The first studies on the perception of trust in higher education emerged in the USA due to the significant fluctuations in graduation rates in higher education since the middle of the 20th century and gained momentum in the next period (Schofer & Myer, 2005). Despite the historically critical importance of higher education institutions in providing information access to a wide segment of the public and achieving long-term goals, public trust in higher education institutions has been given little consideration outside of western countries (EUA, 2011; Johnstone & Marcucci, 2007). For this reason, research on individuals' trust in higher education institutions in non-western societies is extremely limited (Schofer & Myer, 2005). However, trust in higher education plays an important role in the functioning of higher education based on the interaction of policy-makers and other stakeholders (Leveille, 2006).

Higher education institutions are seen as an indispensable element for economic development, democracy, and conscious citizenship in many countries (CHE, 2019; OECD, 2017; Psacharopoulos, 2009). Higher education institutions, which are accepted as a “social contract”, provide the legitimacy they need with the legitimacy of the public, which is a normative control system, apart from the legislation and the resources of the states (Suchman, 1995). Considering the strength and significant impact of this relationship, higher education institutions have an important responsibility in establishing trust and confidence. In this context, there are important areas such as increasing student participation and success, establishing a strong financial balance, developing a smooth and sustainable social contract, and providing quality education under appropriate conditions, which should be taken into account by public authorities as well as higher education institutions (Boer et al., 2010; Eurydice, 2016). On the other hand, for higher education institutions to receive sufficient public support, their goals and objectives must be compatible with social values. Otherwise, unmet expectations can lead to decreased trust and increased questioning of the existence of the institution (Leveille, 2006). This may have a deeper impact, especially in collective and high uncertainty avoidance societies such as Turkey, which has a centralized and rigid bureaucratic educational structure. From this point of view, it is extremely important to create a perception of trust among higher education stakeholders and to monitor it through certain processes.

With the purpose of filling these gaps, the current research focused on determining the perception of trust in higher education institutions, which have an important function in Turkey. Earlier researchers suggest applying a combination of theory and empirical research to the design process when developing and validating a measurement instrument (Frenzel et al., 2016). The advantage of using both theoretical and empirical strategies is to ensure that the measurement tool covers content validity to cover relevant items and content areas (Burić et al., 2018). Accordingly, the research aims to develop and validate the “Trust in Higher Education Inventory” (THEI) in the context of Turkey by adopting a theoretical-empirical approach. In this context, the research questions are summarized as follows:

1. What are the views of higher education students about trust in higher education?
2. Is the THEI reliable?
3. Is the THEI essentially valid?
4. Is the THEI structurally valid?
5. Is the THEI valid externally?

This study was carried out by following the stages of the exploratory sequential design (Creswell, 2014). In the first stage of the research, a semi-structured interview form was developed by the researchers to reveal the main determinants of trust in higher education. In the semi-structured interview form, there are six main questions and four follow-up questions developed based on the relevant literature. The opinions of four experts and two students were taken to ensure items and content validity in the draft interview form. The pilot application of the developed interview form was carried out on two students. As a result of the pilot application, it was decided that all the questions in the interview form were understandable, and the interviews were conducted. Qualitative data were obtained from 20 students attending different faculties of a university in the Marmara region of Turkey and determined by the maximum sampling method. Interviews with the participants lasted an average of 40 minutes. The data were coded with an open and axial coding method. In the second stage of the research, the scale development stages were followed and a pool of 21 items was created based on the qualitative findings. Two new items were added to the item pool by making use of the relevant literature, and a total of 23 items were reached in the first stage. The material validity of the scale was examined by six experts (Krabbe, 2017). In line with expert opinions, four items were dropped, one item was added, and six items were changed due to unsatisfactory confidence ratings. The item pool created at the end of this stage consisted of 20 items. The Likert-type five-point rating scale form was first applied to 390 undergraduate students. Explanatory factor analysis (EFA) was applied to the collected data in SPSS 25 program. The following criteria were used in the EFA procedure: (1) cross-loads of all items fewer than 0.30; (2) all items with regression loadings greater than 0.50 on their intended conceptual factors; (3) three or more conceptual aligned items in one factor; (4) all items without negative error variance (Costello & Osborne, 2005). In this process, six items were removed from THEI. After the EFA procedures, the scale was reapplied to 366 undergraduate students to determine the external validity of THEI over the 14-item model, and confirmatory factor analysis (CFA) was performed using LISREL 8.80.

