The necessity of online disclosure in higher education institutions has been investigated considering different theories. The public accountability perspective suggests that managers are more likely to engage in online disclosure because they have a greater commitment to public accountability and transparency (Coy et al., 2011). Second, according to legitimacy theory-based predictions (Suchman, 1995), increased online disclosure can be a strategic way for HEIs to demonstrate congruence with the goals and norms of the larger society (Chen and Roberts, 2010), with positive effects on institutional reputation, image, and public goodwill. In the Spanish context, the approval of the law on transparency, access to public information, and good governance in Spain has mandated certain reporting requirements for all public universities. Kansal et al. (2018) noted that compliance with minimum reporting requirements is a prerequisite for legitimacy when disclosures are mandatory. Thirdly, according to the stakeholder theory, all stakeholders, both internal and external, have the right to access information about an organisation’s activities and results (Deegan and Samkin, 2009). Specifically, universities must meet the public’s demand for increased oversight and accountability. Providing online information enables stakeholders to satisfy their information requirements and be aware of the process of value creation. This can facilitate support and consent from various stakeholder groups, including students, parents, funding bodies, research councils, the government, employers, and employees (Manes et al., 2018). Therefore, a greater degree of transparency will strengthen the legitimacy of universities in the eyes of key stakeholders, including students, parents, academic and administrative staff, ranking organisations, web agencies, and the general public (Ceulemans et al., 2015). Fourth, the resource dependence theory predicts that increased voluntary disclosure (Vidovich and Currie, 2011) can assist HEIs in gaining access to vital resources, such as donations and funds, and reduce political costs through improved institutional image and reputation (Chen and Roberts, 2010).

Alternatively, with the consolidation of the Internet and the sustained increase in its penetration rates, websites were established as channels that provided citizens with access to information, technological applications, and a variety of resources (Luna, 2017). The websites can be evaluated using a variety of models. Fath-Allah et al. (2017) evaluate 25 contemporary models proposed in specialised literature and identify four stages of digital maturity. The phases are as follows: (a) presence, where government agencies merely present information on the Internet; (b) interaction, where citizens can interact with the public administration; (c) transaction, where citizens can complete transactions and exchanges through the portals; and (d) integration, where different government agencies share information, including social networking applications and e-participation mechanisms. Following this approach, this paper uses a modified version of the global online transparency index developed by Saraite et al. (2018), which consists of four dimensions: E-information, E-Services, E-Participation, and Navigability, Design, and Accessibility (see Table I), to evaluate the online transparency of Spanish universities.

A few prior empirical studies on web disclosure have focused on HEIs (e.g., Crawford, 2012; Díaz et al., 2015; Kirchner et al., 2016; Saraite et al., 2018; Ramírez and Tejada, 2019). Among the existing studies, there are those that examine stakeholders’ perceptions of the importance of web transparency (Evgenievich et al., 2015; Ramírez et al., 2016; Flórez et al., 2017); those that examine the actual web transparency carried out by the university, albeit focusing on one specific dimension such as the information contained on the web page (Crawford, 2012; Flórez et al., 2014; Brusca et al., 2019; Son-Turan and Lambrechts, 2020), E-services (Kim et al., 2014) or E-participation (Díaz et al., 2015; Li and Zhao, 2020) and those that take a more synthetized approach to various dimensions (Pinto et al., 2009; Gallego et al., 2011; Manes et al., 2018; Ramírez and Tejada, 2019). Similarly, a number of studies have examined accessibility and usability issues on university websites (Buenadicha et al., 2001; Hilera et al., 2013; Casasola et al., 2017; İşeri et al., 2017; Acosta-Vargas et al., 2018).

Therefore, there is a need for research that examines the information provided by universities on their websites and the factors that explain this transparency practice.

This study concentrates on some factors that have been considered in previous research in the context of the corporate and public sectors, as well as some factors that are unique to the higher education sector (Saraite et al., 2018; Larrán et al., 2019; Segura-Mariño et al., 2020b; Garde et al., 2021). Seven variables have been chosen in particular: private funding, internationality, gender, academic performance, social media presence, university age, and university size.

