AI and Q Methodology in the Context of Using Online Escape Games in Chemistry Classes

Abstract

The contemporary digital era has fundamentally reshaped pupil education. It has transformed learning into a dynamic environment with enhanced access to information. The focus shifts to the educator, who must employ teaching strategies, practices, and methods to engage and motivate the pupils. New possibilities are emerging for adopting active pedagogical approaches. One example is the use of educational online escape games. In the theoretical part of this paper, we present online escape games as a tool that broadens pedagogical opportunities for schools in primary school chemistry education. These activities are known to foster pupils’ transversal or soft skills. We investigate the practical dimension of implementing escape games in education. This pilot study aims to analyse primary school teachers’ perceptions of online escape games. We collected data using Q methodology and conducted the Q-sort through digital technology. Data analysis utilised both the PQMethod programme and ChatGPT 4-o, with a subsequent comparison of their respective outputs. Although some numerical differences appeared between the ChatGPT and PQMethod analyses, both methods yielded the same factor saturation and overall results.

1. Introduction

In today’s digital age, education faces new challenges that require substantial innovations to traditional teaching methods. Pupils are increasingly losing interest in traditional forms of education, resulting in shorter attention spans and greater passivity in the classroom. This trend compels educators to seek innovative ways to engage and motivate pupils to learn.

In recent years, technological advancements have driven the adoption of digital technologies in many areas, including education. This transition was accelerated due to COVID-19, as restrictions forced educational institutions to adopt strategies and tools suitable for a digital environment (Mondragon-Estrada et al., 2023).

The rapid and sudden transition from in-person to online instruction is associated with challenges and limitations, but also opportunities. Teachers select teaching methods and implement educational activities. They monitor progress, choose and use digital resources, evaluate pupil contributions, foster a positive learning environment, encourage discussions, present content, and provide feedback (Nikolopoulou & Kousloglou, 2022).

Education experts constantly seek innovative teaching and learning strategies and tools adapted to the reality beyond the classroom. These approaches should integrate values and skills that prepare pupils as future citizens, including determination, teamwork, problem-solving, respect, and above all, motivation and interest. Examples include debates, theatrical plays, question and answer sessions, problem-solving tasks, research projects, observation and participation in didactic itineraries, and gamification (Hunt-Gómez et al., 2020).

One of the approaches that has gained attention in the modernisation of the educational process is active learning. This approach, based on the active participation of pupils in learning, not only promotes their interest in the subject matter but also leads to better acquisition and longer-term retention of knowledge (Fraguas-Sánchez et al., 2022). The rise in digital technologies has created new opportunities to implement active learning methods. These include the use of educational online escape games.

Online escape games are an innovative educational tool that combines elements of game design, education, and technology (Huraj et al., 2022). These games immerse pupils into an interactive environment where they have to solve a series of tasks and puzzles, thereby revising and reinforcing the knowledge they have acquired (Kuo et al., 2022). Online escape games not only support active learning but also develop teamwork, critical thinking, and problem-solving skills. Their use is particularly beneficial in science subjects such as chemistry, where the complexity and abstract nature of the curriculum can lead to a loss of pupil interest.

Implementing online escape games helps make learning more engaging. It transforms content into interactive challenges that increase pupils’ motivation and engagement.

This article provides an introduction to game-based learning and escape games, including the presentation of a novel online escape game titled ‘Chemically Pure Substances and Mixtures’. We follow the escape game with a description of the methodology—specifically, Q methodology—and the selected data processing methods, namely ChatGPT and the PQMethod programme. Following the presentation of results, Section 8 compares the outcomes obtained using a classical approach (the PQMethod programme) with those from a novel approach utilising artificial intelligence, specifically ChatGPT. Furthermore, it interprets Factor 1 and Factor 2, addresses the research limitations, its strengths, and future directions.

2. Game-Based Learning

Game-based learning is an effective complement to teaching in schools. It supports the development of learning experiences and helps pupils connect prior knowledge with knowledge gained through play (Pivec & Dziabenko, 2004, as cited in Stojanovska & Velevska, 2018).

Becker (2021) states that game-based learning can be viewed as an approach to teaching with a specific educational goal in mind. The learning task is designed to make learning more interesting and effective.

In an effective play-based learning environment, pupils work towards a goal, make choices, and experience the consequences of those choices. In a safe environment, pupils can experiment and learn actively through making mistakes, thereby practising good behaviour and thought processes. These activities keep the pupils engaged and allow them to transfer the skills learned to real life easily. The main characteristics of play-based learning are as follows:

• It takes place in different scenarios;
• It is based on overcoming challenges;
• It is positive and interesting;
• It creates real-life-like situations that allow children to practise their skills effectively (Andelová, n.d.).

In the case of active learning, the player must make active efforts through various activities to acquire knowledge and skills. The players choose the games based on the knowledge and skills they want to acquire (Gordillo et al., 2021).

Teachers can incorporate games into chemistry lessons either as a component of active learning or as a tool for revision. They can serve as a fun way to introduce a new unit of learning or as a means of reviewing and refining knowledge and practical skills that have been previously taught (Stringfield & Kramer, 2014, as cited in Stojanovska & Velevska, 2018).

The terms “game-based learning” and “gamification” are often mistakenly interchanged, although they represent two distinct teaching strategies. The term gamification, first introduced by developer Nick Pelling in 2002, refers to the use of game elements in a non-game context. The term describes a phenomenon that began to emerge in marketing practice even before the term became common (Fiala, 2019). Gamification implements individual game elements, rather than the whole game itself, outside the game environment (al Fatta et al., 2019).

