• Science education reform is happening worldwide to facilitate
scientific inquiry (SI) ability (Lederman et al., 2019).
• Mastering scientific practices
is an essential element of
scientific literacy (Arnold, 2018)
it is an important
21st-century skill (Dogan, 2017)
Recent Reform of Taiwan
• Reform efforts in Taiwan have added a mandatory ‘Scientific
Inquiry and Practice’ curriculum for all high school students,
emphasizing ‘inquiry-based learning content’ including
1) observing and questioning,
2) planning and carrying out an investigation,
3) argumentation and modelling, and
4) expression and sharing.
• +Argumentation and Modelling, and Expression and Sharing.
Chang, P. S., Lee, S. H., & Wen, M. L. (2022). Developing an inquiry-based laboratory curriculum to
engage students in planning investigations and argumentation. International Journal of Science
Assessment Items using CCR
• Recently, rapid expansion of integrating assessment into
computer-based assessments, which can measure complex
inquiry skills more effectively (Kuo et al., 2015)
1. Are there any differences in scientific inquiry abilities among
demographic characteristics of high school students which
analysis using Bayesian inference?
p-value is being criticized
Traditionally, whether the null
hypothesis (H0) is significant
different or not is decided using
p-value (Frequentist statistic)
However, we have relied
heavily on p-value
Let’s try “Bayesian”
• Considering the ‘replication crisis’:
various alternatives have been
‘banning p-values from journals.’
• Bayesian approaches have become increasingly popular as their
posterior distributions carry more relevant information than p-
values (Kubsch et al., 2021).
• Does Male and Female have similar ability in Scientific Inquiry?
1. Probability of Male to have better SIA than female is true given the data
Bayes Factor (BF10) on gender is 449.36, which means the observed data support 449.36 times more
for the alternative hypothesis
2. Probability of City’s students to have better SIA than county’s students is
true given the data evidence!
scientific inquiry ability among school areas shows that there is very strong evidence
(BF10=3.94×10+134; 2logBF10>10; 2logBF10=269.19) to support the alternative hypothesis.
Comparison Models P(M) P(M|data) BF10 2logBF10
Male and Female Null model 0.50 0.002 1.00 0 -
Gender 0.50 0.998 449.36 5.30
City, County, and Rural Null model 0.50 2.538×10-135 1.00 0 -
School Area 0.50 1.000 3.94×10+134 269.19
Very strong evidence
• İlhan and Tosun (2016) report that gender is not the main predictor of the
understanding level of the scientific inquiry process in Turkish Kindergarten;
age, family income level, and the number of brother/sister is the main predictor.
• inequality between the rural and urban areas of the People's Republic of China
(PRC) has been reported to have a significant gap in education performance
(Zhang et al., 2015).
• Chyi and Zhou (2014) also report that in rural areas, financial stress has become
a significant deterrent to obtaining education since the market reform.