Our sample is based on the total population of Spanish universities as of December 31, 2019 (50 public universities and 26 private universities).

We collected two primary categories of secondary data: university characteristics such as size, age, funding, internationality, presence on social media, leverage, and the Chancellor’s gender, and the global online transparency index (GOTI). In 2019, these data were collected from each university’s website.

This paper’s research is structured in two phases. First, a content analysis of all 76 Spanish universities’ websites was performed between September and December of 2019. The purpose of content analysis is to analyse collected data in a systematic, objective, and reliable manner (Guthrie, 2014; Krippendorff, 1980; Guthrie et al., 2004). According to Krippendorff (1980, p. 21), content analysis is “a research technique for drawing replicable and valid conclusions from data based on their context”. Moreover, according to McMillan (2000), the benefits of content analysis can also be applied to web content analysis. In addition, the content analysis typically yields a disclosure index, which is a numerical indicator that reflects the quantity of information disclosed in order to demonstrate the level of disclosure on the analysed communication channel (Gallego et al., 2011; Lacy et al., 2015). This study employs a modified version of the global online transparency index (GOTI) developed by Saraite et al. (2018) to quantify and evaluate the level of online transparency in Spanish universities. This index consists of 76 items divided into four dimensions: E-Information (it refers to the inclusion of information on the Internet about the main aspects of the university), E-Services (online services that optimise the process of learning and administrative services), E-Participation (online participatory democratic process in order to obtain the involvement of their heterogeneous stakeholders), and Navigability, Design, and Accessibility (characteristics of the webpage that facilitate the involvement of their diverse stakeholders). The E-Information dimension is comprised of seven subcategories: university overview, university organization and governance, financial items, general services: input (of students and resources), teaching services: process, teaching services: output, research services, and community services.

After delineating the information to be included in the disclosure index, a comprehensive analysis of the content of Spanish university websites was conducted. Specifically, the accumulation of data was conducted manually by two researchers who analyzed the websites of the 76 universities in Spain. The authors collected the data between September 2019 and December 2019. These researchers conducted the content analysis, which, according to previous scholars (Guthrie et al., 2004; Striukova et al., 2008), contributes to the precision and dependability of the coding procedure. Initially, a meeting was held to establish the strategy for each indicator. In particular, the researchers coded a sample of websites from ten universities. After discussing the differences and determining the final set of coding principles, the two researchers conducted the analysis independently. In conclusion, the results were reviewed to resolve any discrepancies and eliminate any potential bias. No significant discrepancies were reported, indicating that the coding was reliable. In instances where there was disagreement regarding the score ascribed to a particular item, a third researcher was consulted to determine the appropriate value. Krippendorf’s Alpha was calculated for the purpose of calculating the intercoder reliability test. The average value should be αk ≥ 0.70 (Krippendorff, 2011). The αk = 0.9 was obtained. This is likely due to the fact that there was no significant confusion during the contributor’s training or during the final data collection.

As GOTI is a numerical indicator with the purpose of quantifying the amount of information that is disclosed (Gallego et al., 2011), a dichotomous scoring method was selected in accordance with previous research that suggests an item’s presence is not correlated with its importance (Wood, 2000) and those with comparable research objectives (Buenadicha et al., 2001). Thus, a specific item (di) was assigned a “1” if the information was available on the webpage and a “0” otherwise. The quotient between the aggregates of all item scores (di) and the total number of items observed (76 items) determines the GOTI. In order to convey this as a percentage, this was multiplied by 100.

Taking these factors into account, the GOTI is calculated as follows:

where ∑ i = 1 76 d i is the score obtained in the sub-group of 76 items. The maximum value of the GOTI is 1 and the minimum is 0.

Using Cronbach’s alpha as recommended by Feldt (1969), the internal consistency of the GOTI index was calculated to be αc = 0.68, which is an acceptable value for exploratory research (Hair et al., 2014).

In the second phase, we employed the ordinary least squares (OLS) regression technique to determine the extent to which the selected explanatory variables could affect the level of online transparency.