In practice, online educational escape games are mainly classified under game-based learning. Their main goal is to directly impart knowledge or skills while playing, as opposed to gamification, which focuses on increasing motivation through game elements.

3. Escape Game

Escape games are a relatively new phenomenon in both educational and entertainment settings. They have gained popularity due to their ability to develop teamwork, critical thinking, and problem-solving skills. Although international studies have already addressed this topic, researchers in the Slovak and Czech academic environments have devoted very little attention to this phenomenon. This lack of research highlights the need for deeper exploration of online escape games, particularly in educational contexts.

Escape games, escape rooms, exit games, and puzzle games are all terms referring to a game where the aim is to solve tasks and puzzles in order to escape a room (Nicholson, n.d.).

Depending on the setting, escape games can be divided into two types:

• Real escape games (IRL—In Real Life): Played by pupils face-to-face in various environments, with or without props, or with QR codes;
• Online escape games (OEG): Moved entirely into the virtual world, both in form and organisation (Vaněk, n.d.).

Escape games involve a team of players entering a physical or digital space where they search for clues to solve puzzles, riddles, or mysteries with the aim of solving a particular problem. The objective of escape games is to develop pupils’ skills in a complementary way to other teaching methods (Eukel & Morrell, 2020).

Escape games promote the development of various skills, including teamwork, communication, logical thinking, critical thinking, information seeking, observation, reasoning, pattern recognition, problem-solving, creativity, application of knowledge, and coping with time pressure (Pan et al., 2017).

Morrell and Eukel (2020) view escape games as an effective educational tool that can bridge the generation gap between pupils and teachers.

Chicca and Shellenbarger (2018) claim that escape games are a technology helping to bridge the gap between traditional educational methods and the demands of new generations of learners who require innovative approaches.

Educational escape games offer an ideal environment for active learning. They require participants to utilise their problem-solving skills and progress through the various challenges included in the activity. By facing a variety of puzzles, pupils test their knowledge and develop skills, especially problem-solving skills (Huang et al., 2020). Makri et al. (2021) characterise educational escape rooms as pedagogical activities with defined problems and time constraints that require active participation and collaboration. These activities provide opportunities for active improvement of the learning process.

There are several pedagogical reasons why educational escape games are an engaging way to learn. The main reasons are the development of the so-called transversal or soft skills listed in the diagram in Figure 1 (School Break, n.d.):

4. Related Work

The usefulness of escape games in the teaching process was examined in a descriptive and interpretive exploratory study conducted at the University of Seville, Spain (Hunt-Gómez et al., 2020) to analyse prospective teachers’ perceptions of escape rooms. The results indicated that female pupils highly valued the playful part of escape rooms as a didactic activity that promotes significant learning. They considered escape rooms as an excellent educational resource as well as an appropriate alternative teaching methodology.

The 2021 study on teachers’ perceptions and experiences of escape room play as a learning environment found that, regardless of age, gender, and teaching experience, they found it attractive. During the game, competition and prizes stimulate pupils. Pupils identified the variety of puzzles and the development of teamwork skills as key success factors of classroom escape rooms (Veldkamp et al., 2021). Bray and Dieckmann (2024) reported in their study that participant–educators generally responded positively to escape rooms and recognised their potential in providing personalised learning experiences. Although educators expressed enthusiasm, providing support in the design and implementation of escape rooms remains essential.

Q methodology is a well-established research method primarily employed in various fields, including social sciences, environmental sciences, and medicine (Dieteren et al., 2023). It also holds significant utility in educational research. For instance, Demir (2016) used Q methodology in their study, “An Analysis of Pre-Service Teachers’ Attitudes and Opinions Regarding the Teaching Profession,” with data processed using the PQMethod programme (Demir, 2016). Similarly, Chaaban et al. (2023) applied Q methodology to explore teachers’ perspectives on career development, relying on the PQMethod programme for data analysis (Chaaban et al., 2023).

Q methodology is also valuable for understanding perceptions related to online games and gamification. Chen et al. (2010) investigated “Perception of young adults on online games using Q methodology.” Their findings, processed with the PQMethod programme, revealed that most participants held philosophical objections to online gaming. However, they viewed online games as a social activity and an interactive medium, with particular concerns raised by female participants requiring careful consideration (Chen et al., 2010). In another study, Yıldırım (2017) explored “Pupils’ perceptions about gamification of education using Q methodology.” The data, analysed via the PQMethod programme, indicated a generally positive perception among pupils regarding educational gamification. Key elements identified in this process included the logic of the procedure, emotions associated with it, advancement structure, achievement points, and badges (Yıldırım, 2017).

Recent advancements have also seen the integration of artificial intelligence with Q methodology. Dobečková et al. (2024) compared the application of ChatGPT for five steps of Q methodology against traditional approaches, including the use of the PQMethod programme. These steps encompassed the creation of Q-samples, the formation of correlation matrices, factor analysis, and factor interpretation (Dobečková et al., 2024). Furthermore, Ramlo (2025) explored the utility of ChatGPT specifically in generating Q-samples.

5. Online Escape Game Chemically Pure Substances and Mixtures

When designing an escape room game, it is crucial to balance entertainment with pedagogical objectives. Authors and educators who develop such games must ensure that puzzles and tasks support specific learning outcomes while being challenging enough to motivate pupils towards problem-solving. Another key aspect is a clear and transparent game structure that allows pupils to track their progress and stay oriented throughout the activities (Wiemker et al., 2015).

