Instruction

No specific instructions are required, but we recommend using the following wording: “To what extent do the following statements apply to your studies?” (“Wie sehr treffen die folgenden Aussagen auf Ihr Studium zu?“).

Items

Table 1

Items of the Scale to Assess Study Satisfaction

Note. The unvalidated English version of the Scale is based on the exact wording of the Satisfaction with Life Scale (Diener et al., 1985) with the exception of the words life and live changed to studies and start.

Response specifications

All five items were rated on a seven-point Likert-scale with the response options 1 – completely disagree (trifft überhaupt nicht zu), 2 – disagree (trifft nicht zu), 3 – rather disagree (trifft eher nicht zu), 4 – do not agree nor disagree (teils/ teils), 5 – rather agree (trifft eher zu), 6 – agree (trifft zu), 7 - completely agree (trifft vollkommen zu).

Scoring

All items are positively polarized and are not differentiated into subscales. High values indicate high study satisfaction and low values indicate low study satisfaction. A simple total score for study satisfaction is formed based on all five items. For situations with severe time constraints, study satisfaction can also be measured based on the three-item scale (items 1, 2, and 3) or with a single item (item 3).

Application field

The scale was mainly developed for the assessment of study satisfaction in higher education research. For instance, the scale can be used to identify predictors and consequences of study satisfaction or to investigate the temporal development of study satisfaction over the course of studies. The target group are higher education students of all study programs and all types of higher education institutions (e.g., universities, universities of applied sciences, distance learning universities). The intended survey mode for the scale is online computer-assisted self-administered interviewing, but the scale can also be used with paper-and-pencil self-administered interviewing. Based on the processing time in Study 1, we can confirm that the scale can be answered by most students (88.81 %) in under a minute.

Study satisfaction represents students’ subjective evaluation of their studies in higher education as positive or negative (Westermann & Heise, 2018). Following theoretical ideas on the nature of subjective well-being, study satisfaction is possibly best understood as a domain-specific component of the cognitive-evaluative aspect of subjective well-being (in terms of Diener et al., 1985). Accordingly, study satisfaction shows positive associations with other components of well-being (e.g., positive and negative affect; Grunschel et al., 2016).

In addition, study satisfaction and study motivation should be strongly correlated, since both constructs are understood as encompassing cognitive and affective components (e.g., Bong & Skaalvik, 2003; Eccles & Wigfield, 2020) within the domain of higher education. Empirical studies found positive relations of study satisfaction with the academic self-concept (Kegel et al., 2021) and with positive value (Kryshko et al., 2022) as well as negative relations with motivational costs (Gadosey et al., 2022). Study satisfaction is also negatively related to academic procrastination (Gadosey et al., 2022; Grunschel et al., 2016). In higher education, academic procrastination as a dysfunctional learning behavior is highly prevalent and results in various negative consequences for students (Steel & Klingsieck, 2016). Furthermore, study satisfaction is a domain of academic success that interrelates with other academic success domains. For instance, study satisfaction is a key predictor of dropout intention in higher education studies (Bean & Metzner, 1985; Fleischer et al., 2019). In other words, more satisfied students report lower dropout intentions (Janke, 2020; Kegel et al., 2021).

Even though prior research seems to speak to the importance of study satisfaction, it is generally limited by the applied methodology. It is noteworthy that a lack of consensus between researchers is not the issue as the “Fragebogen zur Studienzufriedenheit” (Schiefele & Jacob-Ebbinghaus, 2006; Westermann et al., 1996) is dominating research on study satisfaction within the German context. However, scale development of this particular instrument was not strongly embedded in a psychological theory. Consequently, the subscales of the instrument (satisfaction with study content, study conditions, and coping with study-related stress) lack a strong justification in their relation to the underlying construct. It is unclear why these fine-grained aspects of study satisfaction are included, while others are neglected (e.g., satisfaction with relationships, personal autonomy, or growth). Furthermore, six out of nine items are negatively worded, which indicates that the instrument is possibly better suited to assess ill- instead of well-being. Finally, the included subscales often seem to mirror related constructs such as personal interest (e.g., “I find my study program really interesting.”; Schiefele & Jacob-Ebbinghaus, 2006). Consequently, it is difficult to pinpoint whether associations of the scale with other aspects of well-being and motivation indicate validity or construct overlap. Taken together, the development of an instrument that represents a purer measure of study satisfaction is strongly mandated.

Item generation and selection

We aimed to ensure that the constructed measure was embedded in a psychological theory. More precisely, we considered study satisfaction as a domain-specific aspect of subjective well-being in terms of Diener et al. (1985). In general, there is no clear rationale in how researchers may approach the development of domain-specific measures of the cognitive component of subjective well-being. Yet there is strong consensus on the measurement of overarching subjective well-being. For this purpose, the Satisfaction With Life Scale by Diener et al. (1985) has been validated in several cultures and translated in many languages including German (Schuhmacher, 2003). For the present scale, we adapted the items of the German translation to the higher education context. This procedure was carried out in two steps. First, all six authors separately adapted the five items to reflect study satisfaction instead of life satisfaction. Subsequently, the items were selected for the scale that all authors considered to be closest to the original wording and most comprehensible.

