Development of a web application for flowering phenology observation and its applicability to climate-related learning in elementary schools

Abstract This study used class practices to verify the applicability of a web application specially developed for the observation in general elementary school classes of the blooming of cherry blossoms with time-lapse digital images. The learning effects of class practices on children’s perceptions of the connection between familiar phenomena and climate change were examined with a questionnaire survey. The results further demonstrated the potential for plant phenology observation using time-lapse images to contribute to students’ acquisition of multi-timescale concepts from seasonal phenomena around them to century-scale climate change. This study can help provide a participatory approach, to an issue that is difficult for many learners to experience directly from the perspective of digital technology on environmental education.


Introduction
Natural areas are important for environmental learners to witness the influences of human activities on natural ecosystems, which is emphasized in climate-related learning (Brownlee et al., 2013).The scientific, social, ethical, and political complexity of climate change as a socioecological phenomenon demands a participatory educational approach that enables learners to meaningfully engage with the complexities of climate-related fact, value, power, and concern at multiple spatio-temporal scales (Rousell & Cutter-Mackenzie-Knowles, 2019).In promoting this approach in contemporary society, where the spatio-temporal gap between human societies and natural areas is large, pedagogical methodologies for participatory engagement with the gap may be inadequate.
The effects of human activities on ecosystems, such as climate change, can be learned in a participatory way through plant phenology-the study of various seasonal phenomena of plants in natural areas (Bombaugh et al., 2003;Nolan & Weltzin, 2011;Vliet et al., 2012).Changes in phenological elements such as leafing, flowering, and fruiting substantially alter the landscape and can be observed by non-specialists including children and young people.Environmental learners are encouraged to observe the process directly (Daniel et al., 2004;McNeill & Vaughn, 2012;Shepardson et al., 2009Shepardson et al., , 2012)), which is why experiential learning through plant phenology in natural areas is highly valuable.Verlie (2017) argues for a climate change learning approach that does not conceptualize humans and climate as independent entities, but rather allows environmental learners to become practical components of the climate process.From this perspective, the presentation of scientific knowledge is likely insufficient, and therefore, the value of practical experiences should be considered.
The educational importance of plant phenology observation as a practical experience has been described mainly in the context of citizen science.Vliet et al. (2012) described the potential of a Dutch phenological network involving over 8,000 citizen volunteers (including children) to raise awareness of climate change and biodiversity.Rosemartin et al. (2021) described the potential of a 10-year effort by the citizen science project of the USA National Phenology Network to help participants understand the effects of climate change on national parks.For a more direct application to educational settings, Bombaugh et al. (2003) reported on the incorporation of Global Learning and Observations to Benefit the Environment (GLOBE) activities in American middle schools.Further, Emery et al. (2019) developed a hands-on science education program through phenological observations for higher education.These studies demonstrate the potential of plant phenology observation as a methodology for observing natural responses to climate change practically, but they lack a longer-term time-scale experience.Most often, citizen scientists might have observed plant phenology directly for a few years, but without being able to participate experientially in long-term phenological observations over which the effects of climate change may be apparent.
Fixed cameras that automatically capture photographs at certain intervals in natural areas and observe temporal changes in the same location provide the opportunity to overcome such challenges.Over a few years, environmental learners can directly observe plant phenology that shows short-term changes independent of climate change.The inclusion of virtual observation through time-lapse photography to complement such direct experience makes it possible to observe plant phenology on a multi-time scale on which long-term fluctuations occur.Although the educational importance of plant phenology using time-lapse photography has been theoretically discussed (Brownlee et al., 2013;Nakamura et al., 2018), its effectiveness has not been sufficiently examined through practical research.
The Cyberforest research project based at the University of Tokyo has been photographing and recording at fixed points (Saito et al., 2015), under the concept of "The Real Nature Beyond the Internet."Through this project, Nakamura et al. (2019) presented a methodology of climate change learning based on the observation of the blooming of cherry blossoms in a Japanese forest using time-lapse images recorded since 1995.However, this method requires considerable labor to print and distribute the time-lapse images.In addition to wasting paper resources, the small size of each image makes it difficult to observe details or for multiple students to collaborate.The use of a digital-based instrument to address such operational challenges would create a situation in which the effects of climate-related learning could be examined.
Climate-related learning has been actively discussed in recent years, mainly from the perspective of environmental and sustainability education (Reid, 2019), but not much practical knowledge exists about the participatory approach in natural areas as described above.For example, Trott (2022) reported that the Science, camera, action!(SCA) program helped children aged 10-12 years acquire a deeper respect for nature, but this was merely conceptual, not real.Swim et al. (2017) conducted a study focusing on nature-based museums as the implementers of climate-related learning and discussed nature as a place in relation to climate change, but its primary concern was focused on the museum as a cultural institution.Practices that look at the authentic nature encompassing multilayered time scales, with short periods of time when real experiences are possible and long periods when virtual experiences through digital technology are valid, have rarely been discussed, especially in climate-related participatory learning.
The purpose of this study was to verify the applicability of a web application, newly developed through this research, for the observation of the blooming of cherry blossoms with time-lapse digital images in general elementary school classes.Furthermore, the learning effects on children's perceptions of the connection between familiar phenomena and climate-related issues through the classes as cases were considered.This study addressed the following research questions: (1) How do the children use the developed web application in an actual elementary school class?(2) What is the effect of the class using the web application from the perspective of climate-related learning?(3) What insights do the class practice provide into the utilization of digital technology on environmental education?