The online escape game we developed targets pupils in the seventh year of primary school and those in the second year of eight-year grammar schools. The game reinforces the lessons within the thematic unit “Chemically Pure Substances and Mixtures.”

When creating the online escape game, we followed the requirements established by the ISCED 2 state educational programme (effective from 1 September 2015), as well as the educational standards and the content of the 2017 textbook Chemistry for the 7th year of primary school and the 2nd year of eight-year grammar school study. This escape game aims to review and consolidate the pupils’ knowledge of the thematic unit “Chemically Pure Substances and Mixtures”. The thematic unit encompasses the following topics:

Composition of substances, Mixtures, Solutions, Expressing the composition of solutions, Mass fraction, Mass fraction in solved tasks, Separation of components from mixtures, and Separation of components from mixtures in practice. The escape game is available in Slovak and can be accessed via a web link or by scanning a QR code (see Figure 2).

Upon clicking, pupils are directed to the introductory page (see Figure 3). On the left side of the introductory page, there is introductory information about the escape game. The text begins with a welcome message, explaining that the image on the right shows a chemistry classroom where pupils will have their lesson. The instructions then follow: “There are several clues hidden in the classroom. In order to get the clues, you have to solve the given problems. After completing all the problems, you will get a password which will open the door and you can go to break with a good feeling that you have done it perfectly:)” This is followed by a wish for success. Finally, the command is given: “Click on START and hurray for the first clue!”

The authors designed the escape game to be flexible, enabling adaptation to different age groups and knowledge levels. We utilised digital tools and multimedia elements that enhance the attractiveness and interactivity of the games. Each escape game is composed of six activities, assignments, and tasks. The tasks were created using various online digital platforms (e.g., https://learningapps.org/ (accessed on 1 January 2024), https://wordwall.net/sk (accessed on 1 January 2024)). The main game page contains interactive spots (see Figure 4). Clicking on these spots leads to a specific task.

The entire game system has its own structure and rules. After clicking “start”, the first clue appears (“All substances are composed of particles.”), which leads us to the first task. Each task begins with a theoretical overview of the given topic. At the end, there is always an image of a beaker, which functions as a link, and when clicked, it takes us to the assignment of the next riddle, crossword, or task. For example, for the topic “Separation Methods”, pupils’ task is to correctly assign the name of the separation method shown in Figure 5.

The correct solution to each task is linked to a clue that will unlock the next activity. Upon correctly solving the activities, a password to open the door is obtained. The pupils then return to the classroom and click on the door handle, virtually leaving the chemistry classroom. They are then redirected to a Padlet board where they can express their feelings or opinions about playing the game (see https://padlet.com/lucboldisova/n-stenka-u-eb-a-ch-mie_2-sggkuw91wrb52lxi, accessed on 21 May 2025). The opinions of Slovak elementary school students after implementing the escape game are displayed there.

6. Methods

The aim of the presented qualitative research was to gather the opinions of primary school teachers regarding escape games, in general, and specifically about newly developed escape games. Q methodology was chosen as the method of data collection due to the nature of the research, since Q methodology emphasises small numbers of participants (McKeown & Thomas, 2013). Q methodology is particularly recommended for exploring novel research domains and serves heuristic purposes (Chráska, 2016). Participants are required to rank-order statements into a quasi-normal distribution (pyramidal array), a process that effectively captures subtle nuances in their perspectives. This makes Q methodology more appropriate for this research than conventional questionnaire surveys. It circumvents the limitations of traditional surveys, such as the tendency for participants to select neutral or “do not know” responses, by compelling them to make forced-choice distinctions between more and less agreeable statements. This approach thereby yields richer insights into their subjective viewpoints on the presented problem.

The results were processed in two ways and subsequently compared. One analysis was performed using PQMethod and the other using artificial intelligence, specifically ChatGPT 4-o (accessed in August 2024). A total of five Q-sorts were compared. The entire pilot study was anonymous. Respondents were Slovak school chemistry teachers with varying lengths of experience who were willing to use and implement the online escape game in their teaching. Three of the respondents were not familiar with escape games and had never used them in an educational setting. The other respondents were aware of escape games. However, they did not specify whether they also used them in an online form or employed them regularly and purposefully in chemistry classes. Respondents were not surveyed for their level of digital competence.

Given the limited number of respondents (n = 5), this qualitative research is characterised as a pilot study and must be interpreted as such. The findings provide preliminary insights rather than generalizable conclusions.

A Facebook group of chemistry teachers was used to reach the respondents. Other authors have taken a similar approach in their research, e.g., Rusek et al. (2022).

Respondents performed a Q-sort using Microsoft Excel. On the first sheet, respondents had instructions, and on the second sheet, a pyramid for Q-sort was prepared, along with statements on a yellow background inserted as pictures (inspired by (Dobečková, 2023)).

Given that the respondents were teachers who regularly employ Microsoft Word, Excel, and PowerPoint in their pedagogical practice, a familiar application from this suite was selected for the Q-sort administration. This strategic choice aimed to streamline participant engagement by eliminating the need to learn unfamiliar software or online environments, thereby minimising potential disincentives for research participation.

The respondents’ prior knowledge of the scoring system enabled them to make more deliberate and precise decisions when placing the statements in the prepared scheme (see Figure 6).