Guided by the validated three-item life satisfaction scale (Kjell & Diener, 2021), we selected the first three items to form a short scale measuring study satisfaction. Furthermore, we chose item three for a single-item measure as it yields the most direct wording to represent overall study satisfaction from our point of view. The five-item scale, three-item scale, and single-item measure were then tested in two samples of students.

Samples

Study 1. Data collection was part of a cross-sectional study in January and February 2024 with the main purpose of examining self-reported cheating of higher education students in the academic context. Higher education students were recruited via social media, posters, and mailing lists. They received either credit points for their studies (only students of psychology or human exercises) or could win one out of five book vouchers for 25€ each. Participants who stated that they had not answered the survey honestly (n = 9) or did not select the instructed response option for two items that were included in the survey to test participants’ attention (n = 4) were excluded. A total of N = 420 (73.10 % female, 23.81 % male, 1.43 % diverse) participants were included in our analyses. They were on average M = 23.31 (SD = 3.89) years old, studied in their M = 4.94 (SD = 2.63) semester, and studied various majors at German universities (e.g., 35 % psychology, 12 % teaching, 7 % law, 4 % computer sciences).

Study 2. Data collection was conducted with online surveys from January to April 2024. The main purpose of the project was to assess study motivation and learning behavior in the 14 days prior to each participants’ individual first exam of the winter semester 2023/2024. We assessed study satisfaction in the preliminary survey before students’ first exam of the semester and in the post survey at the start of the following semester. Participants from all higher education institutions in Germany were recruited via social media, mailing lists, posters, and announcements in various lectures. For their participation, students received either credit points for their studies (only students of psychology or human exercises) or could win one out of 15 monetary rewards each worth 100 euros. Overall, N = 471 students (81.10 % female, 17.62 % male, 0.06 % diverse) were included in the analyses. They studied at 31 German universities, were M = 26.11 years old (SD = 8.69), studied in their M = 3.49 semester (SD = 2.68), and studied different majors with a large proportion of psychology students (83.01 %).

Item analyses

All analyses were conducted using the statistic software R (R Core Team, 2024) with packages lavaan (version 0.6.17; Rosseel, 2012) and semTools (version 0.5.6; Jorgensen et al., 2022). We analyzed the unidimensional structure of the five-item scale with confirmatory factor analyses with robust maximum likelihood estimation. Standardized path coefficients of both studies are displayed in Figure 1. Following recommendations of Hu and Bentler (1998), in Study 1 most fit indices showed overall good fit (CFI = .969, SRMR = .038, c²(5) = 28.467, p < .001, N₁ = 420). Only the Root Mean Square Error of Approximation (RMSEA = .093) was higher than the recommended cut-off value. In Study 2, again, most fit indices showed overall good fit at T1 (CFI = .972, SRMR = .035, c²(5) = 32.047, p < .001, N₂ = 471), but the RMSEA with a value of .103 exceeded the recommended cut-off value. However, the RMSEA often falsely indicates a poor fit in models with small degrees of freedom and small samples (Kenny et al., 2015). Therefore, we report the RMSEA for completeness but in line with recommendations by Kenny et al. (2015) do not dismiss the one-factor model for study satisfaction. In addition, we analyzed the one-factor structure of the three-item scale with confirmatory factor analyses in both studies (Figure 2).

Figure 1

One Factor Confirmatory Factor Analysis for Study 1 (left) and Study 2 (right) for the Five-Item Scale

Note. Standardized path coefficients, N₁ = 420, N₂ = 471.

Figure 2

One Factor Confirmatory Factor Analysis for Study 1 (left) and Study 2 (right) for the Three-Item Scale

Note. Standardized path coefficients, N₁ = 420, N₂ = 471.

Item parameters

Table 2

Means, Standard Deviations, Skewness, Kurtosis, and Selectivity of the Manifest Items in Study 1 / Study 2

Note. Scale ranging from 1 (completely disagree) to 7 (completely agree), N₁ = 420, N₂ = 471, M = mean, SD = standard deviation, selectivity = corrected part-whole-correlation.

Objectivity

The survey is meant to be used as an online or paper-based survey and was validated online in CASI mode. Standardized instructions and items are presented so that the survey can be conducted objectively.

Reliability

Internal consistency of the five-item and three-item scale was determined using McDonald's Omega-h and Cronbach’s Alpha. In both studies the coefficients for the five-item scale showed good internal consistency (α₁ = .83, ω₁ = .82, α₂ = .85, ω₂ = .83). For the three-item scale, internal consistency was also good in both studies (α₁ = .80, ω₁ = .81, α₂ = .81, ω₂ = .82). For Study 2 we could estimate a retest-reliability of r = .71 (p < .001) based on data of N = 301 students who participated in both measurement points.