Methodology and methods
The research consisted of three steps: web application development, classroom-based application, and questionnaire surveys and data analysis.As 141 participants in this research were deemed insufficient to discuss only quantitative data, a mixed-methods design was used which also included qualitative data, drawing on previous research on climate-related learning practices (Trott, 2022).These processes were conducted in Japanese and translated into English as needed for this study.We explained the purpose of this research to the subject students through their classroom teachers and obtained their consent to participate.The ethical review committee at the institution to which the corresponding author belongs waived the ethical review for a classroombased practice and questionnaires in a classroom.

Web application development
The web application employed in this research was constructed using HTML and JavaScript, to allow it to operate on common tablet devices.The application is thus independent of the device's operating system and can be used with any modern web browser, including Chrome, Safari, Firefox, and Edge.This application runs on any device that can use a web browser, not just tablets, but its screen layout was designed for use on a typical 10-inch class-size tablet device.
The main components of the application were as follows: a list display of daily fixed-point images from April 16th to May 15th by year since 2001, a magnified display of each daily image, the registration of full-bloom date by year, and transition graphs of the full-bloom dates and temperature data (Figure 1).These components allowed users of the application to observe the daily images in detail, compare them with those of the previous and following days, and quickly confirm the annual change of days determined according to certain criteria, such as the full bloom date of cherry blossoms and their relationship to the temperature of each year.The images used for the application were daily time-lapse images of the University of Tokyo Chichibu Forest, taken as part of the University of Tokyo Cyberforest Project, which had been reported by as an example of observation on paper media.For temperature data, the average temperature from February to April was used as a representative value by year.The three-month average temperature showed a strong correlation with the full bloom dates based on observations of these images (Nakamura et al., 2019).
The developed application is used from the user's terminal via Internet.The English and Japanese versions of the application, titled "Observation of cherry blossoms," are available to the public and can be used freely by anyone.The images on the site were updated after 2019 when the application was developed.As of 28 April 2022, images up to 2021 are available.Although the application title is specific, it is not intended to exclude observations of other phenomena (e.g., observing the unfolding leaves of a beech tree is also possible through this application).