The Q-sorting data were first processed according to Chráska (2016), producing two tables: one containing the statements with the highest level of agreement and the other containing the statements with the highest level of disagreement (see Table 1 and Table 2). The results were then processed using the PQMethod programme (Schmolck, 2014), which calculated the correlation matrix and performed principal component analysis (hereinafter referred to as PCA). For comparison, the results were processed by an artificial intelligence tool, specifically ChatGPT 4-o, which created the correlation matrix and performed PCA with data interpretation. The prompt was as follows (the original prompt was written in Czech):

“Hi, could you please perform a factor analysis? Please (1) determine whether the data is suitable for factor analysis using (a) Kaiser-Meyer-Olkin measure of sampling adequacy (b) Bartlett’s test of sphericity—chi square test, number of degrees of freedom, significance (c) communalities (d) using Pearson coefficients (2) perform a factor analysis in which you include a) a table with factor loadings (b) a table in which you mark with the symbol X the factor loadings exceeding the value of the significant correlation measure, i.e., 0.360 (c) a table with factors in which the eigenvalue, S.E. z-score of the factor, number of defining variables, average reliability coefficient, composite reliability, correlation between factor scores of the factors will be listed (3) for individual factors calculate the z-score and find out the characteristic significant statements for individual factors, i.e., factors to which respondents assigned extreme values to the given factor, while the ratings between the factors differed significantly. (4) interpret the factors to which respondents assign a high level of importance? Thank you.”

Subsequently, ChatGPT was asked additional questions to obtain additional data, including “Could you do a PCA?” and “Could you please create a table with factor loadings?”.

Previous research, such as Dobečková et al. (2024), has also used ChatGPT to process data obtained through Q methodology.

7. Results

The Q-sorting data were first processed according to Chráska (2016) into a table containing statements with the highest level of agreement ordered from the highest average rating by respondents (see Table 1) and a table containing statements with the highest level of disagreement ordered from the lowest average rating by respondents (see Table 2).

Table 1. Statements with the highest level of agreement.

Ranking	Q Statements (Number and Statement)	Average Rating	Standard Deviation
1.	45. The terminology in the OEG is consistent with the terminology used in the subject of chemistry at primary school.	2.2	1.30
2.	48. Pupils repeated their knowledge from the thematic unit in a fun way through OEGs.	2.0	1.00
3.–5.	37. OEGs can be a tool to make the teaching of chemistry more attractive.	1.8	1.30
	50. OEGs are a way to increase pupils’ motivation.	1.8	1.10
	51. Online escape games encourage active learning methods and the use of diverse teaching strategies.	1.8	0.45
6.	31. I find repetition through OEG an interesting way of repeating a thematic unit.	1.6	1.52
7.–10.	28 The repetition through OEG was interesting for the pupils.	1.4	1.67
	30. The use of the teaching material in the lessons met my expectations.	1.4	1.34
	38. OEGs increase pupils’ interest in repetition.	1.4	1.52
	44. OEGs are aligned with the content performance standard.	1.4	1.52

OEG = abbreviation for online escape game.

Table 1. Statements with the highest level of agreement.

Ranking	Q Statements (Number and Statement)	Average Rating	Standard Deviation
1.	45. The terminology in the OEG is consistent with the terminology used in the subject of chemistry at primary school.	2.2	1.30
2.	48. Pupils repeated their knowledge from the thematic unit in a fun way through OEGs.	2.0	1.00
3.–5.	37. OEGs can be a tool to make the teaching of chemistry more attractive.	1.8	1.30
	50. OEGs are a way to increase pupils’ motivation.	1.8	1.10
	51. Online escape games encourage active learning methods and the use of diverse teaching strategies.	1.8	0.45
6.	31. I find repetition through OEG an interesting way of repeating a thematic unit.	1.6	1.52
7.–10.	28 The repetition through OEG was interesting for the pupils.	1.4	1.67
	30. The use of the teaching material in the lessons met my expectations.	1.4	1.34
	38. OEGs increase pupils’ interest in repetition.	1.4	1.52
	44. OEGs are aligned with the content performance standard.	1.4	1.52

OEG = abbreviation for online escape game.

Table 2. Statements with the highest level of disagreement.

Ranking	Q Statements (Number and Statement)	Average Rating	Standard Deviation
1.	10. The lessons have been processed in a cluttered manner.	−2.6	1.34
2.–3.	13. The assignment for each task was inadequate.	−2.4	0.89
	3. OEGs do not belong in the primary school curriculum.	−2.4	1.34
4.	1. I consider OEGs in the classroom to be a waste of time.	−2.2	1.30
5.–6.	9. The design of the OEG was not age-appropriate for the pupils.	−2.0	1.87
	2. OEGs are uninteresting for pupils.	−2.0	1.58
7.–8.	12. The OEG used inappropriate terminology and terms inappropriate to the age and knowledge of the pupils.	−1.8	1.30
	11. The tasks were challenging. They did not match the knowledge and abilities of the pupils in the given grades.	−1.8	0.84
9.–10.	41. The OEG is difficult to operate and navigate.	−1.6	3.21
	34. Trying out the prepared OEG encouraged me to try to create my own online escape game.	−1.6	1.82

OEG = abbreviation for online escape game.

Table 2. Statements with the highest level of disagreement.

Ranking	Q Statements (Number and Statement)	Average Rating	Standard Deviation
1.	10. The lessons have been processed in a cluttered manner.	−2.6	1.34
2.–3.	13. The assignment for each task was inadequate.	−2.4	0.89
	3. OEGs do not belong in the primary school curriculum.	−2.4	1.34
4.	1. I consider OEGs in the classroom to be a waste of time.	−2.2	1.30
5.–6.	9. The design of the OEG was not age-appropriate for the pupils.	−2.0	1.87
	2. OEGs are uninteresting for pupils.	−2.0	1.58
7.–8.	12. The OEG used inappropriate terminology and terms inappropriate to the age and knowledge of the pupils.	−1.8	1.30
	11. The tasks were challenging. They did not match the knowledge and abilities of the pupils in the given grades.	−1.8	0.84
9.–10.	41. The OEG is difficult to operate and navigate.	−1.6	3.21
	34. Trying out the prepared OEG encouraged me to try to create my own online escape game.	−1.6	1.82

OEG = abbreviation for online escape game.

