Multivariate image analysis: use of a portable application in a Problematized Experimental Activity (PEA) approach
Name: GABRIEL TEIXEIRA MALACARNE
Publication date: 27/02/2026
Examining board:
Name |
Role |
|---|---|
| ANDRÉ LUIS SILVA DA SILVA | Examinador Externo |
| PAULO ROBERTO FILGUEIRAS | Presidente |
| PAULO ROGERIO GARCEZ DE MOURA | Coorientador |
| RAFAEL DE QUEIROZ FERREIRA | Examinador Interno |
Summary: The study evaluated the applicability of multivariate image analysis (MIA) for the teaching of chemometrics through problematized experimental activity (PEA) in higher education in chemistry, investigating the contributions of the use of RedGIM® as a technological resource for the interpretation of multivariate chemical data. The research adopted a mixed approach, combining qualitative and quantitative methods, with an experimental design for chemometric analysis via partial least squares regression (PLS) and didactic intervention to assess grounded theory (GT). For the application of the PLS method, the determination of the antioxidant capacity of 21 honey samples was carried out by means of the FRAP method associated with UV-Vis spectrophotometry, in conjunction with the registration and treatment of digital images in RedGIM®. The Shapiro-Wilk test was applied to verify the normality of the data, and the paired T-test was applied to indicate whether there was no statistically significant difference between the results of the RedGIM® and UV-Vis. The results showed that the generated model presented satisfactory metrics (R2 = 0.995; relative error = 4.6%; RMSEC = 0.877 µmol L1; RMSEP = 0.832 µmol L1 and bias = 2.9 × 1015 µmol L1), showing good adjustment and prediction capacity, with internal consistency and absence of overfit. Statistically, the tests indicated that there is no significant difference between the values obtained by the proposed method (RedGIM®) and by the reference method (UV-Vis). In educational context, the GT analytical process indicated that the PEA promoted the integration of chemical concepts, multivariate data interpretation, and laboratory practices, as evidenced by participants’ reports and the activities carried out throughout the intervention. These findings suggest the occurrence of meaningful learning, grounded in David Ausubel’s theory, as well as the development of metacognitive skills. According to the statistical analysis, the metacognitive indicators A1, A2, A3, P3 and C4 showed significant differences between the beginning and the end of the application. The use of RedGIM® through MIA demonstrated performance comparable to the reference analysis using UV-Vis in the application of the determination of the antioxidant capacity of honey samples. In addition, RedGIM® contributed as an effective technological resource, allowing students to apply chemometrics concepts in a practical and accessible way.