Classroom-based application
Single periods (45 minutes) of a Japanese elementary school's 5th-and 6th-grade classes (ages 11 and 12) were used for this study.The class-based research methods were conducted in two 6th-grade classes on 6 February 2020, and in three 5th-grade classes on the following day, 7 February 2020.Class sizes ranged from 25 to 33 students.A total of 141 students, 63 in the sixth grade and 78 in the fifth grade, participated in this study.The class was not conducted by the classroom teachers but by three fourth-year university students in the teacher training program who had been familiarized with the objectives of this study.As it was deemed difficult to observe 19 years of cherry blossom patterns (from 2001 to 2019) in a 45-minute class period for the observation, data were provided for 12 years (from 2008 to 2019), as this was considered a time scale easier to imagine given its similarity to the target students' age.
The following three points based on the practice of Nakamura et al. (2019), in which the same fixed-point images were used, were set as the goals of the class: (1) To understand that the blooming time of cherry blossoms changes every year; (2) To understand the relationship between temperature and the full bloom dates of cherry blossoms based on the graph; and (3) To compare predicted long-term changes with short-term (12-year) changes and infer correlations with environmental issues.The class plan consisted of a 10-minute introduction, 23 minutes of main content, and a 12-minute summary.This class plan was made available on the web along with the application.
In the introduction, pictures of cherry trees in spring, summer, fall, and winter were presented to arouse students' interest in the seasonal changes around them.The progress of the "cherry blossom front," which is similar to weather fronts, moving northward across a map of the Japanese Islands was then shown, and students were asked the following two questions: "Where did cherry blossoms begin to bloom?" and "You recognized that the blooming day is different depending on location.Then, would it be the same at the same location every year?"For the main content, students were divided into groups of four to observe cherry blossoms through a web application using one 10-inch class tablet per group (Figure 2).In the first 15 minutes, students determined the full bloom dates of cherry blossoms for 12 years (from 2008 to 2019) through their observations of the photographs on the web application.Here, when there seemed to be multiple full bloom dates in a certain year, students were told to select the earliest date among them.Through this observation, students considered what causes changes in the full bloom date of cherry blossoms.In the last 8 minutes of the main content, students considered the relationship between the cherry blossoms and temperature by drawing graphs of changes in the full bloom dates of cherry blossoms and temperature data using the application's functionality.
For the summary, students came to understand the influence of temperature changes on a global scale on the cherry blossoms and thought about the future through the following three questions: first, "What influence can be expected if global warming progresses?," to which the students shared their thoughts.Next, a graph of average temperature changes over the past 100 years by the Japanese Meteorological Agency was presented with the second question, "The average temperature only increased by one degree during the 100 years.Do you feel the change is small or significant?"Related to this question, the teacher added that cherry trees may not bloom after 100 years in the area of the elementary school (Yamanashi Prefecture) because cold temperatures during the cold season are necessary for cherry trees to bloom.Finally, the third question, "What do you think about future changes in cherry blossoms as a result of temperature changes?" was asked, and students wrote their responses to this question on a worksheet that was distributed by the authors at the beginning of the class.

Questionnaire survey and data analysis
Immediately following the class, a questionnaire survey was conducted on the usability of the web application (Qa1 in Table 1) and students' thoughts on the three goals of the class (Qa2 in Table 1).Questions regarding goal (1) (Qa2-1, 2, 3 in Table 1) were also conducted before the class, as it was deemed feasible to ask these questions before class (Qb1, 2, 3 in Table 1).Questions regarding goal (2) (Qa2-4, 5, 6 in Table 1) and goal (3) (Qa2-7, 8, 9 in Table 1) were answered only after the class.All responses to the questions were chosen from a five-point rating: i) agree, ii) somewhat agree, iii) neither agree nor disagree, iv) somewhat disagree, and v) disagree.
For the quantitative analysis, the number of responses to each option for each question was first simply tabulated.The most positive of the five options for each response was then scored as 5-points and the most negative as 1-point.Corresponding t-tests were used to check for significant differences in the means for the three items for which the same questions were asked before (Qb1, 2, 3 in Table 1) and after (Qa2-1, 2, 3 in Table 1) the class.In addition, based on the answers provided for each of the three sets of pre-post questions, students were divided into two groups: those who had positive changes before and after the class and those who did not, along with Wilcoxon's rank-sum test was used to confirm whether there was a significant difference in the tendency to answer each other question between the two groups.The significance levels for all statistical tests were set at 5%.
For the qualitative analysis, each student's free response on the worksheet to the question posed to the students at the end of the class, "What do you think about future changes in cherry blossoms as a result of temperature changes?" was targeted, and the results were interpreted in combination with the quantitative analysis.

Q_ID Question
What do you think of the following?

Qb1
This winter is quite different from last winter.

Qb2
The plants around me do not look so different every year.
Qb3 I find many differences between this year and last year in my daily life.

What did you think of the iPad application used in today's class?
Qa1-1 I had fun using the application.

Qa1-2
There was not enough time to use the application.
Qa1-3 I found the application inconvenient at times.
Qa1-4 I would like to use the application more.

What did you think of the following after today's class?
Qa2-1 This winter is quite different from last winter.

Qa2-2
The plants around me do not look so different every year.
Qa2-3 I would like to find many differences between last year and this year in my daily life.
Qa2-4 I discovered by myself that the full bloom dates of cherry blossoms are related to the temperature.

Qa2-5
Even the slightest change in temperature has meaning for plants around me.

Qa2-6
It is not necessary to observe the full bloom dates of cherry blossoms, but it is enough to measure the temperature with a thermometer.

