CHARACTERIZATION OF TRAINING MODELS IN DATA LITERACY: PERSPECTIVES FOR TEACHER TRAINING

Objective: This article presents the results of a study on training practices and models for data literacy at the university level. Theoretical Framework: Data literacy is an emerging and novel construct that provides the training required in a digital and digitized society. Method: To achieve this purpose, the training proposals based on courses, offered in the period from 2018 to 2022, and available in Spanish, English, Portuguese, French and Italian, were previously identified. Then, different categories were established that allowed the grouping of the data obtained, which were recorded in a file created for this purpose. Results and Discussion: The quantitative analysis carried out makes it possible to determine which country offers the most training in data literacy; the most common type of practice, the modality under which it is usually offered and the methodology applied; its cost and the target audience. It also confirms the existence of links between the different variables examined that characterize the type of training offered. Research Implications: It confirms the urgent need for a transversal training model in data literacy that responds to its main paradigms and parameters. Originality/Value: This study contributes to the literature by characterizing, for the first time, the training practices in data literacy, which will make it possible to detect the shortcomings and areas for improvement that these training scenarios present.


INTRODUCTION
The constant production and consumption of data constitutes the characteristic feature of 21st century society.Through the continuous management and use of information technologies, millions of data are generated every day, of which a large amount is generated 3 and/or transferred by ourselves unconsciously, both in our professional and personal environments.The searches we carry out on the Internet, the messages we send through social networks, the interaction with the intelligent technologies that surround us and that we even have incorporated, constitute, among others, a constant source of data.And we live in a world where "information grows in a way that is as extraordinary as it is disproportionate and uncontrollable" (Carbonell-Alcocer & Gertrudix -Barrio, 2019, p. 2).This constant and massive production of data by all citizens, no longer limited to specific environments, has generalized data issues throughout society (Verdi, 2023).On the other hand, the pandemic caused by COVID-19 revealed the exponential increase in data, its diversity and its not always reliable nature, but also the benefits and opportunities that data brings with it in terms of progress in the knowledge, decision-making and formation of opinions based on truthful information.However, in this digital, technical and data-enabled environment , a large part of citizens is still not aware of the power that data has nor does it know how to interpret it for their own benefit, hence the need to train them in the proper use of it and design skills that form proactive ( Carmi et al., 2020) and empowered citizens and prevent, among others, the uncritical (Gálvez de la Cuesta et al., 2020), illegal and uncontrolled use of data, as well as the social exclusion of the individual.
Given this environment where data is an inexhaustible source of knowledge and opportunities (Villegas-Muro, 2023), Raffaghelli (2020) asks what type of educational interventions should be designed and promoted to address this complex and multi-layered problem that is datafication.of society, to which Pangrazio and Sefton -Green (2020) respond by suggesting that data literacy can be a useful response and that it should be considered by democratic societies as a strategic element to adapt to the digital world through greater investment and commitment, both in formal and informal education.

