Strategies to develop life skills while Geography learners use ChatGPT-4o to interpret contour maps

  • 0

Abstract

The rapidly changing 21st century requires learners to develop a set of life skills during education and training in order to be successful in future careers. In the South African context, the South African Qualifications Authority (SAQA) have identified seven critical crossfield outcomes (CCFOs) and five developmental outcomes (DOs) which should guide South African education and training. These outcomes can be translated as life skills (LSs) which learners must develop while they are busy with subject content. Despite this requirement of SAQA, little research has been conducted to suggest strategies for developing the LSs while learners are busy with subject content.

The primary purpose of this research was to design strategies for developing the LSs while Geography learners use ChatGPT-4o to interpret contour maps. However, due to lack of research regarding (a) ChatGPT-4o’s ability to interpret contour maps and (b) the adoption of ChatGPT-4o for this purpose, the primary research question could not be answered without investigating whether or not ChatGPT-4o can interpret contour maps, and if it would be adopted for this purpose. Thus, the research was conducted within an overarching multi-phase mixed methods research design, to answer the following questions in consecutive phases:

  • To what extent can ChatGPT-4o interpret contour maps?
  • Which factors can drive the adoption of ChatGPT-4o to interpret contour maps?
  • How can LSs be developed while Geography learners use ChatGPT-4o to interpret contour maps?

An exploratory case study was conducted during the first phase to explore ChatGPT-4o’s ability to interpret contour maps. A contour map is a map that an educator draws using contour lines to represent imaginary slopes and landscapes in order to help learners identify these elements on topographical maps. Topographical maps are contour maps of real landscapes that are drawn by cartographers. I engaged in six conversations with ChatGPT-4o, based on themes to be covered in Grade 12 Geography classrooms during interpretation of 1:50 000 topographical maps. The results showed that ChatGPT-4o’s theoretical knowledge of contour maps, slopes and landforms was good, but also that it made mistakes and hallucinated convincingly. ChatGPT-4o could not use contour lines to demonstrate different types of slopes, but it could identify four of the types of slopes on a contour map. Its motivation for these slopes was good. However, ChatGPT-4o could not identify and motivate a terraced slope. Although it knew what a concave and a convex slope is, it could not distinguish between the two slopes and hallucinated convincingly. Based on the findings, it could be deduced that ChatGPT-4o can be used to interpret contour maps, as long as the user reads all responses critically to identify and rectify errors.

A quantitative study was conducted during the second phase to better understand if ChatGPT-4o would be adopted to interpret contour maps. For this purpose, the Artificial Intelligence Adoption Prediction Model (AIAPM) was used to calculate an Adoption Prediction Score (APS). Unlike previous studies, I had to allocate half a point if a question of the AIAPM could not be answered by a definite ‘yes’ or ‘no’. Based on the APS (seven out of nine), it could be predicted that ChatGPT-4o would be adopted to interpret contour maps in geography classrooms. Thus, the AIAPM suggests that strategies and policies must be in place to promote responsible use of ChatGPT-4o during interpretation of contour maps. However, responsible use of technologies is only one of twelve LSs to be developed in South African classrooms.

Lastly, design-based research was conducted during the third phase to identify the LSs that should be developed in Geography classrooms while learners use ChatGPT-4o to interpret contour maps. The LSs (consisting of the seven CCFOs and the five DOs) were compared with the 21st century skills to be developed in the American context, future-ready practical life skills (FRPLSs) and the geographical knowledge, skills and attitudes to be developed in Geography classrooms. As seven of the LSs were integrated in the geographical skills, techniques and attitudes to be developed in Geography classrooms, I focused on developing strategies to promote the development of the rest of the LSs while learners use ChatGPT-4o to interpret contour maps. The strategies were designed from a learning-oriented approach assessment and are based on theories related to cooperative learning, self-regulation and self-directed learning. The first strategy is to use cooperative base groups (CBGs) to develop teamwork skills. The second strategy is to create opportunities for self-regulation and self-directed learning, to enable learners to manage themselves and their activities effectively. The last strategy is to create opportunities for learners to explore a variety of learning strategies and to reflect on the efficiency of these strategies. I also indicate which of the other LSs can be developed during application of these strategies.

The research contributes on a practical level by providing theory-based strategies for developing LSs while learners use ChatGPT-4o to interpret contour maps. These strategies may be useful in other contexts as well, for example while learners use ChatGPT-4o for other purposes in Geography classrooms. More research must be conducted to better understand if, and how, ChatGPT-4o is already used in Geography classrooms, and how AI plagiarism can be mitigated.

Keywords: ChatGPT-4o; contour maps; critical crossfield outcomes; developmental outcomes; Geography classrooms; geographical knowledge; life skills; topographical maps

 

  • This article’s featured image was created with Canva.

 

Lees die volledige artikel in Afrikaans

Strategieë om lewensvaardighede te ontwikkel terwyl Geografie-leerders ChatGPT-4o gebruik om kontoerkaarte te vertolk

  • 0

Reageer

Jou e-posadres sal nie gepubliseer word nie. Kommentaar is onderhewig aan moderering.


 

Top