The Effect of Artificial Intelligence-Supported Sustainable Geography Education on the Preparation Process for the IGEO Olympiad

Abstract

This research aims to examine the effect of artificial intelligence (AI)-supported sustainable geography education on the preparation process for the International Geography Olympiad (IGEO). Research was designed according to the simultaneous triangulation design, which is one of the mixed-methods designs. The research is a quasi-experimental model in terms of revealing the effects of independent variables (IGEO) on dependent variables (artificial). In this study, a quasi-experimental design with a pre-test–post-test control group was used. In this mixed-method study, quantitative data were obtained from questionnaires and achievement tests, while qualitative data were obtained from semi-structured interviews with students and teachers. The quantitative data collection tools used in the study were a mapping literacy achievement test and a problem-solving skills perception scale. The data were obtained from students across various class sections of the same school. Qualitative data were collected through semi-structured individual interview forms, observation forms, participant diaries, and focus group interview forms. Hierarchical regression analysis and ANOVA were used to analyze the statistical data, and the inductive analysis technique was used to analyze the qualitative data. The findings show that AI-supported sustainable geography education improves spatial thinking skills, individualized learning, and learning motivation. In the IGEO exam, students answered the field questions.

1. Introduction

According to Edward Fredkin, “There are three great events in history. The first is the formation of the universe. The second is the beginning of life. The third is the emergence of artificial intelligence.” Based on this sentence, it can be said that the potential of artificial intelligence and the point it can go are far beyond what was imagined. It is certain that this technology, which is advancing at a dizzying pace, adds power to education from different perspectives and will create greater momentum in overcoming the problems encountered in the teaching process.

The International Geography Olympiad (IGEO) is a prestigious competition for high school students worldwide, assessing advanced spatial analysis, geographic information systems (GIS), map interpretation, and fieldwork skills. Preparing for such a competition requires not only curriculum-based knowledge but also higher-order thinking and problem-solving abilities.

Artificial intelligence offers new opportunities to enhance student success by creating personalized learning environments. In the information age, geography education is evolving to emphasize not only spatial knowledge but also environmental awareness, sustainability, critical thinking, and multidimensional analysis. IGEO serves as a key platform to evaluate these competencies among young individuals and to indirectly compare geography education systems across countries [1]. Its results provide valuable insights into the effectiveness of national education policies, curriculum designs, and pedagogical approaches.

Turkey has been participating in these Olympiads since 2004 and has achieved significant rankings in various years [2]. However, Turkey’s success in the IGEO has fluctuated when compared to some countries at the global level. This situation can be associated with various factors such as the structure of geography education in Turkey, student selection systems, teacher qualifications, and the up-to-dateness of the curriculum [3]. On the other hand, the high levels of success achieved by developed countries in these exams are mostly based on modern educational methods such as interdisciplinary education approach, applied learning environments supported by field studies, and intensive use of geographic information systems (GIS) [4,5].

In our knowledge-based age, geography education aims not only to provide individuals with spatial knowledge but also to help them gain skills such as environmental awareness, analytical thinking, sensitivity to global problems, and sustainable solution production. In this context, IGEO has become an important platform where high school students are measured internationally based on these skills, and where the effectiveness of countries’ geography education systems can be indirectly observed [1]. The successes demonstrated by students provide important clues about a country’s geography teaching approach, curriculum structure, teacher qualifications, and educational policies [4].

IGEO [6] is an elite academic competition held in a different country each year, measuring the knowledge, skills, and thinking abilities of high school students in the field of geography worldwide. It was first organized in 1996 and is an international geographic competition held in a different country each year. This organization, carried out under the supervision of the IGU, aims to popularize the discipline of geography at a global level, to discover qualified young geographers, and to ensure academic interaction at an international level [1].

IGEO’s exam structure is designed in a multidimensional manner. The competition generally consists of three main parts:

• The Written Exam contains open-ended questions that require analysis, interpretation, and synthesis based on theoretical knowledge. Questions cover areas such as human geography, physical geography, environmental problems, and regional analysis.
• The Visual Test (Multimedia Test) measures spatial perception, geographic interpretation, and problem-solving skills through photographs, maps, graphics, and satellite images.
• The Fieldwork and Practical Assignments consist of practical assignments that assess participants’ skills in field data collection, analysis, and geographic problem solving.

This multi-component structure transforms IGEO from being a solely knowledge-focused competition into a unique platform that evaluates high-level cognitive skills such as critical thinking, analytical questioning, visual literacy, and problem solving (Lambert & Morgan, 2010) [7].

Artificial intelligence-supported sustainable geography education includes a contemporary teaching approach that aims to develop students’ sustainability-based geographic knowledge, skills, and values by using the opportunities offered by artificial intelligence technologies. In this approach, technology is used not only as a tool but also as a support element that deepens learning and provides a perspective compatible with sustainable development aims.

