cultural communication. Applied Linguistics, 5(3), 189-195.

22.

Ueno

Utsuo

Iinaga

(1991). Picture Vocabulary Test–Revised. Tokyo: Nihon Bunka Kagakusha.

23.

Waterman

A. H.

Blades

Spencer

(2001a). Interviewing children and adults: The effect of question format on the tendency to speculate. Applied Cognitive Psychology, 15, 521-531.

24.

Waterman

A. H.

Blades

Spencer

(2001b). Is a jumper angrier than a tree? Psychologist, 14(9), 474-477.

25.

Zelazo

P. D.

Frye

Rapus

(1996). An age-related dissociation between knowing rules and using them. Cognitive Development, 11(1), 37-63.