To analyse the determinants of online transparency, the following model is proposed for both of them:

where GOTI is the global online transparency index of university; β0 is the constant; β1-β7 is the coefficient of the explanatory variables (size, age, funding, internationality, presence in social media, leverage, and gender of Chancellor); εi is the error or disturbance terms of university.

Table II summarizes the hypothesis, definitions of variables, proxies and expected signs.

Globally, the analysis of university websites revealed an online transparency level of 53.64 percent (see Table I). This result is consistent with the assertion that, despite the fact that the web can be an effective and useful tool for disclosing information to the broad range of stakeholders involved in university activities, its use by Spanish universities is below the desired level (Brusca et al., 2019; Ramírez and Tejada, 2019).

Regarding subindexes, this study demonstrates that universities do not place equal emphasis on all aspects. In particular, it was observed that “Navigability, Design, and Accessibility” and “E-participation” had higher values than “E-Information” and “E-Services”. This result is consistent with the findings of Saraite et al. (2018), who found that the top 100 universities in the world are aware of the importance of having a website with adequate “Navigability, Design, and Accessibility”, while “E-Information” was the least valued dimension.

Focusing on categories, the mean value of “E-Information” is 51.39 percent, indicating that, on average, Spanish universities do not disclose many of the items that should be disclosed in accordance with stakeholder theory. Specifically, Spanish universities are more likely to disclose items related to “General services: Input of students and resources”: (73.50%) and “Teaching services: Process” (64.74%), which is not surprising given that teaching appears to be the primary mission of most universities. In addition, the subtopic University organisation and governance (73.19%) are extensively accessible online. The lowest scores pertain to financial items (31.39%), highlighting the dearth of transparency regarding the financial performance statement and financial ratios. Additionally, the subsection “Community services” registered a low result, achieving an average value of 34.21%, indicating that despite the emerging third mission, characterised by synergy with the external community of industries and research centres, Spanish universities disclose a low quality and quantity of this information on their websites. Previous studies (Gallego et al., 2011; Católico, 2012; Flórez et al., 2017; Saraite et al., 2018; Ramírez and Tejada, 2019) have highlighted the need for higher education institutions to exert more effort on E-information.

The results, however, indicate that there is a high level of interest in “Navigability, Design, and Accessibility” (the mean value is 60.53%). Specifically, the obtained values indicate that the websites of Spanish universities have a high level of navigability and accessibility but not interactivity. The item with the highest usage rate was the presence of a section used specifically to access each category of information (92.11%). Similarly, the majority of cases included a search function on their pages, and the internal links of the university were distinguished from external links in 85.52 percent of instances. However, only a few cases (10.53%) included hyperlinks as a tool alongside the supplied information. These findings are consistent with those of other studies (Hilera et al., 2013; Casasola et al., 2017; İşeri et al., 2017; Acosta-Vargas et al., 2018), indicating that there is still work to be done before the entire university community is fully accesible.

The “E-Participation” subindex represented 55.26 percent of the total items. It was striking to discover that only 10.53 percent of universities offered online opinion polls when analysing this category. Also receiving a low score was the student complaints and suggestions box (23.68%). In contrast, it was observed that contact information for those in command of the services was abundant, with a value of 81.58%. These results indicate that the mechanisms implemented by Spanish universities to improve e-participation are still in their infancy, which is consistent with the findings of Díaz et al. (2015) and Li and Zhao (2020), who assert that higher education institutions need to pay more attention to facilitating the e-participation process and providing the necessary channels to receive feedback from the public.

Finally, “E-Services” dimension received the lowest score (47.37%). In terms of this category of information, it is important to note that 65.79 percent of the studied universities offered downloadable forms and applications for managing official study procedures. In only 26.32 percent of all cases, it was possible to complete the entire process online, including payment, for unofficial studies.

Therefore, despite the fact that the Internet can be an effective and useful tool for disseminating information to the vast array of stakeholders engaged in university activities, its use by Spanish universities falls short of expectations. In this regard, these results demonstrate the need for universities in Spain to make greater efforts to disseminate pertinent information and establish interaction mechanisms that facilitate effective communication. Moreover, our findings are consistent with those of Ramírez and Tejada (2019), Brusca et al. (2019), and Andrades et al. (2021), who discovered that Spanish universities are attempting to create websites with the properties recommended by the literature. Moreover, if we compare our results with those of Gallego et al. (2011), we can see that, despite the 2013 implementation of the law on transparency, access to public information, and good governance, the situation regarding the characteristics of the websites of Spanish universities has not changed after ten years.