The IGEO results are directly related not only to the individual success of countries but also to the general adequacy of their education systems. For this reason, the IGEO is also considered an “indicator of educational quality” [8]. In particular, the currency of the curriculum, teacher competence, the use of technological tools (e.g., GIS and remote sensing), out-of-school learning activities, and students’ interdisciplinary thinking skills stand out as the main factors determining the level of success in the Olympics [4,9].

In the 21st century, education systems are being restructured with both the impact of digital transformation and the increase in global sustainability problems. In this process, geography education stands out as a critical discipline in terms of developing environmental awareness, acquiring spatial thinking skills, and growing up as responsible individuals towards sustainable development goals (SDGs) [10]. Sustainable geography education aims to enable individuals to comprehend the relationships between the environment, society, and economy in a multidimensional way, while also providing students with skills to produce solutions to global problems on a local scale [11].

In this context, rapid developments in educational technologies, especially the integration of artificial intelligence (AI)-based tools into teaching processes, are transforming the quality of geography education. Artificial intelligence provides a more effective learning experience by providing instant feedback to students in areas such as data analytics, personalized learning, map literacy, and spatial analysis [12,13]. AI-supported learning environments allow students to model complex geographical processes and explore sustainability principles through various scenarios [14].

In particular, those organized at the international level, such as IGEO, provide platforms where students can demonstrate their high-level geographic skills. These competitions require multifaceted competencies such as spatial analysis, problem solving, graphic and map reading, while also measuring sensitivity to global environmental problems. Therefore, examining the impact of AI-supported sustainable geography education practices in the preparation process for Olympiads such as IGEO is an important area of research in terms of both educational technologies and interdisciplinary teaching.

This study aims to examine the contribution of artificial intelligence-supported sustainable geography education to students’ preparation processes for the IGEO Olympiad. The research not only evaluates the effects of digitalization in education on geography teaching but also questions how sustainability-based pedagogical approaches can be integrated with digital tools.

This study is important in terms of showing how digital transformation in education can be used in geography teaching. It also reveals the potential role of AI technologies in raising individuals who will contribute to sustainable development goals. It is thought that the study findings will guide both geography teachers and education policy makers in developing technology-supported sustainable education models and will also contribute to providing an effective preparation model for students who aim for success at the international level.

The main purpose of this research is to “examine the effect of AI-supported sustainable geography education on the preparation process of the IGEO.” In this context, the following questions were sought in the study.

• What is the effect of AI-supported sustainable geography education on the IGEO preparation process?
• What is the level of problem-solving skills among students preparing for the IGEO?
• What is the impact of AI-supported sustainable geography education on students’ problem-solving skills during IGEO preparation?

2. Method

The research was conducted using a mixed-method design. According to the authors of [15], the preference for the mixed method is because the main hypothesis in the research is used together and intertwined to provide a more accurate expression of questions and problems. Thus, the mixed method is based on the principle of collecting, analyzing, and using qualitative and quantitative data in a single study. According to Johnson and Christensen (2008) [16], this method covers both qualitative and quantitative research methods. At the same time, this research was designed according to the simultaneous triangulation design, which is one of the mixed-method designs [17]. Quantitative data were collected with pre-test and post-test achievement tests and perception surveys; qualitative data were collected through interviews.

2.1. Participant Group

For IGEOs from all over Türkiye, the research process was planned as an 8-week activity process, but the activities were extended to 10 weeks to equalize the groups. A total of 5 different activities were used in the research process. With these activities, the participants gained the opportunity to apply the effect of sustainable geography education supported by artificial intelligence on the IGEO preparation process. Thus, 41 students from the experimental and control groups and 41 students from the control group were equalized and included in the research. In the sixth stage, which is the last stage of the sampling process, the participants who were interviewed in the focus group after the experimental process were determined. The multiple hierarchical regression analysis applied during the sample selection is presented below. The sample of this study consisted of 102 students enrolled in high school in the city center of Eskisehir. In accordance with the quasi-experimental design with pre-test–post-test control groups, students were initially assigned to experimental and control groups using a random number table to ensure randomization. However, to control for potential pre-existing differences between groups in terms of academic achievement, hierarchical regression analysis was conducted to create matched groups. In the regression model, students’ academic achievement, parental educational status, previous educational experiences, AI self-sufficiency, GIS experience, and levels of internal and external locus of control were included as independent variables. A significance level of p < 0.05 was used, and the model showed a coefficient of determination of R² = 0.42. Based on the standardized beta coefficients obtained from the regression, two groups of 41 students each (experimental and control) were formed with closely matched predicted values. This approach aimed to minimize individual differences between groups at baseline and to more accurately assess the effect of the experimental intervention (see

Table 1. Results of multiple hierarchical regression analyses.

