Data Visualization Using Boxplots: Comparison of Metalloid, Metal, and Nonmetal Chemical and Physical Properties

Self Cite

This laboratory experiment was divided into four parts. In the first part, students evaluate previously published data and check their normality using histograms, Q−Q plots, the Shapiro−Wilk test, and boxplots. In the second part, two different groups were compared. First, data normality and homoskedasticity were checked by using the Shapiro−Wilk test and Levene's test, respectively. Then the group mean was compared using the Student, Welch, and Mann−Whitney tests. In the third part, the famous Iris flower data set was analyzed, where data normality and homoskedasticity were checked. Then, three independent groups were compared using ANOVA, without correction, and using the Welch and Brown-Forsythe corrections. The Iris flowers data set was also compared using the nonparametric version of ANOVA (Kruskal−Wallis test) and post hoc tests such as the Tukey test (parametric) and Dunn's test (nonparametric). In the fourth part, students measured the density of water using four pieces of glassware (a beaker, a graduated cylinder, a volumetric pipet, and a graduated pipet). After the experiment, the data' normality and homoskedasticity were checked. Then, the data was compared using ANOVA and post hoc tests. The results of these tests were visually interpreted using raincloud and discriminant plots. These data provide a starting point for teaching precision, accuracy, descriptive statistics, and hypothesis tests. Statistical analysis was carried out using Jeffreys's Amazing Statistics Program (JASP), which is free and open-source software that provides several graphical interpretations of the hypothesis test.

Section: Methodsmentioning

confidence: 97%

Section: Methodsmentioning

confidence: 99%

Section: ■ Experimental Overviewmentioning

confidence: 99%

“…The distribution plot (Figure 1) shows that the data were negatively skewed. 12 Thus, the data have a negative skewness (Table 1).…”

Section: Descriptive Statisticsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Teaching Descriptive Statistics and Hypothesis Tests Measuring Water Density

Silva de Souza,

Borges

2023

Self Cite

Enhancing Statistical Education in Chemistry and STEAM Using JAMOVI. Part 2. Comparing Dependent Groups and Principal Component Analysis (PCA)

Sequeira,

Borges

2024

Self Cite

In statistics and chemometrics education for chemistry and STEAM (Science, Technology, Engineering, Arts, and Mathematics) students, gone are the days of laborious calculations and conclusions based solely on p values. Free and graphical user interface (GUI) software like JAMOVI streamlines calculations, generates informative plots, and fosters a deeper understanding beyond just p values. This article serves as a comprehensive guide for instructors on utilizing JAMOVI to teach essential parametric and nonparametric tests commonly encountered in these fields. Our focus centers on a paired samples t test, Wilcoxon rank sum test, repeated measures ANOVA (RMANOVA), and Friedman test. We also delve into principal component analysis (PCA) using the MEDA plugin, which generates highquality colored plots that visually elucidate trends between groups. Through 21 guided questions, students will assess data normality, compare dependent groups using both parametric and nonparametric tests, explore comparisons between dependent groups, and observe trends using PCA. These STEAM-contextualized questions and practical examples empower educators to seamlessly integrate JAMOVI into their teaching, enhancing the learning experience for students.

Enhancing Statistical Education in Chemistry and STEAM Using JAMOVI. Part 1: Descriptive Statistics and Comparing Independent Groups

de Souza,

Sequeira,

Borges

2024

This laboratory experiment was devoted to teaching descriptive statistics and comparing independent groups to STEAM (Science, Technology, Engineering, Arts, and Mathematics) students using open-source and graphical user interface software. Students answered 21 questions using JAMOVI in previously published data sets to learn fundamental statistics concepts. It was divided into four parts. In the first part, descriptive statistics were carried out (mean, median, standard deviation, interquartile range, data normality, and skewness). In the second part, data normality was checked by using visual inspection of plots (histograms and Q− Q plots). In the third part, two independent groups were compared. In the fourth part, more than two independent groups were compared. Normally, comparisons between two or more groups are presented in many textbooks, and a normal and homogeneous distribution of the data is assumed. Only parametric tests were taught, while nonparametric tests were not presented. Thus, data normality was checked using hypothesis tests (Shapiro−Wilk, Kolmogorov−Smirnov, and Anderson−Darling tests). Then, homogeneity was checked using Levene's and Bartlett's tests. Normality and homogeneity were also checked using a visual inspection of plots. Once normality and homogeneity were checked, parametric tests were used (t test and ANOVA). If the normality of the data was not checked, nonparametric tests were used (Mann−Whitney and Kruskal−Wallis tests).