Validity

Using data from Study 1 and the first measurement point of Study 2, we analyzed the correlations of our five-item scale, three-item scale, and single item with the three subscales of study satisfaction by Schiefele and Jacob-Ebbinghaus (2006) to examine associations with existing measures for study satisfaction as an indicator for validity. Each of our measures showed a strong positive correlation with the subscales of study satisfaction. Correlations were highest with the subscale satisfaction with study content.

In addition, we used data from Study 2 to examine how our three measures correlated with academic self-concept (Dickhäuser et al., 2002), intrinsic value (Schnettler et al., 2020), motivational costs (Schnettler et al., 2020), the behavioral as well as the emotional component of academic procrastination (Bobe et al., 2022), and dropout intentions (Bäulke et al., 2021). The correlations found for motivational constructs were likewise strong and, as expected, positive for the academic self-concept and intrinsic value but negative for motivational costs. The relation of study satisfaction with learning behavior in the form of academic procrastination was negative and smaller than the correlations of study satisfaction with motivational constructs. The correlation with the academic success domain of dropout intention was negative and moderately high. These results show that our scale for measuring study satisfaction relates to study-relevant variables as we expected based on theory and previous empirical findings, indicating validity.

Table 3

Correlations of the Five-Item Scale, Three-Item Scale, and Single Item to Assess Study Satisfaction With Further Study-Relevant Variables

Note. N₁ = 420, N₂ = 471, three-item scale contains items one, two, and three, single item is item three, all correlations were significant with p < .01.

Descriptive statistics (scaling)

The five-item and three-item scales to assess study satisfaction showed comparable descriptive statistics in Study 1 and Study 2 (see Table 4). This finding is similar to the comparison of the five-item and three-item Satisfaction with Life Scale (Kjell & Diener, 2021).

Table 4

Means, Standard Deviations, Skewness, and Kurtosis of the Five-Item Scale and Three-Item Scale in Study 1 / Study 2

Note. Scale ranging from 1 (completely disagree) to 7 (completely agree), N₁ = 420, N₂ = 471, M = mean, SD = standard deviation.

Further quality criteria

This scale enables a very time- and cost-efficient assessment of study satisfaction. In Study 1 mean processing time of the survey page to assess study satisfaction including instruction, the five items of interest, and two control items was 41.71 seconds (SD = 30.83 seconds) and 88.81 % of participants completed this survey page in under 60 seconds.

We examined measurement invariance for gender (female and male) in Study 1 and Study 2, for study program (psychology and others) in Study 1 and Study 2, for the type of higher education institution (distance learning university and traditional face-to-face learning university) in Study 2, and over time in Study 2. For this purpose, we compared fit indices of progressively restricted multigroup CFAs by using cut-off values suggested by Chen (2007; metric: ΔCFI ≤ −.010, ΔRMSEA ≥ .015, ΔSRMR ≥ .030; scalar: ΔCFI ≤ −.010, ΔRMSEA ≥ .015, ΔSRMR ≥ .010; strict: ΔCFI ≤ −.010, ΔRMSEA ≥ .015, ΔSRMR ≥ .010). The five-item scale showed strict invariance for gender (Table 5), study program in Study 2 (Table 6), type of higher education institution (Table 7), and over time (Table 8).

Table 5

Measurement Invariance for the Five-Item Scale to Assess Study Satisfaction for Gender in Study 1 and Study 2

Note. N₁ = 407, female group n = 307, male group n = 100, N₂ = 465, female group n = 382, male group n = 83, all c² are significant with p < .001.

Table 6

Measurement Invariance for the Five-Item Scale to Assess Study Satisfaction for Study Program in Study 1 and Study 2

Note. N₁ = 420, psychology group n = 142, other study programs-group n = 278, N₂ = 471, psychology group n = 391, other study programs-group n = 80, all c² are significant with p < .001.

Table 7

Measurement Invariance for the Five-Item Scale to Assess Study Satisfaction for Students From Distance Learning Universities and Face-to-Face Universities in Study 2

Note. N = 471, students from distance learning universities n = 195, students from traditional universities n = 276, all c² are significant with p < .001.

Table 8

Measurement Invariance for the Five-Item Scale to Assess Study Satisfaction for Students Over Time in Study 2

Note. N = 301, all c² are significant with p < .001.

Acknowledgement

We would like to thank all participants who spent their time completing the surveys and thus supported our research projects. Carola Grunschel, thank you very much for your continuous support at all levels for the realization of both studies. Many thanks to Jeannine Kecker and Moira Andrä for the implementation of Study 1, and to Lucas Wloch for the realization of Study 2.