THEORETICAL FRAMEWORK
Data literacy provides the training required for the selection, analysis and management of data, tasks that face, in turn, a situation of explosion and interest that is accompanied by the promotion and recognition of digital competence.In this way, data literacy, closely related to information literacy, which is part of a broader set of competencies called multiple literacies (Marzal, 2020), provides the necessary training to interact in the contemporary data society, by constituting the starting point in the acquisition of knowledge and the development of skills that allow access, measurement, analysis, critical evaluation, presentation, management, ethical use 4 and preservation of data, in everyday and professional contexts.This literacy also consists of formulating and answering questions from data sets through a process of inquiry, considering the ethical use of the data (Marín et al., 2022).The objective to achieve is to be able to make informed decisions based on data, in any scenario, and in this context "it is important to recognize the omnipresence of data, its fluidity" (Milanés & Feria, 2023, p. 2).
In order to offer training programs in data literacy, it is necessary, on the one hand, to start from a holistic and inclusive approach (Atenas et al., 2023), and, on the other hand, to reach a common definition of data literacy and identify, in a clear and consensual manner, the knowledge and skills that comprise this metaconstruct (Beck & Nunnaley , 2021;Henderson & Corry , 2021).In short, data literacy must be provided with a more complete theorization and practical development ( Pangrazio & Sefton -Green, 2020), as well as the design of pedagogies that encourage the development of transversal skills that incorporate critical, ethical and political dimensions of data (Athens et al., 2023).
In this sense, the findings that we show in this article fully coincide with those obtained in the systematic reviews of the literature carried out by Ghodoosi et al. (2023), Pinto et al. (2023) and Cui et al. (2023), where the need to develop and provide greater training in data literacy in the university environment is highlighted.
The enormous amount of data that is produced permanently, in all areas and at all levels, makes the unavoidable task of training current and future generations of university students in the acquisition, development and implementation of skills a priority.and data literacy skills in order not only to promote good practices in working with data, but also to take advantage of that talent during and throughout the process of identification, location, collection, analysis, interpretation, use, dissemination and reuse of the data, so that all this results in a real benefit for society as a whole.
As can be seen in the study carried out, breaking the talent gap in data literacy inevitably means introducing changes or making adjustments to the programs of the subjects taught in university degrees.And, to do this, "it is necessary to incorporate data into management and teaching practice by designing teaching approaches with the use of data" (Martín, Gallego & Iglesias, 2023, p. 153).
Hence, it is imperative to design good training models in data literacy and adequate training practices, since the appropriate use of data and its application in various contexts and scenarios will promote the construction of knowledge that enables the anticipation and resolution of problems.Likewise, carrying out this process will allow, in turn, the identification, addressing and comparison of the need for political measures supported by evidence, as well as 5 the discernment of effective practices in a general or specific manner, benefiting social and educational systems.individuals in order to improve training.The relevance or significance of this approach lies in the availability of data for the common good, facilitating its use by researchers, government and private entities, emerging companies, as well as by citizens themselves (Martín-González and Iglesias-Rodríguez, 2021 ;Martín, Gallego & Iglesias, 2023).
In this regard, we agree with Ghodoosi et al. (2023, p. 113) when he says that "teaching in data literacy is complicated by three factors: the continuous change of technologies and the data they generate, the changing context of companies and the breadth of contexts in which data literacy is relevant"; Consequently, it is imminent that teachers and students are carriers of the competencies, capacities and prospective knowledge that are needed to innovate and prosper (European Commission, 2018).
In the field of teacher professional development, there has been an extensive body of research and theorizing that addresses various models and approaches aimed at improving pedagogical practice and educational effectiveness.Among the prominent authors in this field are Marcelo, Imbernón, Darling-Hammond or Fullan, among others, whose contributions have significantly influenced the understanding and conceptualization of teacher training models.
These authors have explored and contributed to the design of teaching professional development models, in some cases, based on reflective learning (Imbernón, 2007;2020), which emphasize critical reflection as a driver of continuous improvement and advocate for programs training programs that encourage self-assessment and self-reflection (Allen, 2008;Corrall , 2017).These types of programs especially promote professional autonomy and the ability to adapt to changing educational contexts.
In other cases, they have opted for comprehensive approaches to teacher professional development, as is the case of Marcelo, who places emphasis on the notion of "situated training" and the development of "disruptive experiences" (Vaillant and Marcelo, 2021;Marcelo and Marcelo, 2021).In this case, great importance is given to linking and connecting training with everyday practice through practical experiences and continuous reflection as a foundation for personal and professional growth ( Kitchen and Petrarca, 2016).The author argues that connecting theory and practice improves pedagogical skills; and, consequently, the exchange of effective and collaborative practices (Darling-Hammond , 2017), which will result in a significant improvement in professional development.
To all this, it could be added that, in the field of institutional training models, to achieve the proposal made by the authors cited above, effective leadership is required that supports 6 teacher professional development and, consequently, the successful implementation of innovative practices (Fullan, 2020).
As can be seen, these approaches not only focus on the individual growth of the teacher, but also on the creation of collaborative learning environments that take advantage of various training modalities to enrich the educational experience (Hargreaves and O'Connor, 2018;Bolick et al. ., 2019;Fullan, 2020;