Qa2-7
Observing the full bloom dates of cherry blossoms is related to thinking about the future of the earth.

Qa2-8
Even if more people observe cherry blossoms carefully, the future will not change eventually.

Qa2-9
Observing the full bloom dates of cherry blossoms over the 12 years has led me to think about the future 50 or 100 years from now.

Results
All five class practices were conducted as planned.It was confirmed that all groups of students used the web application successfully and drew reasonable transition graphs of the full-bloom dates through 15-minute observations of the photographs.Each classroom teacher reviewed the progress of the class practices, and no problems were noticed.Through this class practice, valid responses to the questionnaires were obtained from all 141 students who participated in the classes using the web application.
Three out of four questions on the usability of the application resulted in a high average score of 4.6 or higher (Qa1-1, Qa1-3, Qa1-4 in Table 2 and Figure 3).The remaining question, which asked whether students had enough time to use the application, yielded a relatively low mean score (Qa1-3 in Table 2 and Figure 3).However, all groups were able to complete the 12-year observation, which suggests that there were no major problems in the set time of use.Given that many students wanted to use the application more (mean of Qa1-4), it seems likely that students who were positively motivated to observe more carefully might have felt that they had insufficient time as a result.
In the immediate post-class responses to questions about students' thoughts on the three goals of the class, the average score suggested that the class goals were generally met by the group of students as a whole (Table 2 and Figures 3-5).The question "The plants around me do not look so  different every year" (Qa2-2 in Table 2 and Figure 4) had a relatively low mean score, but the mean scores for all three questions related to the class goal (1), including this one, were significantly higher based on corresponding t-tests after the class than before (Figure 4), which indicates some effectiveness related to goal (1).Referring to the class contents, the images used were mainly not of plants around them, and the annual differences were summed up in the context of minute rather than major differences (but ones that could have serious effects if they accumulated over a long time).These factors might have resulted in the students' thoughts that they could not strongly disagree with the reversal item that familiar plants do not look "so different" every year.
Regarding the responses to the question about the class goal (1), the number of students who had a positive change before and after the class was 39 in Qb1 to Qa2-1, 66 in Qb2 to Qa2-2, and 70 in Qb3 to Qa2-3, respectively, out of 141 students (Table 3).
The group of students who had a positive change in "This winter is quite different from last winter" (Qb1 to Qa2-1) were significantly more positive in their responses to four questions (Qa1-4, Qa2-3, Qa2-5, Qa2-9), which included those that indicated a direct understanding of each goal of the class, immediately after class than the group who did not (Table 3 and Figure 6).As this question (Qb1 in Table 2) had a high average and the most positive response from before the class was also slightly higher at 50 out of 141 students, the positive change group includes a large number of students with relatively low interest in seasonal phenomena.This could be interpreted as indicating that the students' motivation to study for the three goals of the class was piqued by their interest in the application (Qa1-4) and the accompanying increased interest in seasonal phenomena, as represented by two students' free writings quoted below.I don't want to see the cherry blossoms disappear because they are one of the cultural elements that I would like to preserve in Japan.I am sure that the cherry blossoms are not the only thing that is affected by global warming, and as far as I know, there are not many benefits of global warming.If this application is widely used, it can be used for research and study, and I think it will be an interesting and fun way to learn about nature in today's world.I wanted to do something about global warming when I learned that some cherry trees have stopped blooming because of the change in temperature between now and the past.It was fun to use the app to find out when the cherry blossoms were in full bloom now and in the past.I would like to learn more about cherry blossoms.
The group of students who had a positive change in "The plants around me do not look so different every year" (Qb2 to Qa2-2) were significantly less (not more) positive in their responses to the following question, "Observing the full bloom dates of cherry blossoms over the 12 years has led me to think about the future 50 or 100 years from now." (Qa2-9 in Table 3 and Figure 7) In interpreting this result, it should be taken into account that this question among the three questions about the class goal (1) has a relatively low response, although significantly increased average from pre-class to post-class.As already noted, a certain number of students seemed to recognize annual changes as minute throughout the class, and it is likely that many of them were included in the group that did not change positively on this question.Since they recognized that these minute annual changes could lead to serious problems over the long term, they could strongly agree that 12 years of observation of the full bloom dates of cherry blossoms had led them to think about 50 or 100 years from now.Indeed, in the free writing of the students belonging to the non-positive group, there were several descriptions in which they seriously considered the possibility that the accumulation of minute changes in temperature could lead to the cherry trees not blooming in the future, as represented by two students' writings quoted below.
I was surprised to see how many things can be changed by a slight change in temperature, even though the temperature does not change much.Global warming changes so many things, and it was a little scary to think that we might not be able to do what we normally do, in the future because of it.
When I thought about the possibility that the cherry trees might not bloom someday, I realized that even if the temperature does not rise that much now, we will have to be careful in the future, and that we must take good care of cherry trees.
The group of students who had a positive change in "I (would like to) find many differences between last year and this year in my daily life" (Qb3 to Qa2-3) were significantly more positive in their responses to four questions (Qa1-3, Qa2-5, Qa2-6, Qa2-7), which included those regarding the class goals, immediately after class than the group who did not (Table 3 and Figure 8).The significantly positive responses suggest a deep understanding by the students regarding the class goals in that the seasonal phenomena of plants are closely related to temperature, and that observing them can lead to consideration of the future of the earth.Furthermore, they were able to recognize that their deep understanding was also connected to their daily lives (Qa2-3).The fact that this positive group also responded significantly positively to the usability of the application (Qa1-3) suggests that students who were able to handle the application, rather than simply use it, were able to learn more deeply regarding the class goals, as indicated by two students' free writings quoted below.I heard that if global warming continues, there may be no more cherry blossoms.As the temperature will continue to rise, we are trying to help people through the use of programming.Of course, we also want to help the earth.
I was surprised to hear that global warming would make it impossible to see cherry blossoms.I was horrified to think that we would not be able to see such beautiful cherry blossoms anymore.I want to do what we can do to prevent global warming.The iPad was very easy to use.The application was also very easy to understand, and I learned a lot about cherry blossoms.

