Prelogical Test: An Alternative Tool for Early Detection of Learning Difficulties

Abstract

Difficulties during the preschool age commonly lead to children who cannot solve problems, organize information and create meaning. It is necessary to predict factors that may affect their future learning. The aim is to develop an evaluation tool, to be applied in groups and that can easily evaluate results, to detect future learning problems in children of 3-6 years old. Computational intelligence techniques could contribute greatly to analyze results and to detect patterns that otherwise would not be apparent. Two protocols were implemented: an Indirect Variables Protocol (IVP) which captures children's personal data, and a Direct Variables Protocol (DVP) that assesses the graphic production of each student. The protocols were applied to a sample of 165 children of 3 to 6 years. The data was processed employing clustering algorithms. The clustering algorithm grouped the data into two distinct clusters. The first contains the cases of children who had difficulties in solving almost all the activities. The second contains those cases that correctly performed most activities. The age was not a relevant variable in this sample. Although the presented tool is still in testing phase, it showed to be appropriate to detect learning difficulties. Some variables considered resulted relevant predictors of children's performance. Cluster algorithms allow the isolation of groups of childr en whose performance is lower than the expected one. However, it is necessary to increase the size and diversity of the sample to confirm these tendencies.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 0 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.