Regression	Model	Variable	β	t	R	R2	R2 ch	F ch
	1	Al self -sufficiency	0.18	4.03 *
		Parental education status	0.10	2.21 *	0.28	0.08	0.08	14.75 *
PEE		GIS experience	0.11	2.73 *
		Academic achievement	−0.10	−2.98 *
	2	Internal control	−0.18	−4.62 *
		External control	0.08	2.41 *	0.35	0.12	0.04	16.28 *
	1	AI self-sufficiency	−0.17	−7.51 *
		GIS experience	0.10	2.83 *
AISE		Academic achievement	−0.13	−3.47 *
		Parental education status	0.09	2.39 *	0.41	0.17	0.17	23.52 *
	2	Internal control	−0.14	−4.12 *
		External control	0.11	2.98 *	0.45	0.21	0.04	15.78 *
	1	GIS experience	−0.14	−3.63 *
		AI self-sufficiency	−0.12	−3.23 *
		Academic achievement	−0.13	−3.48	0.30	0.09	0.09	11.53 *
PE		Parental education status	−0.08	−1.65
	2	Internal control	−0.10	−2.65 *
		External control	0.22	5.97 *	0.39	0.15	0.06	25.08 *

* p < 0.01.

).

Based on the figure above, a hierarchical regression analysis was conducted for sample selection related to AI-supported sustainable geography education. The analysis examined students’ academic achievement levels, parental education status, GIS experience, and AI self-efficacy. Additionally, internal and external loci of control dimensions were analyzed. It was found that the sub-dimension of prior learning in AI-supported sustainable geography education explained 45% of the variance, AI self-efficacy accounted for 32%, and parental education status accounted for 23%. Furthermore, AI-supported sustainable geography education was found to negatively predict the internal locus of control, while positively predicting the external locus of control. This finding has significant educational implications. Specifically, it suggests that students engaged in AI-assisted learning environments may attribute their academic successes and failures more to external factors, such as technology or instructional design, rather than to their own efforts or abilities. This shift in the locus of control orientation could influence motivation, self-regulation, and ultimately learning outcomes. Therefore, educators and curriculum developers should consider strategies to enhance students’ sense of personal agency and internal control within AI-supported settings, ensuring that technology acts as a facilitator rather than a determinant of student achievement.

Data Collection Tools

In accordance with the nature of the mixed design within the scope of the research, the data collection process was carried out multidimensionally. The data collection tools used in the research, which were shaped according to the simultaneous triangulation design from mixed-method models, were determined as follows. In this research, quantitative data were collected with the “IGEO achievement test (IGEO-AT)” and “Problem solving skill perception scale (PSSPS)” developed by the researcher. In the qualitative dimension of the research, data were collected by the researcher again through multiple data collection tools, namely “focus group interview”, “semi-structured individual interview”, and “participant diaries” with 11th-grade secondary school students (see

Table 2. Independent samples t-test results for the gender variable of participants.

Group	Female/Male	Total
Experiment	24/17	41
Control	19/22	41

).

2.2. IGEO Achievement Test Scale

After the pilot application, item discrimination and difficulty values were determined for data analysis, and the average item difficulty and item discrimination of the developed test, as well as Kuder Richardson-21 (KR-21) and Cronbach Alpha values, were determined as reliability. When the analysis results were evaluated, the 25-item “IGEO-AT” was completed because the item discrimination index was not low in any item. According to Lawshe [18], based on the assessment of 10 experts, none of the items fell below the content validity criterion of 0.62; therefore, no items were excluded from the scale. Through these procedures, the scale achieved statistical content validity. Consequently, the finalized scale form, consisting of 25 items, was deemed ready for implementation.

In the achievement test, the average difficulty (p) was calculated as 0.824, and the average discrimination (r) was calculated as 0.861. Based on the item analysis results, it can be said that an achievement test with high difficulty, high discrimination power, and reliability was obtained. While the KR-21 reliability coefficient was determined as 0.872, the Cronbach Alpha reliability coefficient was determined as 0.981.

The perception scale for problem-solving skills for high school students consists of 20 items in its original form. In order to ensure the construct validity during the scale adaptation process, factor analysis was applied to the data obtained by the researcher. For this reason, EFA and CFA were applied to the data obtained during the construct validity process of the scale. In the exploratory factor analysis of the scale, Kaiser–Mayer Olkin (KMO) coefficient was calculated as 0.876, and the Bartlett test was found to be significant at 0.000.

2.3. Problem-Solving Skills Scale for High School Students

Based on the exploratory factor analysis, it was decided that the scale was grouped under two factors. Based on these results, confirmatory factor analysis was conducted in order to test the two-factor scale structure. In carrying out the confirmatory factor analysis, the Lisrel package program was used. As a result of the analysis, x2 = 520.05, df = 217, p = 0.000 < 0.001; RMSEA = 0.052; X2/df = 3.34; NFI = 0.95; NNFI = 0.91; CFI = 0.92; GFI = 0.92; AGFI = 0.90.

2.4. Semi-Structured Individual Interview Forms

A semi-structured individual interview form was prepared to be applied to the volunteer participants in the experimental group. Interviews are frequently used in qualitative research. Criticisms are mentioned regarding the use of the concept of “data” in qualitative research [19]. Interviews are a frequently used data collection tool and are a source. Semi-structured individual interviews are widely used in qualitative research paradigms and are considered an important data source [20,21].