Results from the EFA procedure for THEI yielded an original one-factor solution accounting for approximately 46.68% of the variance with eigenvalues >1, characterized by an adequate sample size (Kaiser-Meyer-Olkin = 0.92; Bartlett's test of sphericity χ2 = 2437.441, df = 91, p < 0,001). In this process, the same 11 items were removed from the items in sample groups, and the THEI with a single factor was left with 14 items. The described explanatory model was evaluated in CFA, which allows better detection of items with weak loads and violating the structure. A multi-criteria approach was adopted for acceptable model fit (Schumacker & Lomax, 2010). As a result of the analysis, the DFA model gave a good model fit index with the 14-item THEI (χ2 = 153.68, df = 74, p < 0.001; GFI = 0.93, AGFI = 0.90, CFI = 0.97, NFI = 0.94, NNFI = 0.96, SRMR = 0.050 RMR = 0.05, RMSEA = 0.060, RFI = 0.93, IFI = 0.97). As a result of the internal consistency analysis of the model obtained, it was determined that the Cronbach alpha values had a range between 0.90 and 0.91 and the average value was 0.91. These results indicated that the developed scale was reliable enough and that the model could be used meaningfully in further analyzes.

Boer, H. de, Jongbloed, B., Enders, J., & File, J. (2010). Progress in higher education reform across Europe, Funding reform, Volume 1: Executive Summary and main report. Centre for Higher Education Policy Studies. Retrieved January 9, 2022, from http://doc.utwente.nl/88692/ Burić, I., Slišković, A., & Macuka, I. (2018). A mixed-method approach to the assessment of teachers’ emotions: Development and validation of the Teacher Emotion Questionnaire. Educational Psychology, 38(3), 325-349. CHE. (2019). 2019-2023 stratejik planı [2019-2023 strategic plan]. Retrieved December 28, 2021, from http://www.sp.gov.tr/upload/xSPStratejikPlan/files/U8mRM+YOK_2019-2023_Donemi_Stratejik_Plani.pdf Costello, A. B., & Osborne, J. W. (2005). Best practices in exploratory factor analysis: Four recommendations for getting the most from your analysis. Practical Assessment, Research & Evaluation, 10(7), 1–9. Creswell, J. W. (2014). A concise introduction to mixed methods research. Sage. Enders, J., de Boer, H., File, J., Jongbloed, B., & Westerheijden, D. (2011). Reform of higher education in Europe. In J. Enders, H. d. Boer, & D. Westerheijden (Eds.), Reform of higher education in Europe (pp. 1– 10). Sense Publishers. EUA. (2011). Aarthus declaration 2011: Investing today in talent tomorrow. European Universities Association. Retrieved January 3, 2022, from http://www.eua.be/Libraries/Policy_Positions/Aarhus_Declaration_2011.sflb.ashx Eurydice. (2016). National student fee and support systems in European higher education 2016/17. Europen Commision. Frenzel, A. C., Pekrun, R., Goetz, T., Daniels, L. M., Durksen, T. L., Becker-Kurz, B., & Klassen, R. M. (2016). Measuring teachers’ enjoyment, anger, and anxiety: The Teacher Emotions Scales (TES). Contemporary Educational Psychology, 46, 148-163. Johnstone, B. & Marcucci P. (2007). Worldwide trends in higher education finance: Cost sharing, student loans, and the support of academic research. UNESCO. Krabbe, P. F. M. (2017). The measurement of health status: Concepts, methods and applications from a multidisciplinary perspective. Elsevier. Leveille, D. (2006) Accountability in higher education: A public agenda for trust and cultural change. Research & Occasional Paper Series. University of California. OECD. (2017). The state of higher education 2015-16. Organisation for Economic Co-operation and Development. Psacharopoulos, G. (2009). Returns to investment in higher education a European survey. Higher education funding reform project. European Commission. Schofer, E. & Meyer, J. (2005). The worldwide expansion of higher education in the twentieth century. American Sociological Review, 70(6), 898-920. Schumacker, R. E., & Lomax, R. G. (2010). A beginners guide to structural equation modeling. Routledge Suchman, M. C. (1995). Managing legitimacy: Strategic and institutional approaches. Academy of Management Review, 20(3), 571-610.