The following sections present the results from the PQMethod programme and ChatGPT analyses.

7.1. Analysis of Results by PQMethod

A correlation matrix was created using the PQMethod programme (see

Table 3. Correlation matrix created by PQMethod.

SORTS	1	2	3	4	5
1	100	63	22	47	44
2	63	100	26	37	52
3	22	26	100	26	33
4	47	37	26	100	48
5	44	52	33	48	100

).

After performing PCA, the programme grouped respondents into two factors based on the calculated factor loadings. The values are shown in

Table 4. Factor loadings according to PQMethod.

QSORT	Factor 1	Factor 2
1	0.7899X	−0.3288
2	0.7943X	−0.2579
3	0.5044	0.8231X
4	0.7173X	−0.0088
5	0.7785X	0.0716

, with X indicating the defining sort.

PQMethod then generated a list of significant statements for Factor 1 and for Factor 2 (see below).

7.2. Analysis of Results by ChatGPT

ChatGPT generated a correlation matrix (see

Table 5. Correlation matrix generated by ChatGPT.

SORTS	1	2	3	4	5
1	1.000	0.631	0.222	0.466	0.438
2	0.631	1.000	0.256	0.369	0.517
3	0.222	0.256	1.000	0.261	0.330
4	0.466	0.369	0.261	1.000	0.483
5	0.438	0.517	0.330	0.483	1.000

) and scree plot (see Figure 7).

Next, ChatGPT calculated the KMO value (equal to 1), Bartlett’s Chi-Square (equal to 64.652437), degrees of freedom (equal to 10), and significance p-value (equal to 4.72 × 10⁻¹⁰). According to ChatGPT’s interpretation, the data are “very suitable for factor analysis” according to the KMO value. ChatGPT further interpreted the data as follows:

“High value (64.65) with low p-value (4.72 × 10⁻¹⁰) confirms that the correlation matrix is significantly different from the unit matrix, supporting the appropriateness of factor analysis. The low p-value confirms that the correlations between variables are not random.”

(Open AI, 2024)

ChatGPT performed a PCA, which it processed into

Table 6. Factor loadings of two factors derived from ChatGPT analysis.

QSORT	Factor 1	Factor 2
1	−0.487	−0.355
2	−0.490	−0.279
3	−0.311	0.889
4	−0.442	−0.010
5	−0.480	0.077

, showing the PCA factor loadings for the two factors. ChatGPT interpreted Table 6 as follows:

“PC1 has a negative loading for all teachers, meaning that this component represents a common pattern among all teachers’ ratings. High negative loadings indicate that teachers share common characteristics or are rated similarly in the context of this component. Given that PC1 explains 52.6% of the variability, we can say that this component dominates teachers’ ratings and represents the main factor influencing all of them. PC2 has positive loadings especially for item 3 (authors’ note—meaning respondent 3), while the other items have lower or negative loadings. This means that this teacher (item 3) is perceived differently than the other teachers in the second principal component. PC 2 explains 17.1% of the variability, suggesting that this component captures specific differences that are not as strong as the main component but are still significant.”

ChatGPT was also asked to interpret the factor analysis, which ChatGPT interpreted as follows:

“Factor 1: Teachers 1, 2, 4, and 5: These teachers have significant loadings on the first factor (Factor 1). That is, their ratings are influenced by similar characteristics or characteristics that are captured by this factor. Thus, Factor 1 likely represents some common aspect or pattern that is present in the ratings of these teachers. For example, it may be their overall teaching effectiveness, their approach to teaching, or some other common feature.

Factor 2: Teacher 3: This teacher has a significant loading on the second factor (Factor 2), indicating that their ratings are influenced by different characteristics than the other teachers. Thus, Factor 2 captures specific characteristics that distinguish this teacher from others. This may be his or her specific teaching style, interaction with pupils, or some other unique aspect of his or her teaching.”

(Open AI, 2024)

8. Discussion

The discussion focuses on the comparative analysis of data processed by the PQMethod programme and ChatGPT, the interpretation of Factor 1 and Factor 2, and the study’s limitations, strengths, and future directions.

8.1. Comparison of PQMethod and ChatGPT Data Processing

Comparing the results obtained by the PQMethod programme and AI, the same conclusions were reached, albeit via different paths. Both approaches created an identical correlation matrix. Both performed PCA. Although the numerical values differed (see Table 4 and Table 6), the results were consistent: respondents were sorted into two factors, with Factor 1 saturated by four respondents and Factor 2 by one respondent. Different numerical values may stem from several factors, including variations in factor loading methodologies, limitations of the ChatGPT model in performing Varimax rotation, or other variables requiring further investigation in this field.

ChatGPT has been used in previous research for factor analysis, for example, in 2024 (Dobečková et al., 2024), where its use for factor analysis proved inappropriate due to its inability to perform Varimax rotation.

When ranking the statements in each factor according to their Q-score values (Q-SVs), the first ten statements with the highest level of agreement and the last ten statements with the lowest level of agreement are always the same using both the PQMethod programme and ChatGPT, except that for some, the order is reversed (see

Table 7. Comparison of results from ChatGPT and PQMethod analyses for Q-score values (Q-SVs) and rankings of the 10 statements displaying the highest agreement among respondents within Factor 1.