2.5. Focus Group Interview

In this study, where a focus group interview was also used as a data collection tool, personal opinions of students determined through purposeful sampling regarding the effect of AI-supported sustainable geography education on IGEO and the effect of AI on problem-solving skills were analyzed. The main purpose of the focus group interview is to obtain deep and detailed information with many dimensions regarding the basic qualities, such as the approach, knowledge, skills, experience, perception, emotion, behavior, attitude, and habitual life of the participants related to the subject to be researched (Bowling [22,23,24,25,26]). In total, 12 + 12 participants were selected from the experimental group on a voluntary basis and received high and low scores from the map literacy skill IGEO-AT and the PSSPS. The average duration determined for a healthy focus group interview varies between 1 and 2 h [7]. Sampling: The qualitative component of the study comprised a purposive sample of 24 voluntary participants from the experimental group. Participants were stratified into two homogeneous subgroups based on their scores on the IGEO-AT and the PSSPS: high scorers (n = 12) and low scorers (n = 12). The focus group sessions were conducted with an average duration of approximately one hour per group, consistent with the 1–2 h timeframe recommended in the extant literature [25]. Approximately one hour of focus group interview was conducted for each participant. Coding Procedure: Qualitative data were systematically analyzed using NVivo 15 software. The analytical process commenced with open coding, followed by the aggregation of related codes into thematic clusters. This iterative coding approach facilitated in-depth engagement with the data to derive meaningful categories and emergent themes. Reliability: To ensure coding reliability, two independent coders separately coded the dataset. Intercoder agreement was quantitatively assessed through Cohen’s Kappa coefficient, yielding a value exceeding 0.80, indicative of strong concordance. This high level of agreement attests to the rigor and consistency of the coding procedure. Thematic Saturation: Data collection was concluded upon achieving thematic saturation, defined by the absence of novel themes or codes during successive focus group discussions. This methodological rigor bolsters the trustworthiness and validity of the study’s qualitative findings. Approximately one hour of focus group interview was conducted for each participant.

2.6. Observation

The experimental process of the study was recorded by the researcher by keeping unstructured observation notes. The observation was made to determine whether the activities in the classroom where the experimental process was carried out were suitable for the purpose of the study, to what extent the participants were affected by the map literacy skill activities, and to ensure the credibility of the data obtained from the measurements made with the scale and the interviews made after the application. The lesson hours when the observation was made were recorded by expressing the variables of date, place, and duration.

2.7. Participant Diaries

Another data collection tool used in this study was participant diaries. During the 10-week experimental process, at the end of each activity, participants were asked to write diaries about their expectations for the course and whether they met these expectations. All 41 students in the experimental group participated in the participant diaries, which were collected on a voluntary basis.

3. Experimental Process of the Research

After the post-test results were analyzed, individual interviews were conducted with the students, and the collected data were added to the data set to be analyzed. The process carried out by the researcher is once again explained below with the list of experimental process steps.

Summary.

• The “IGEO-AT” was developed by the researcher as a pre-test and post-test at the beginning and end of the experimental process. At the same time, the “Problem solving skill perception scale for high school students” was also used as a pre-test and post-test. The scale development process was carried out with students from other classes, apart from the main research group.
• Experimental and control groups were randomly assigned.
• In the selection of experimental and control groups, the number of students in both classes was considered and equalized.
• Randomization was implemented in the selection of the experimental and control groups; however, students’ academic achievements, parental educational status, previous educational experiences, AI self-sufficiency, GIS experience, and levels of internal and external loci of control were also considered. Therefore, the study exhibits quasi-experimental characteristics. Furthermore, power analyses were conducted (α = 0.05, 1 − β = 0.80) to mitigate potential misinterpretations arising from small sample sizes and large effect sizes during sample selection.
• The experimental group students were informed about both the IGEO-AT and problem-solving skills in advance, and the benefits of gaining these skills were explained by the researcher.
• Activities for the IGEO preparation process were implemented by the researcher in a 10-week program using artificial intelligence support in geography education.
• The activities were implemented face-to-face in the classroom by the researcher, after obtaining legal permissions, in a gain-oriented manner.
• Individual interview forms were evaluated individually and daily by the researcher.
• The teacher who taught in the control group was interviewed by the researcher to check whether the same learning outcomes were covered and whether the content was not interfered with. The process was carried out by the teacher of that class at the discretion of the parties.
• In the experimental group, activities for the student-centered IGEO-AT, previously prepared by the researcher, were applied, and the lesson plans for these activities, the researcher before each lesson.
• “Individual evaluation forms” were distributed to students so that they could evaluate themselves during these activities. These forms were examined, and the students’ progress was evaluated during the process.
  The aspects they encountered and had the most difficulty in developing, as well as their other gains in the process, were examined.
• Regular weekly meetings were held with the experimental group teacher to evaluate the previous week and plan for the next week.
• In addition to the 3 h weekly schedule of the Social Studies course, the researcher held an additional 1 h meeting with the students and received feedback from the students.
• The experimental group class was divided into groups of a maximum of 4 students, considering the number of students, and each group was given problem situations that they had to solve using their map literacy skills. During the implementation of the experiment, students were guided for varying periods of time for each problem situation, ensuring that they sought solutions to the problem and used their map literacy skills while doing so.
• At the end of the experiment, each group made a presentation in class on a skill they had acquired regarding the use of artificial intelligence in geography education and the solution to the problem they had found.
• The presentations were opened to discussion by other groups within the scope of “critical thinking” skills, and the problem solutions obtained were supported by developing new ideas and different perspectives.