METHODOLOGY
This study collects the results obtained from the analysis of the training practices identified in terms of data literacy, at a national and international level.To locate it, a query was made through the Google search engine, using the search equations "data literacy" AND "University", and "data training" AND "university", and their corresponding variants in English, Portuguese, Italian and French.Furthermore, the research was limited to the period between 2018 and 2022, and training models and practices based on courses (MOOC, webinar , blog, etc.) were prioritized.
The information corresponding to each of the 82 training practices that make up the final sample was compiled in a file with the following categories: In this way, we obtained information about the type of training practice developed to improve data literacy of different groups.To do this, we investigate the nature of the practice, that is, whether it is a Specialization Diploma, MOOC, Seminar or Webinar ; the modality in which Plan.Additionally, we expand this exploration by providing a brief description of the practice in the same questionnaire.Finally, regarding the context of the instruction, we examine the target audience to which it is directed and the country in which it is taught, leaving a space to indicate the university center from which the teaching is given.
In total, the data collection form prepared for this purpose is made up of nine questions: five of them with a single possible answer, two with multiple answers, two with a short answer and one open.Given the methodological characteristics of the items considered, all of them were analyzed with the SPSS version 28.1 program.

RESULTS
The results corresponding to the descriptive and inferential analysis of each of the closed questions included in the questionnaire are presented below.
In Table 1, you can see the value that is most repeated in each variable (mode).Broadly speaking, it is concluded that Spain would be the country in which the most training in data literacy is developed; The most widespread type of practice are seminars, workshops and courses, mainly in person.Regarding the cost, it is observed that the value that is most repeated is "Paid" and the methodology used would be, in essence, interactive.
Finally, universities and training centers advocate long-term instruction that makes it possible to delve deeper into the contents of the field of study.Consequently, much of the instruction lasts between 12 and 24 months and is aimed, for the most part, at people who already have a bachelor's degree, bachelor's degree or diploma.With respect to the differences between the type of training and the territory (figure 1), it is observed that there is a dependence between these variables (p<0.04),therefore, depending on the country, a different training offer can be enjoyed.The interactive methodology along with the expository methodology is imposed in the training plans of the Expert Degrees and in the Specialization Diplomas (37.1%; n=13); On the contrary, seminars, courses or workshops show their preference for the interactive method (38.6%; n=17) and the expository method (25%; n=11) (figure 6).

Figure 6
Relationship between the type of training practice and the methodology that follows.As seen in Figure 7, most of the Specialization Diplomas/Expert Titles integrate data science (91.4%; n=32), data management (85.7%;n=30) and big data (60%; n=21); In fourth position would be the content directed towards the search and recovery of data (37.1%;n=1) and, subsequently, that interested in the protection, security and ethics of data (31.4%;n=11).15 undergraduate students (11.4% and 18.2% respectively).On the contrary, the general public could enjoy any type of training practices ( webinar , MOOC, SPOC...), but with a very reduced teaching offer (figure 8).