Statement (Number and Statement)	Q-SV		Ranking
	ChatGPT	PQMethod	ChatGPT	PQMethod
45. The terminology in the OEG is consistent with the terminology used in the subject of chemistry in primary school.	2.25	4	1	1
38. OEGs increase pupils’ interest in repetition.	2	3	2	3
31. I find repetition through OEGs an interesting way of repeating a thematic unit.	2	4	3	2
43. OEGs encourage active learning in pupils.	1.75	2	4	10
46. The language used in the OEG is age-appropriate for the pupils.	1.75	3	5	5
51. Online escape games encourage active learning methods and the use of diverse teaching strategies.	1.75	2	6	6
50. OEGs are a way to increase pupils’ motivation.	1.75	2	7	8
48. Pupils repeated their knowledge from the thematic unit in a fun way through OEGs.	1.75	3	8	4
35. OEGs are time-consuming to prepare.	1.5	2	9	7
28 The repetition through OEGs was interesting for the pupils.	1.5	2	10	9

The Q-score values represent the level of agreement, with increasingly positive values indicating stronger agreement. The ranking scale spans from 1 to 51, where rank 1 denotes the strongest level of agreement, and agreement decreases as the rank number increases. OEG stands for online escape game.

,

Table 8. Comparison of results from ChatGPT and PQMethod analyses for Q-scores (Q-SVs) and rankings of the 10 statements displaying the highest disagreement among respondents within Factor 1.

Statement (Number and Statement)	Q-SV		Ranking
	ChatGPT	PQMethod	ChatGPT	PQMethod
12. The OEG used inappropriate terminology and terms inappropriate to the age and knowledge of the pupils.	−1.5	−2	42	42
11. The tasks were challenging. They did not match the knowledge and abilities of the pupils in the given grades.	−1.75	−2	43	44
4. OEGs are not an appropriate means of reviewing pupils’ knowledge.	−1.75	−2	44	43
13. The assignment for each task was inadequate.	−2.0	−2	45	45
1. I consider OEGs in the classroom to be a waste of time.	−2.25	−2	46	46
9. The design of the OEG was not age-appropriate for the pupils.	−2.5	−3	47	49
2. OEGs are uninteresting for pupils.	−2.5	−3	48	47
10. The lessons have been processed in a cluttered manner.	−2.75	−4	49	50
3. OEGs do not belong in the primary school curriculum.	−2.75	−3	50	48
41. The OEG is difficult to operate and navigate.	−3	−4	51	51

The Q-score values represent the level of disagreement, with more negative values indicating stronger disagreement. The ranking scale ranges from 1 to 51, where a higher rank (e.g., 51) signifies the highest level of disagreement. OEG stands for online escape game.

,

Table 9. Comparison of results from ChatGPT and PQMethod analyses for Q-scores (Q-SVs) and rankings of the 10 statements displaying the highest agreement among respondents within Factor 2.

Statement (Number and Statement)	Q-SV		Ranking
	ChatGPT	PQMethod	ChatGPT	PQMethod
37. OEGs can be a tool to make the teaching of chemistry more attractive.	4	4	1	1
41. The OEG is difficult to operate and navigate.	4	4	2	2
33. If I had more games prepared in this way, I would definitely use them in my teaching.	3	3	3	3
44. OEGs are aligned with the content performance standard.	3	3	4	4
48. Pupils repeated their knowledge from the thematic unit in a fun way through OEGs.	3	3	5	5
29. Teachers do not have the time to create such interactive materials on top of their duties.	2	2	6	6
30. The use of the teaching material in the lessons met my expectations.	2	2	7	7
45. The terminology in the OEG is consistent with the terminology used in the subject of chemistry at primary school.	2	2	8	8
50. OEGs are a way to increase pupils’ motivation.	2	2	9	9
51. Online escape games encourage active learning methods and the use of diverse teaching strategies.	2	2	10	10

The Q-score values represent the level of agreement, with increasingly positive values indicating stronger agreement. The ranking scale spans from 1 to 51, where rank 1 denotes the strongest level of agreement, and agreement decreases as the rank number increases. OEG stands for online escape game.

and

Table 10. Comparison of results from ChatGPT and PQMethod analyses for Q-scores (Q-SVs) and rankings of the 10 statements displaying the highest disagreement among respondents within Factor 2.

Statement (Number and Statement)	Q-SV		Ranking
	ChatGPT	PQMethod	ChatGPT	PQMethod
1. I consider OEGs in the classroom to be a waste of time.	−2	−2	42	42
10. The lessons have been processed in a cluttered manner.	−2	−2	43	46
11. The tasks were challenging. They did not match the knowledge and abilities of the pupils in the given grades.	−2	−2	44	45
17. I use interactive aids often in my classes.	−2	−2	45	43
40. Teachers do not have sufficient technical equipment in schools to use OEGs in the classroom.	−2	−2	46	44
12. The OEG used inappropriate terminology and terms inappropriate to the age and knowledge of the pupils.	−3	−3	47	47
14. OEGs as a means of repetition and reinforcement of knowledge are not suitable for all pupils.	−3	−3	48	48
34. Trying out the prepared OEG encouraged me to try to create my own online escape game.	−3	−3	49	49
13. The assignment for each task was inadequate.	−4	−4	50	50
16. I consider the use of interactive applications in the classroom to be important.	−4	−4	51	51

The Q-score values represent the level of disagreement, with more negative values indicating stronger disagreement. The ranking scale ranges from 1 to 51, where a higher rank (e.g., 51) signifies the highest level of disagreement. OEG stands for online escape game.