4. Analysis of Data

The data collected in the study were analyzed in a stepwise manner in accordance with the structure of the design used in the research process. As a result of the experimental process, a post-test was applied to the experimental and control groups. First, the skewness value of the post-test scores of the experimental group was determined as 0.312 and the kurtosis value as −0.754; the skewness value of the pre-test scores of the control group was determined as 0.295 and the kurtosis value as −0.876. In the Shapiro–Wilk test, the p-value of the experimental group was measured as 0.740, and the p-value of the control group was measured as 0.643. Since the skewness and kurtosis values of the pre-test scores of the experimental and control groups were between −1.5 and 1.5, and the Shapiro–Wilk test results were higher than 0.05, it was concluded that the distribution was normal. In the next step, the homogeneity status of the post-test scores was examined by using the Levene homogeneity test. According to the test in question, the p-value of the post-test scores of the experimental and control groups was determined as 0.868. Therefore, depending on the statistical result, it was concluded that the post-test scores of the groups showed a normal distribution.

In qualitative research, the analysis of data is to identify patterns and then reveal the relationship between the phenomena [21]. In the inductive analysis process, the researcher identifies codes from data, themes from codes, and findings from themes. In this study, the data collection and data analysis processes were carried out together. The data were analyzed using the NVivo 12 program. Data analyses were carried out simultaneously with the data collection process of the research. In the analyses made after the conclusion of the research, codes were first obtained by coding line by line, and themes were obtained from the codes.

4.1. Ethical Process of Research

Ethical principles were complied with without any process. Participants in the research were involved during the initiation and completion of the research process. Participants in the research were informed that the data obtained from the research would be used only for scientific purposes and that their confidentiality would be protected. Additionally, ethical permission was obtained from the Anadolu University Ethics Committee (date: 2025). Additionally, due to the participants in the study being minors, parental consent was obtained through a parental consent form. Documentation confirming the receipt of parental approval was submitted to the sent form. The data included in the study may be used for educational purposes. However, it may not be used for commercial purposes. The names included in the study may also be used after being changed. This study was approved by the Ethics Committee of Anadolu University (Approval No: 162819, Date: 17 February 2025). Informed consent was obtained from all adult participants and from the parents or legal guardians of minors. All data were anonymized and stored in a password-protected file accessible only to the research team.

4.2. Findings

4.2.1. Findings of the First Sub-Problem

In this section, which is the first sub-problem of the research, the pre-test IGEO-AT results of the experimental and control group students are included.

As shown in

Table 3. Pre-test t-test scores obtained by students in the experimental and control groups on the “IGEO-AT”.

Groups	N	$\overline{\mathbf{X}}$	ss
Experiment	41	96.67	9.72
Control	41	98.95	12.65

, it was determined that the mean pre-test score of the experimental group students to whom the IGEO-AT was applied was 96.67, and its standard deviation was 9.72, while the mean pre-test score of the control group students was 98.95, and its standard deviation was 12.65. Thus, no significant difference was found between the pre-test scores of the experimental group students and the control group students.

When

Table 4. Independent samples t-test results for pre-test scores of experimental and control groups.

Tests	Groups	N	$\overline{\mathbf{x}}$	ss	sd	t	p
Pre-test	Experiment	41	96.91	9.62	39	0.72	0.46
	Control	41	97.93	10.81

is examined, the pre-test score average ($\overline{\mathrm{X}}$) of the IGEO-AT score of the experimental group is 96.91; the pre-test score average of the control group is 97.93. In addition, the t-test applied to the groups was determined as 39 degrees of freedom, the t-value as 0.723, and the p-value as 0.464. Therefore, it was concluded that there was no significant difference between the pre-test results of the experimental and control groups t (39) = 0.72; p = 0.46). In other words, it was determined that the results of the experimental and control groups were similar and that the IBT results of the groups were similar within this scope.

In this section, which is the first sub-problem of the research, the post-test IGEO achievement test results of the experimental and control group students are included.

When

Table 5. The level of map literacy skills of the experimental and control group students according to their post-test t-test scores from the “IGEO-AT”.