Figure 8
Relationship between the target audience and the type of training practice Source: Own elaboration More than 20% of training practices are aimed at graduates, graduates and diploma holders, this means that this group has more options to choose between the face-to-face modality (40%; n=12), blended face-to-face (40%; n=4) and online (39.3%; n=11).
Undergraduate and postgraduate students find themselves in this same situation, although with a smaller training offer.As for teaching and research staff, other groups (such as PAS or business professionals) or the general public, would have options both online and in blended or exclusive face-to-face settings.This will depend on the country you are in.
In Figure 9 you can see how all groups have access to content on science and data management, in addition to big data.Specifically, graduates, diploma holders and graduates also have access to subjects related to data search and recovery (24.1%; n=7) and on data protection, security and ethics (20.7%; n =6).
Those who have already completed a postgraduate or master's program would have greater access to content related to data description (FAIR; 16%; n=4) and data preservation (36%; n=9) than the rest of the recipients .
Doctoral students can obtain broader information on content related to information and/or data literacy, on data management plans or data protection, security and ethics (42.9%; n=3, in all cases).
Teaching and research staff, other professionals and the general public have access to the main contents, as well as the rest of the studies.Finally, the general public, despite having a limited training offer, can delve into content related to data management and science (60%; n=4 and 60%; n=3, respectively), but not over big data.Therefore, it is the responsibility of the university to prepare students and graduates for professional and civic life, since the ability to work with data is a crucial skill in a wide range of disciplines and professional fields such as science, technology, engineering, social sciences, medicine, journalism and many others (Robertson & Tisdall , 2020, Usova & Laws , 2021).
This implies having the ability to design learning activities that involve the exploration and The analysis of data literacy training practices identified in the university context, at a national and international level, confirms the growing importance of this competence in higher education.
According to the results obtained, Spain consolidates itself as a leader in data training by registering a greater number of proposals (34.1%), followed by countries such as France (22%) and the United Kingdom (19.5%).
More than half of the training proposals identified (53.7%) take place through seminars, courses or workshops that are taught in person (36.59%), although they are closely followed by those provided to through virtual means (34.1%).Almost three quarters of the internships offered are paid (70.7%).
The learning content included in the instruction refers mainly to data science, data management and big data (25.1%,22.6% and 12.3% respectively), which are the areas relating to the data where there is a greater demand for professionals in the current labor market, both inside and outside Spain.
Regarding the existing relationships between the country where the training center is located and the rest of the variables analyzed, it is observed that the type of training varies depending on its geographical location, as well as the methodology applied.However, the in-person modality is the predominant one, as is payment for the instruction received.Likewise, there is a direct relationship between the duration of the training and the country where it is offered.
Likewise, a link is established between the type of training practice and the modality in which it is taught, the duration of the practice and the methodology applied.However, in this variable there is no direct relationship with the cost of instruction or with the learning contents, since, in all cases, those related to data science, data management and big data stand out.
The connection present between the people to whom the instruction is directed and the rest of the variables analyzed is also an aspect to take into account, especially in relation to the type of training offered; although it is not decisive in terms of the modality in which the training is given, nor the cost or the learning contents.
In short, the study carried out allows us to characterize the training offer provided in the university context at a national and international level, in terms of data literacy, and conclude that this varies considerably in terms of approach, methodology and scope.
However, there are recurring patterns that encompass the integration of data in the 19 curricula, both undergraduate and graduate, either through the inclusion of specific courses on data literacy or through the insertion of modules or related activities in courses.already existing in various disciplines, such as computer science, statistics, social sciences, health sciences and business, among other areas.On certain occasions, an interdisciplinary approach is adopted by recognizing that data literacy is a crucial competence in different fields of knowledge, thus allowing training practices in this area to be part of the programs of numerous disciplines and not are restricted to a particular sphere.
On the other hand, data literacy involves understanding the ethical and legal aspects linked to the use of data.Therefore, many data literacy programs emphasize the relevance of addressing ethical issues such as privacy, security, and fairness in data manipulation and analysis.
Even though data literacy is gaining importance in many universities, some limitations and challenges can be identified, as the scope of training practices on this topic can still vary between institutions and countries.Some universities have well-established programs, courses, seminars and/or workshops in data literacy, while others have a limited or even no focus.
Another limitation relates to access to appropriate resources and technologies to learn and practice data literacy.This is because, depending on the geographical location of the universities and the financial resources at their disposal, issues related to infrastructure, software and resources to support the teaching of data literacy.
Added to all this, it is worth noting that the field of data literacy is continually evolving due to rapid changes in the data landscape, which, in turn, can make it difficult to review and adapt training practices in this field.scope and, therefore, the possibility of providing relevant and updated education in this field.
In summary, taking into consideration the results achieved in this research, the main challenge that data literacy presents is to develop a transversal training model that can be adapted throughout higher education, at a national and international level, and that responds to its main paradigms.and parameters.Well , as Ghodoosi et al. (2023, p.123), "Data literacy is dynamic and, therefore, there is a likely need to create innovative strategies to better teach the emerging dimensions of data literacy." ___________________________________________________________________________ Rev. Gest.Soc.Ambient.| Miami | v.18.n.8 | p.1-22 | e08421 | 2024.