Table III contains the principal descriptive statistics of the independent variables used in the regressions.

To test the seven hypotheses, an OLS regression model was performed with web transparency as the dependent variable (i.e., the global online transparency index, GOTI).

The results of the OLS regression model used to test the seven hypotheses are presented in Table IV. Multicollinearity (variance influence factor test) and heteroscedasticity were evaluated as hypotheses underlying the regression model (Hidalgo et al., 2011). To evaluate multicollinearity issues, the variance inflation factor (VIF) score was calculated for each independent variable. In this instance, the maximum VIF for variable leverage is 3.176, eliminating multicollinearity issues. In addition, White’s and Breusch-Pagan’s heteroscedasticity analyses were conducted with negative outcomes (Table IV). With an adjusted R-Squared of 0.45, the regression model is statistically significant (P-value 0.01).

Three out of seven variables are statistically significant in general: internationality, size, and leverage.

Internationality specifically has a positive coefficient and is statistically significant at the 1% level (p-value 0.01). This conclusion can be derived from the fact that a higher level of information disclosed by universities can attract more international students, for whom university websites are the primary source of information regarding the activities, services, and financial condition of the universities. Previous research on the level of online information disclosure in universities (Manes et al., 2018; Gallego et al., 2011) has found a positive relationship between internationality and online information disclosure. Some features, such as an English-language website and audio and video files, offer international students friendlier and more accessible disclosure.

Similarly, we discovered a correlation between size and voluntary disclosure. The coefficient for size is positive and statistically significant at the 5% level (p-value 0.05). This result confirms, in accordance with stakeholder theory and political costs theory, that large universities disclose more information to satisfy the information needs of a growing number of stakeholders (taxpayers, governments, students, research centres, and external funders) who demand high levels of transparency and to reduce external pressure and political costs (White et al., 2007). According to the legitimacy theory, larger universities are more visible and subject to public scrutiny, which may prompt them to disclose more information in order to enhance their legitimacy (Andrades et al., 2021). In addition, the findings are consistent with those of previous research on universities (Maingoth and Zeghal, 2008; Gallego et al., 2011; Saraite et al., 2018; Católico, 2012; Garde et al., 2013; Andrades et al., 2021).

Leverage is negative and statistically significant at the 10% confidence level (p-value 0.10). This demonstrates that leverage has a negative effect on the extent of web-based disclosure by universities as the quantity of information to disseminate decreases in proportion to leverage. This finding is consistent with Gallego et al. (2011), who noted that Spanish universities with greater leverage disclose less information on their websites than universities with minimal leverage. In contrast to the findings of other studies (Gordon et al., 2002; Ntim et al., 2017; Nicoló et al., 2020), we find that leverage has no significant effect on the extent of disclosure in the context of higher education.

On the other hand, there were no significant associations between web transparency and the variables “age, funding, social media presence, and gender of the Chancellor.” Specifically, age and funding have negative coefficients that are not statistically significant, whereas social media presence and the gender of the Chancellor have positive coefficients that are not statistically significant. These results are consistent with previous studies that found no relationship between public or private funding and online transparency (Gallego et al., 2011; Flórez et al., 2017; Brusca et al., 2019); studies that found no significant relationship between university’ age and online transparency (Gallego et al., 2011; Bisogno et al., 2014; Manes et al., 2018; Garde et al., 2021); and studies that found that gender of the Chancellor is not a factor affecting the online transparency in higher education institutions (Garde et al., 2020; Andrades et al., 2021).

In conclusion, internationality, size, and leverage appear to be three appropriate factors on which universities can act to develop strategies that positively influence the level of online information transparency. While it appears that the trend regarding the greater or lesser implementation of transparency best practices via the Internet is not influenced by the “age, funding, social media presence, and gender of the Chancellor,” these factors do appear to play a role.