Tests	Groups	N	$\overline{\mathbf{x}}$	SS	U	Z	p	Cohen’s d
Final Test	Experiment	41	14.42	723
					167,000	−4.325	0.006 *	0.81
	Control	41	8.24	531

* p < 0.05.

is examined, the post-test mean score of the experimental group for the IGEO-AT questions is 14.42; the post-test mean score of the control group is 8.24. When the post-test mean scores of the participants in the experimental and control groups are taken into consideration, the difference between the groups’ achievement test analysis level mean scores is determined to be significant in favor of the experimental group at the U value of 167.00 and the p value at the 0.006 significance level. Therefore, it is concluded that artificial intelligence-supported sustainable geography education influences the IGEO preparation process. A Cohen’s d value of 0.81 indicates a large effect size. The effect size was large, d = 0.81, 95% CI [0.33, 1.29], indicating a substantial difference between the experimental and control groups.

4.2.2. Findings of the Second Sub-Problem

In this section, which is the second sub-problem of the research, the pre-test problem-solving skill level (PSS) results of the experimental and control group students are included.

As shown in

Table 6. Findings regarding the PSSPS of the experimental and control group students according to their pre-test, independent samples t-test scores from the “problem solving skills scale”.

Groups	N	$\overline{\mathbf{X}}$	ss
Experiment	41	101.83	8.865
Control	41	100.21	9.436

, it was determined that the mean pre-test scores of the experimental group students to whom the problem-solving skills perception scale (PSPS) was applied were 101.36, and its standard deviation was 8.865. It was concluded that the mean pre-test scores of the control group students to whom the PSSPS was not applied were 100.21, and its standard deviation was 9.43.

When

Table 7. Independent samples t-test results for post-test scores of experimental and control groups.

Tests	Groups	N	$\overline{\mathbf{x}}$	ss	sd	t	p
Pre-test	Experiment	41	92.91	9.62	39	0.75	0.46
	Control	41	96.54	11.84

is examined, the pre-test score average of the experimental group for the IBM SPSS Statistics 30 questions ($\overline{\mathrm{X}}$) is 92.91 and the pre-test score average of the control group is 96.54. In addition, the t-test applied to the groups was determined as 39 degrees of freedom, t-value as 0.745, and p-value as 0.460. Therefore, it was concluded that there was no significant difference between the pre-test results of the experimental and control groups, t (39) = 0.75, p = 0.46.

When

Table 8. Findings regarding the problem-solving skills of the experimental and control group students according to their post-test independent samples t-test scores from the “problem solving skills scale”.

Tests	Groups	N	$\overline{\mathbf{x}}$	SS	U	Z	p	Cohen’s d
Final Test	Experiment	41	13.92	814
					184.000	−4.312	0.007	0.79
	Control	41	6.28	437

is examined, the mean score of the experimental group in the problem-solving skill scale is 13.92; the mean score of the control group is 6.28. When the mean scores of the participants in the experimental and control groups are taken into consideration, the difference between the groups’ problem-solving skill perception scores mean scores is determined to be 184.000, and the p-value is determined to be significant at the significance level of 0.007 in favor of the experimental group. In other words, it has been determined that the problem-solving skills of the experimental group participants have positively differed with the artificial intelligence-supported geography education. Similarly, a Cohen’s d value of 0.79 indicates a large effect size. The effect size was large, d = 0.79, 95% CI [0.31, 1.27], indicating a substantial difference between the experimental and control groups.

4.2.3. Findings of the Third Sub-Problem

In this section, which is the third sub-problem of the research, regression analysis of the pre-test results of the effect of artificial intelligence-supported geography education on problem-solving skills of students preparing for IGEO in the experimental and control groups is included.

The sub-dimensions of the IGEO activities (map interpretation skills, perception of spatial distribution, determination of location on the map, preparation and interpretation of tables, graphics, diagrams, problem-solving on the map, map knowledge and perception of time, and change and continuity on the map) are the dependent variables of the research. The effect of the artificial intelligence support variable was analyzed by multiple regression analysis (

Table 9. Regression analysis of the effects of artificial intelligence-supported IGEO activities applied to experimental group students on students’ problem-solving skills.

Sub-Dimensions	Unstandardized Coefficients		Standardized Coefficients	T	p
	B	Std. Error	Beta
Map interpretation skills	3.700 0.197	0.313 0.071	0.234	13.207 3.319	0.000 *
Perceiving spatial distribution	0.232	0.062	0.212	2.324	0.000 *
Locating on the map	0.232	0.267	−0.001	−4.307	0.000 *
Preparing and interpreting tables, graphs, and diagrams	0.396	0.187	−0.302	−2.201	0.000 *
Solving problems on a map	0.142	0.62	−0.256	2.546	0.000 *
Map information	0.156	0.62	0.187	3.091	0.000 *
Perceiving time, change, and continuity on the map	0.238	0.154	0.249	2.742	0.000 *
R	0.874
R2	0.862
Adjusted R2	0.857
Standard error	0.511
F	12.452				0.000 *
Durbin–Watson	1.63

* p < 0.01.

). According to the findings, it was determined that the multiple regression model established was significant (F = 12.452, p < 0.01) and explained 84.7% (R² = 0.862) of the effect of map literacy skills on problem-solving skills.