Applicability of the web application in elementary school classes
The applicability of the web application for cherry blossom observation to elementary school classes, one of the main considerations of this research, was confirmed by the progress of the actual class practices and the results of the questionnaire immediately after the classes.Compared to the previous method of printing on paper (Nakamura et al., 2019), a larger image can be displayed, making it easier to observe the details of blooming conditions and enabling collaborative learning in which multiple students view and interact with the same image together.The labor required for preparation depends on the terminal environment of the school where the class is held, so it is difficult to make a general evaluation.However, the application developed in this research can be used with any device with a 10inch or larger screen that can use a Web browser, even if it is not a tablet, and classes can be held with one terminal for every four students, so there are probably many schools where the application can be implemented within the range of ordinary terminal environments.
The information and communication environments in schools are rapidly being upgraded, driven in part by the recent COVID-19 pandemic.In Japan, the educational policy of the so-called "GIGA School Concept" is being implemented to provide one terminal per student from AY 2020 onward (Kihara, 2021).Elsewhere, for example, the European Commission has issued a Digital Education Action Plan (2021-2027), which aims to reduce the disparities in the information and communication environment in schools all over the region.In the near future, the number of schools where students can learn using the web application of this research will rapidly increase.Furthermore, the application is not limited to use in schools, but could also be used for online learning at home and other locations.The class practices in this research were limited to one 45-minute class which emphasized the applicability to compulsory education schools that tended to have overcrowded curricula.However, considering the fact that many students wanted to spend more time using the application, it might be effective to develop a more advanced program that combines school classes and home study.

Effects of the class from the perspective of climate-related learning
The three goals set for the class practices in this research were generally achieved for the student group as a whole.This indicates that the effects of learning related to climate change were observed in more detail than in previous research in which similar observational learning was conducted in a paper-based format (Nakamura et al., 2019).This would mean that the potential for plant phenology observation using time-lapse images to contribute to students' acquisition of multi-timescale concepts covering seasonal phenomena around them to century-scale climate change has been further augmented.It could also be considered a process by which students can personalize climate change through the familiar phenomenon of cherry blossom blooming.
However, several points should be noted when examining the details.The contents of the classes in this research could have been more effective for students who were less interested in seasonal phenomena in terms of their direct effects on the class goals.In other words, it might be important to have an additional program for students who are more interested in seasonal phenomena to encourage more advanced learning.We should also consider the possibility that the students' mastery of the application, rather than simply using it, could have led to deeper learning relevant to the class goals.It would then be worth considering a more time-intensive program management that would allow more students to become expert users of the application.Furthermore, if the priority of the program is to encourage students to think about the long-term effects of climate change, such as 50 or 100 years from now, it might be useful to emphasize that the annual changes are minute and difficult to notice, in relation to the possibility that the accumulation of annual changes could lead to major problems.In this practice, an impressive episode of cherry blossoms that may not bloom after 100 years was used to emphasize completing the lesson as a 45-minute class, but it is worth considering that several concerns were raised that making students fearful may decrease problem-solving behavior (Morton et al., 2011;Wang & Chen, 2022).