).

ChatGPT proved to be a suitable choice for PCA as it reached the same conclusions as the PQMethod programme. Further analysis of ChatGPT’s use for PCA is necessary, as the data revealed slight differences that require investigation. This investigation is crucial for assessing the reliability of using artificial intelligence in processing Q methodology data in the form of PCA. The advantage of using ChatGPT for data processing is that it allows researchers to explain the different steps to them. Thus, it not only offers raw data but also provides a helping hand throughout the entire process of data processing and interpretation. Due to the current lack of research, it is essential to verify AI-generated results using established tools, such as the PQMethod programme.

8.2. Data Interpretation for Factor 1

The data suggest that teachers saturating Factor 1 perceive online escape games positively, which aligns with the findings of a study conducted by Bray and Dieckmann (2024).

These findings also align with constructivist approaches to education and modern theories of motivation and learning. Online escape games promote experiential learning, which focuses on connecting concrete experiences, observations, and reflections to form abstract concepts and generalisations, and testing the implications of concepts in new situations (Kolb, 1984; Ryder & Downs, 2022). In an educational context, such activities naturally lead to a deeper understanding of the subject matter and foster both independent and collaborative problem-solving.

Respondents perceived that online escape games increase pupils’ interest in repetition, considering this method of repetition an interesting way to revisit the thematic unit for both them and the pupils, as well as a fun form of repetition. Furthermore, they agreed with the statements that online escape games encourage active learning methods and the use of diverse learning strategies and help increase pupils’ motivation.

From a motivational perspective, Self-Determination Theory (SDT) can also explain teachers’ positive attitudes. SDT proposes specific factors that affect people’s ability to develop and sustain motivation for particular behaviours, thereby decreasing their dependence on external support (Alberts et al., 2024). Central to SDT are the basic psychological needs for autonomy, competence, and relatedness, which, when satisfied, lead to optimal human functioning (Howard et al., 2024). We can thus infer that if pupils perceive the task as meaningful and manageable, their intrinsic motivation and willingness to learn increase. When motivation becomes internalised, actions can become self-determined, at which point interventions based on extrinsic forms of motivation may no longer be needed (Alberts et al., 2024).

The respondents saw a disadvantage in the time needed to prepare an online escape game.

The respondents attributed the highest level of agreement to the statement that the terminology in online escape games is consistent with the terminology used in the chemistry subject in primary schools, and they felt that the language used in online escape games is age-appropriate for the pupils.

Respondents saturating Factor 1 expressed the highest level of disagreement with the statement that online escape games are challenging to navigate and control. Online escape games simultaneously fulfil the conditions for experiencing “flow”. Researchers initially defined flow as the holistic sensation that people feel when they act with total involvement (deMatos et al., 2021). According to research in the educational environment, there is a relation between flow and pupils’ interest and understanding during lectures (Culbertson et al., 2015). Flow mediates the constant interaction between the individual and the virtual environment across various contexts, such as learning and gaming, to provide experience enrichment and diverse outcomes (e.g., emotions and satisfaction) (deMatos et al., 2021).

They disagreed that online escape games did not belong in the primary school curriculum, that the material was not explicit, was uninteresting to pupils, was not age-appropriate for pupils, or that the assignment was inadequate. They also disagreed that online escape games were not an appropriate means of reviewing pupils’ knowledge, that online escape games were challenging, were not aligned with pupils’ knowledge and ability in the year groups, that appropriate terminology was not used, or that concepts were not suitable for pupils’ age and knowledge.

This is consistent with expressing the greatest degree of agreement with the statements above, which refines and confirms the data obtained.

Finally, it is crucial to mention that the format of online escape games also reflects the needs of the current generation of pupils, who have grown up in a technology-rich environment. These pupils are often referred to as “digital natives” (Prensky, 2001). However, more recent research indicates that this designation is misleading and unsubstantiated—while young people commonly use digital technologies, this does not automatically imply their effective utilisation for educational purposes (Kirschner & De Bruyckere, 2017). Educational tools, such as online escape games, must therefore be designed with an emphasis on pedagogical objectives and appropriateness, rather than solely on a format attractive to a “technologically raised” generation.

8.3. Data Interpretation for Factor 2

Factor 2 was defined by a single respondent who strongly agreed that online escape games can make chemistry education more engaging. However, this respondent also highlighted challenges related to usability and navigation within the game environment. Despite these concerns, the respondent acknowledged that if more ready-made tools of this nature were available, the respondent would use them in teaching. The respondent emphasised the alignment of these games with curricular requirements and specialised terminology, as well as their contribution to the revision and consolidation of pupils’ knowledge. In the respondent’s view, online escape games promote active forms of learning and diverse learning strategies, contributing to increased motivation. Conversely, the respondent considered creating respondent’s own online escape games to be overly time-consuming and did not perceive a personal need to master the development of digital games.

This perspective aligns with the Innovation Resistance Model, which suggests that individuals, despite potential benefits, may hesitate to adopt innovations due to perceived complexity (Laukkanen et al., 2007), conflict with prior beliefs, or because the innovation threatens to create changes in their well-established routines (Ram & Sheth, 1989). In this instance, the respondent does not rule out the potential of escape games, but their creation represents a barrier that the respondent does not perceive as surmountable.