4.3. Findings Regarding the Interview Questions Obtained from the Participants

In this section, student opinions are directly included.

“…The use of artificial intelligence in sustainable geography lessons has made our lessons fun…”

“…IGEO is a very big and difficult exam, but artificial intelligence makes our job easier…”

“…Solving problems is now easier. Because there is artificial intelligence …”

“…Learning the geography of countries we have not visited or seen with artificial intelligence is very helpful…”

“…Sustainable geography education not only prepares you for IGEO, but also prepares you for living…”

5. Conclusions

In this section, separate results for the three sub-problems of the research are given.

Results on the Impact of Artificial İntelligence-Supported Sustainable Geography Education on the IGEO Exam Preparation Process

The study was conducted in experimental and control groups before a pre-test was applied. No significant difference was found in the preparation process of the artificial intelligence-supported geography education of the experimental and control groups for the IGEOI exam. After the experimental process, a post-test was applied, and it was found that there was a significant difference between the problem-solving skills of the experimental and control groups in favor of the experimental group.

The pre-test and post-test scores of the experimental and control groups were compared with each other, and it was determined that there was a significant difference between the pre-test and post-test scores of the experimental group, while there was no significant difference between the pre-test and post-test scores of the control group. It was determined that there was no such problem. Therefore, it was concluded that artificial intelligence-supported geography education implemented in the 11th-grade geography course increased success in the preparation process for IGEO.

This study examined the effects of artificial intelligence-supported sustainable geography education on the IGEO preparation process and revealed that it provided significant improvements in students’ higher-order thinking skills, geographic literacy levels, and problem-solving abilities. Participating students stated that they felt more competent in the basic skill areas of IGEO, especially map interpretation, graphic analysis, and relating environmental problems. In addition, the interactive presentation of sustainability-based content with artificial intelligence increased both interdisciplinary thinking and awareness of issues such as the climate crisis, disaster management, and urbanization in students [27,28].

According to the results of the study, educational processes carried out using artificial intelligence tools have gained a more dynamic, student-centered, and critical thinking-based structure compared to traditional methods. This has supported both the individual learning speed of students and strengthened the interdisciplinary knowledge integrity in preparation for high-quality exams such as IGEO [12]. Therefore, sustainable geography education supported by artificial intelligence has ceased to be a structure that only transfers information and has built a learning environment in which students are active participants.

Results Regarding the Problem-Solving Skill Level of Students in the IGEO Exam Preparation Process

Intelligence during the IGEO preparation process contributed to the level of awareness regarding their problem-solving skills. Among the results obtained is that students’ problem-solving skills have made significant progress in gaining self-confidence. It has been determined that students are now more successful in solving problems that they previously had difficulty with after the interviews.

The findings of this study show that artificial intelligence-supported geography education has a significant and multifaceted effect on students’ problem-solving skills. It has been observed that students can perform spatial analysis, interpret data, and develop solution alternatives by interacting with artificial intelligence tools (e.g., Google Earth Engine and ArcGIS AI), especially in solving geographic scenarios. This shows that students do not only remember information but also become active in Bloom’s higher-level cognitive stages, such as analysis, synthesis, and evaluation [29].

Similarly, Bednarz [30] and Jekel et al. [31] state that the integration of geography education with artificial intelligence technologies supports students’ problem-solving thinking structure and especially provides development in spatial literacy skills. This study also reveals similar results; a significant increase in students’ systematic and multidimensional thinking skills was observed, especially when analyzing sustainability-based geographic problems.

AI-supported learning processes increase students’ motivation and develop a more creative and critical perspective in their search for solutions are also confirmed in studies conducted by Holmes et al. [12]. In addition, this study found that AI-supported activities allow students to develop versatile solutions in both individual and collaborative learning environments. This is directly related to the cognitive flexibility and applied problem-solving competencies required in exams such as IGEO.

On the other hand, it was noted that some students had difficulty interpreting artificial intelligence tools and needed guidance during this process. This finding coincides with Luckin et al. [32]’s emphasis that artificial intelligence may not always be effective for students without teacher guidance. Therefore, it can be said that a structured learning environment should be created under teacher guidance for artificial intelligence-supported geography education to be effective.

In conclusion, this study shows that artificial intelligence technologies are not only a technological innovation in geography education, but also a pedagogical transformation tool based on problem-solving. In this respect, the study offers an important perspective on how new-generation geography education should be structured in preparation for high-level applications such as IGEO.

Results on the Effect of Artificial Intelligence-Supported Sustainable Geography Education on Problem-Solving Skills of Students Preparing for the IGEO Exam

The results of the research also included the fact that students who were in a state of stress at the beginning of the research process were able to manage the crisis and produce solutions for coping with stress.

By working in groups, students have learned to work together, in other words, to move forward in the same boat, based on the principle of aggregation that is appropriate to the nature of the geography course. Learning to work cooperatively in a group despite their differences is also one of the results of the research. It can be said that students have made progress in working together. Students who could not come together and work together before can now work in groups and successfully contribute to themselves and the course. Students who learned group dynamics have also experienced the happiness of being able to work like an orchestra.