Insights into digital technology on environmental education from the class practice
If we reconsider the process of the class practices in this study as a case of climate-related learning from the perspective of digital technology on environmental education more generally, the possibility arises of a participatory approach to an issue that is difficult for many learners to experience directly.What should be re-emphasized here is that digital technology is valuable not as efficiency enhancers or substitutes for direct experience, as feared by Payne (2006), but rather as extensions and complements of direct experience.Real observation of familiar seasonal phenomena is spatiotemporally extended by virtual observation of time-lapse images, and the learning gained from the virtual observation complements and deepens the familiar real observation.This process may enable education "in" environments of any place and time on Earth, which is truly the object of environmental education, beyond the constraints of real space.Greenwood and Hougham (2015) state, "Never before has so much computing power been wielded by so many people in networks that literally cover the planet," (p.100) but this is only for human-inhabited areas.Computer power does not necessarily extend to natural areas.When it comes to the natural areas, computer power may serve to promote education for deep wonder (Schinkel, 2020;Washington, 2018).It is not "Bringing distant places closer, and potentially misrepresenting what they are" (Greenwood & Hougham, 2015, p. 111), but rather glimpsing distant places and creatively wondering what they are beyond the Internet.

Conclusion
This study developed a web application to observe the blooming of cherry blossoms with timelapse digital images and verified its applicability through class practices in general elementary school classes.A larger image can be displayed on the application, making it easier to observe the details of blooming conditions and enabling collaborative learning in which multiple students view and interact with the same image together.
The learning effects of class practices on children's perceptions of the connection between familiar phenomena and climate change were examined through a questionnaire survey.The results demonstrated the potential for plant phenology observation using time-lapse images to contribute to students' acquisition of multi-timescale concepts from seasonal phenomena around them to century-scale climate changes.We could consider the process of class practices in this study as a case of possibility arises of a participatory approach, to an issue that is difficult for many learners to experience directly from the perspective of digital technology on environmental education.The results of the present practice suggest that the degree and nature of the effects differ a little depending on differences in readiness, such as interest in seasonal phenomena and digital devices.In the future, a comprehensive discussion considering the practice with older students is warranted.

Figure 1 .
Figure 1.On-screen display of the developed web application.

Figure 2 .
Figure 2. Learning with tablets in the class practice.

Figure 3 .
Figure 3. Simple tabulation of the responses to the questionnaire on the usability of the web application, after the class.The Qa1-1, 2, 3, and 4 are linked to those in table 1.The Qa1-2 and Qa1-3 are reversal items.

Figure
Figure 4. Simple tabulation of the responses to the questionnaire asked before and after the class regarding goal (1) of the class.The Qb1, 2, 3, and Qa2-1, 2, 3 are linked to those in table 1. Regarding Qb1 and Qa2-1, Qb2 and Qa2-2 (reversal items), and Qb3 and Qa2-3, each of which asked almost the same questions, and all combinations showed a significant increase by corresponding t-tests in positive responses after class compared to before class.

Figure
Figure 5. Simple tabulation of the responses after the class to the questionnaire on goals (2) and (3) of the class.The Qa2-4, 5, 6, 7, 8, and 9 are linked to those in table 1.The Qa2-6 and Qa2-8 are reversal items.

Figure
Figure 6.Results of the stratified analysis of positive and non-positive groups for the Qb1 to Qa2-1.The Q_IDs are linked to those in table 1. Asterisks (*) indicate significant differences by Wilcoxon's rank-sum test.

Figure
Figure 7. Results of the stratified analysis of positive and non-positive groups for the Qb2 to Qa2-2.The Q_IDs are linked to those in table 1. Asterisks (*) indicate significant differences by Wilcoxon's rank-sum test.

Figure
Figure 8. Results of the stratified analysis of positive and non-positive groups for the Qb3 to Qa2-3.The Q_IDs are linked to those in table 1. Asterisks (*) indicate significant differences by Wilcoxon's rank-sum test.