The Technology Acceptance Model (TAM) provides further explanation for this stance, indicating that the acceptance of a technological tool is shaped not only by its perceived usefulness but also by its ease of use (Davis, 1993). Although the respondent evaluates the games as beneficial and meaningful, their technical complexity and the respondent’s lower digital self-efficacy may diminish the respondent’s willingness to experiment further with them.

The respondent disagreed with the statement that teachers do not have sufficient technical equipment in schools to use online learning games in the classroom. The respondent considers online escape games in the classroom a waste of time, but often uses interactive tools in the classroom. The respondent also rejects the notion that the curriculum lacks clarity or that the tasks are too challenging and misaligned with the knowledge and abilities of pupils in the given grades.

The view of the teacher saturating Factor 2 can be considered sceptical about online escape games in terms of the importance of using interactive applications in teaching, the willingness to try creating respondents’ online learning games, or regarding the simplicity of online escape games and orientation in them.

Scepticism regarding the significance of interactive applications in teaching can also be reflected through the prism of teachers’ professional identity. One aspect of the relationship between teachers’ identity and curriculum practice is that when programmes and curricula change, teachers lose a sense of themselves. School change, then, results in new stories to live by. Thus, teacher resistance against school change may also reflect an effort to maintain a “story to live by” (Beijaard et al., 2004).

Unlike the respondents in Factor 1, who perceive escape games as an attractive and enjoyable form of learning, the respondent in Factor 2, while accepting the educational potential of online escape games, does not approach them with enthusiasm and creative initiative. This suggests a pragmatic or cautious approach to using educational technologies.

However, as previously mentioned, Factor 2 is saturated by only one respondent. Consequently, this single-case factor represents a unique perspective that, while enriching our understanding of the diversity of opinions, precludes concluding broader group patterns.

8.4. Limitations

A key limitation of this study is the number of respondents. Given this limited sample size, this study is considered a pilot study. Furthermore, only one respondent saturates Factor 2, which indicates that the factor may reflect an individual perspective rather than a broadly shared experience. This singularity introduces a limitation to the study, as it may influence the interpretation of the findings and restrict the generalizability of the results.

Artificial intelligence tools are continuously evolving. At the time of this research, the chosen AI tool (ChatGPT 4-o) was unable to perform Varimax rotation, a standard rotation method commonly used in evaluating Q methodology results. This specific limitation of the selected AI tool must also be taken into account when interpreting the conclusions of the present study. It also underscores the importance of understanding the technical boundaries of AI tools and highlights the need for the continued development of AI models to better accommodate advanced statistical procedures.

Moreover, the observed differences in PCA results between ChatGPT and PQMethod warrant further exploration, and future research should investigate the underlying causes of these discrepancies. Clarifying these differences is essential to assess the reliability and replicability of AI-assisted data analysis. These differences may stem from variations in underlying computational method data handling or the absence of customizable rotation settings in the AI environment.

Despite the aforementioned limitations, this study offers important conclusions and insights, particularly regarding the potential utility of artificial intelligence tools in qualitative research. Nonetheless, a critical awareness of AI current capabilities and limitations remains essential to ensure responsible and accurate use.

8.5. Strengths and Future Directions

This study highlights the potential of artificial intelligence tools, specifically ChatGPT-4o, for statistical evaluation, factor analysis, and data interpretation in qualitative research. Given the continuous advancements in AI tools, it is anticipated that they will evolve to perform functions currently unavailable, such as Varimax rotation in ChatGPT. Following the integration of this capability, it would be advantageous to replicate this study and compare the outcomes.

This pilot study revealed that respondents perceived online escape games positively, primarily due to their attractiveness, ability to promote active learning, and capacity to consolidate and extend pupils’ knowledge. Future research on online escape games would benefit from expanding the respondent sample, not only within the Slovak Republic but also in other countries, to facilitate a comparative analysis of the findings. Furthermore, subsequent studies could build upon this research by comparing the results and approaches of other artificial intelligence models, such as Google’s Gemini, in addition to ChatGPT.

9. Conclusions

This pilot study explored primary school teachers’ perceptions of online escape games as a pedagogical tool, revealing nuanced insights into their potential and challenges. Given the pilot nature of this study and its limited sample size (n = 5 respondents), the findings reflect specific perceptions within this group and are not intended for broad generalisation. The findings indicate that online escape games are largely perceived positively, particularly for their attractiveness, ability to promote active learning, and capacity to consolidate and extend pupils’ knowledge. The distinct viewpoints identified through Q methodology highlighted two main perspectives: Factor 1 respondents (n = 4) emphasised the engaging, age-appropriate, and well-structured nature of online escape games, underscoring their significant potential to enhance pupils’ interest in learning. Conversely, the respondent representing Factor 2 (n = 1) acknowledged certain technical and time constraints yet still recognised the motivational potential and innovative character of these games.

The methodological comparison between the PQMethod programme and ChatGPT for data analysis, while exhibiting some numerical differences, consistently yielded the same two-factor saturation. This suggests ChatGPT’s promising potential as a future tool for the statistical processing of qualitative research data, particularly in principal component analysis.

In conclusion, online escape games represent a valuable and innovative addition to educational practice, fostering active learning and skill development. The insights from this pilot study underscore the potential of online escape games to transform traditional learning environments into dynamic, interactive spaces. These findings provide a foundational understanding of teachers’ perceptions within this specific context, emphasising the need for further research with larger and more diverse samples to explore generalizability. Future research, as detailed in Section 8.5, should build upon these findings by addressing the identified limitations (Section 8.4), exploring broader contexts, and further refining the integration of these games into curricula to maximise their pedagogical impact.