Students who gain creative thinking skills gain problem-solving skills with artificial intelligence-supported geography education, and the result that they also have the capacity to produce projects is significant among the research results. The emphasis that they have become individuals who can produce solutions to the problems they experience with projects is also among the research results. At the same time, students have reached a level where they can struggle with the difficulties in life, in addition to just preparing for the IGEO exam.

They stated that, like the research, students’ problem-solving skills improved in artificial intelligence-supported geography courses and that it also facilitated the preparation process for the IGEO exam [13,14,33]. As a result of their studies, they stated that since map skills are related to geographic inquiry skills, having a paradigm suitable for project-based learning increased students’ interest in the course, and students who previously had low motivation for inquiry and problem-solving-based activities developed positive attitudes towards geography and social studies courses and studies.

Researchers state that digital game-based education uses students’ knowledge and competencies in problem-solving, decision-making, and active learning [34,35,36]. Digital games have significant potential for many disciplines [9]. Digital games help solve complex problems. It is also among the results that the individual’s spatial skills improve in terms of self-awareness and self-confidence [33,37].

According to Mayo [18], digital games, unlike mass media tools used in education, are a tool that allows individuals to develop a multifaceted perspective and are also highly interactive in terms of pedagogy [26,37]. The use of games in the education–training process still creates prejudice. However, the research conducted draws attention to a point beyond these stereotyped views. While educational digital games accelerate access to information in the education-training process, they have made it difficult to forget the learned information. In addition, they have increased the level of retention of the learned information [18]. The results of the research include that the use of artificial intelligence support in the preparation process of the IGEO exam and all similar exams, such as PISA and TIMMS, not only increases the students’ academic success, but also increases the level of problem-solving skills and allows individuals to acquire 21st-century skills.

The findings show that artificial intelligence technologies are not limited to the use of digital tools in the context of geography education but also create a pedagogical transformation. Studies conducted specifically for IGEO exams aim to provide students with skills such as “data-based decision making,” “spatial relationship building,” and “systematic thinking” [30]. In this context, it has been seen that artificial intelligence-supported applications offer students the opportunity to simplify complex processes such as data interpretation and geographic modeling.

Before AI-supported sustainable geography education, students’ perceptions of a sustainable environment were characterized by neglecting socio-economic and environmental dimensions, difficulty in relating sustainability to daily life, and a weak understanding of recycling [38]. However, after the educational process, students’ perceptions of sustainability changed significantly. For example

• They began to evaluate sustainability through its environmental, social, and economic dimensions.
• They acquired the ability to develop solutions for sustainable environmental issues at local and global scales.
• They discovered that artificial intelligence technologies can contribute to sustainability.
• They developed a perception of behaving more harmoniously with nature.
• A sustainable awareness toward reducing carbon footprints emerged.

Significant changes in students’ perceptions of sustainability were observed following the AI-supported education process. Through AI-supported content analyses, interactive scenarios, and personalized learning pathways, students started to internalize sustainability as a multidimensional concept.

Artificial intelligence in education in areas such as learning analytics, predictive modeling, and personalized guidance has been confirmed in the literature [18]. In this study, it was concluded that activities carried out with tools such as Google Earth Engine and ArcGIS AI improved students’ creative solution-producing skills both individually and in groups. In terms of the development of high-level cognitive skills (analysis, synthesis, evaluation) expected by IGEO, the integrated and guided use of these technologies plays a critical role [31].

Student views have shown that sustainable geography education conducted with artificial intelligence supports not only exam success but also value-based gains such as environmental awareness, social responsibility and global citizenship awareness [39]. In this respect, the study shows that artificial intelligence can be used as a strategic tool in raising individuals equipped with 21st century skills and high environmental and ethical sensitivity.

It would be more beneficial to increase the quality of education and training through smart systems or embedded systems without relying solely on computers [40]. By creating smart classes, instant monitoring of student and teacher interactions through sensors, and instant monitoring of students’ motivation for the lesson will undoubtedly increase success. Although artificial intelligence offers many exciting developments, especially for improving education in the world, it can be said that it is still in the early stages of its use. It can be stated that more trials and research should be conducted for artificial intelligence tools to be successfully applied in educational institutions.

Future studies should consider employing longitudinal designs to examine the long-term effects of artificial intelligence-supported education on students’ learning outcomes and problem-solving skills. Expanding the research to include diverse educational settings and subjects beyond geography would enhance the generalizability of the results. Furthermore, comparative studies investigating different AI tools and pedagogical approaches could provide insights into best practices for integrating AI in education. Finally, exploring the impact of AI-supported learning on student motivation, attitudes, and collaborative skills would offer a more comprehensive understanding of its educational benefits.

At the same time, what individuals can do for a sustainable nature can be developed starting from the basic learning levels of students. Similar studies can be developed for different types of exams within the context of sustainability. Finally, it is recommended that similar studies be conducted based on different student groups and different variables.