-
this information allowed us to delve deeper into the characteristics of the training practices offered in Brazil, Spain, the United States, Italy, Portugal, the United Kingdom and Switzerland in the field of data literacy during the period examined (2018-2022).
Characterization of Training Models in Data Literacy: Perspectives For Teacher Training ___________________________________________________________________________ Rev. Gest.Soc.Ambient.| Miami | v.18.n.8 | p.1-22 | e08421 | 2024.7 this instruction is carried out (in-person, online, blended), the cost or not of the training and its timing.Likewise, we were able to identify the proposed methodology, depending on whether it is expository or interactive, or a mixed composition; and what is the curricular content of each of the practices.Specifically, given the breadth of subjects that can be considered in data literacy, we limited the options to nine: Data management; Big data; Data science (analysis and/or visualization); Data Description (FAIR); Data search and recovery; Data protection, security and ethics; Data Preservation; Data Management and Information and/or Data Literacy . Soc. Ambient.| Miami | v.18.n.8 | p.1-22 | e08421 | 2024understanding of the data obtained, these were grouped based on the following relationships: firstly, the results of the associations between countries were analyzed according to the type of training practice, the cost it has, the group to which it is directs, the modality , the methodology and the duration of each practice; Secondly, the correspondences between the modality of the training practice, the type of training, the recipients, the duration and the cost of this training were examined.And, finally, the correlation between the public to which the training practice is directed, the content of the training and the cost was studied.4.1 RESULTS OF THE ANALYSIS OF THE RELATIONSHIPS BETWEEN COUNTRIES BASED ON THE VARIABLES STUDYIn this section, the relationships between the countries and the type of training practice they offer, the modality in which it is developed and the cost of each of them are presented.To do this, the chi-square statistic (p-value) has been considered as an indicator of independence or not between the variables studied, starting from the null hypothesis (H0: all variables are independent).

Figure 1
Figure 1Relationship between the type of training practice and the country

Figure 2
Figure 2Relationship between the modality of the training practice and the country

Figure 3
Figure 3Relationship between the cost of training and the country

Figure 4
Figure 4Relationship between the duration of the training practice depending on the country.

Figure 5
Figure 5Relationship between the type of training practice and the modality in which it is taught Source: Own elaboration Finally, 22.7% (n=10) of the seminars, courses or workshops are free and 52.3% (n=23) of the training practices carry a cost, as does obtaining the Specialization Diploma or Expert Title.On the other hand, the webinars are completely free and the MOOCs, SPOCs and NOOCs allow you to obtain a teaching certificate upon payment.Finally, the content of the different training practices on data literacy is multiple and diverse.While those that address a single topic are limited, it is common for the same training to bring together different contents on the field of study.
___________________________________________________________________________ Rev. Gest.Soc.Ambient.| Miami | v.18.n.8 | p.1-22 | e08421 | 2024.14).With a smaller and similar presence, there is teaching in information and/or data literacy and on data management plans (20%; n=7); On the contrary, only four of these training practices integrate data description (11.4%).For their part, seminars, courses or workshops maintain the trend of the previously mentioned practices by showing their preference for data science (61.4%; n=27) and data management (53.3%;n= 23).However, this type of training practice has, unlike the previous ones, a greater interest in information and/or data literacy (25%; n=11) and, to a lesser extent, in big data (20.5% ; n= 9) or data search and recovery (11.4%; n=5).As for the two webinars included in the study, both offer content in line with the trend of the rest of training practices.

Figure 7
Figure 7Relationship between the type of training practice and the topics studied .

Figure 9
Figure 9Relationship between the target audience and the contents addressed by each training practice

Figure 10
Figure 10Relationship between the target audience and the cost of the training practice ___________________________________________________________________________ Rev. Gest.Soc.Ambient.| Miami | v.18.n.8 | p.1-22 | e08421 | 2024.18 analysis of data, and the ability to provide feedback to the different profiles of people interested in receiving this training in a way that responds to their needs and abilities (European Union , 2022).
Thus, for example, in Spain and