The M\u00fcnster Questionnaire for the Evaluation of Seminars \u2013 Revised<\/strong> (German: M\u00fcnsteraner Fragebogen zur Evaluation von Seminaren \u2013 revidiert<\/em>, abbreviated as MFE-Sr<\/strong>) is a specialized psychometric assessment tool developed at the University of M\u00fcnster (Westf\u00e4lische Wilhelms-Universit\u00e4t M\u00fcnster) designed for student evaluations of educational quality within higher education seminar settings. Stemming from the modular M\u00fcnster Evaluation System (M\u00fcnsteraner Fragebogen zur Evaluation<\/em>, MFE), the MFE-Sr was engineered to balance psychometric rigor with extreme methodological economy, providing actionable, multidimensional feedback while minimizing survey fatigue during high-stakes end-of-semester examination periods. The core psychometric core of the instrument comprises 15 primary items organized into four correlated first-order dimensions: Instructor & Didactics<\/em> (Dozent & Didaktik<\/em>; 6 items), Excessive Demand \/ Overload<\/em> (\u00dcberforderung<\/em>; 3 items), Participants \/ Peer Interaction<\/em> (Teilnehmer<\/em>; 3 items), and Instructional Materials<\/em> (Materialien<\/em>; 3 items). These core evaluative items are presented alongside 13 contextual and diagnostic items (yielding 28 total items) measuring attendance, prep time, curricular motives, spatial\/environmental conditions, learning gain, and open-ended feedback.<\/p>\n Psychometric validation using confirmatory factor analysis (CFA) on a curated sample of N<\/em> = 657 university students confirmed the hypothesized four-dimensional factor structure with robust goodness-of-fit indices: χ² = 301.0 (df<\/em> = 84), Tucker-Lewis Index (TLI) = .95, Comparative Fit Index (CFI) = .96, and Root Mean Square Error of Approximation (RMSEA) = .06. The internal consistency coefficients (Cronbach’s α) demonstrate solid reliability across all subscales despite their brevity: Instructor & Didactics (α = .89), Excessive Demand (α = .77), Participants (α = .82), and Materials (α = .83). Criterion and construct validity are supported by substantial correlations with overall course satisfaction (ranging up to r<\/em> = .89 for Instructor & Didactics) and significant discrimination across diverse course types (η² = .33). Core evaluative items employ a 7-point Likert response scale ranging from 1 (“strongly disagree”) to 7 (“strongly agree”), supplemented by a distinct “cannot be answered sensibly” option. The MFE-Sr represents a benchmark in higher education quality assurance, offering evidence-based diagnostic capability tailored for modern web-based institutional evaluation platforms<\/span>.<\/p>\n MFE-Sr, student evaluation of teaching, teaching quality, higher education quality assurance, psychometrics, confirmatory factor analysis, instructional design, academic overload, didactic competence, seminar evaluation<\/p>\n The M\u00fcnster Questionnaire for the Evaluation of Seminars \u2013 Revised (MFE-Sr) was developed, standardized, and validated by academic researchers affiliated with the Institute of Psychology at the Westf\u00e4lische Wilhelms-Universit\u00e4t M\u00fcnster and partner institutions in Germany:<\/p>\n The overarching purpose of the M\u00fcnster Questionnaire for the Evaluation of Seminars \u2013 Revised (MFE-Sr) is to provide higher education institutions, department chairs, and university instructors with a psychometrically validated, standardized, and highly efficient instrument to assess instructional quality in interactive academic seminars. Student evaluations of educational quality (frequently abbreviated as SET or SEEQ) have served as an indispensable pillar of university quality management for over half a century (Schmidt & Lo\u00dfnitzer, 2010). However, traditional evaluation instruments developed during the pen-and-paper era were often cumbersome, frequently consisting of 40 to 80 items. In the contemporary European higher education landscape shaped by the Bologna Process, academic semesters culminate in congested testing windows characterized by severe student exam stress and administrative overload (Bechler & Thielsch, 2012). Long questionnaires distributed under these conditions inevitably trigger survey fatigue, elevated non-response rates, and indiscriminate straight-lining or acquiescence bias.<\/p>\n The MFE-Sr was specifically engineered to overcome these systemic barriers by offering a hyper-economical core module tailored for web-based digital administration (G\u00f6ritz et al., 2005; Haaser et al., 2007). In higher education, seminars possess fundamentally distinct pedagogical dynamics compared to mass lectures: while lectures prioritize unidirectional knowledge dissemination, seminars depend heavily on student-led presentations, reciprocal group discourse, preparation compliance, collaborative inquiry, and critical literature analysis. Consequently, evaluating seminars requires an instrument that does not solely evaluate the lecturer in isolation, but instead captures the multifaceted social and academic ecosystem of the classroom. The MFE-Sr fulfills three essential organizational objectives, as conceptualized in higher education quality research (Souvignier & Gold, 2002):<\/p>\n Furthermore, the MFE-Sr purposefully integrates diagnostic bias variables\u2014including student course prioritization (first choice vs. compulsory enrollment), self-reported absenteeism, weekly preparation hours, and environmental adequacy (acoustics and room facilities). By isolating these extraneous situational and motivational influences, the scale enables instructors and administrators to interpret student ratings within their proper ecological context.<\/p>\n The construct assessed by the MFE-Sr is multidimensional instructional quality in academic seminars<\/em>. Rather than treating teaching quality as a monolithic or unifactorial attribute, modern psychometrics recognizes that classroom learning environments are composed of distinct, interacting pedagogical and psychosocial subsystems. The core psychometric architecture of the MFE-Sr comprises 15 items mapping onto four correlated first-order dimensions, complemented by diagnostic context variables:<\/p>\n The Instructor & Didactics<\/em> subscale measures the pedagogical competence, communicative effectiveness, and structural organization exhibited by the instructor. In the setting of an academic seminar, didactic effectiveness does not merely entail clear oratory; it demands the skillful moderation of intellectual discourse, responsive engagement with student inquiries, cognitive scaffolding, and the ability to maintain conceptual coherence. This 6-item dimension assesses whether the instructor provided a structured overview of the disciplinary domain (Item 8), employed clarifying real-world examples (Item 9), responded constructively and respectfully to student questions and comments (Item 10), presented content in an intellectually stimulating manner (Item 11), maintained a transparent conceptual syllabus\/outline throughout the semester (Item 12), and managed instructional time efficiently (Item 13). High scores on this subscale reflect superior pedagogical expertise, cognitive clarity, and interactive responsiveness.<\/p>\n The Excessive Demand<\/em> subscale quantifies student cognitive and temporal strain. Grounded in Cognitive Load Theory<\/a>, academic learning fails when instructional demands exceed students’ working memory capacity or available study resources. This 3-item subscale captures subjective overburdening across three specific vectors: conceptual difficulty (Item 14: seminar content being too complex or difficult), instructional pacing (Item 15: the pace of knowledge transmission being excessively rapid), and out-of-class workload (Item 16: time investment required exceeding student coping capacity). In contrast to the other subscales, low-to-moderate values on this scale indicate healthy cognitive challenge, whereas elevated scores signal pathological cognitive overload, pacing mismatch, or excessive curricular demands.<\/p>\n The Participants<\/em> subscale reflects the academic engagement, preparation culture, and active participation of the student cohort. Unlike frontal lectures, the pedagogical success of a seminar is deeply co-constructed by the peer group. Even the most gifted instructor cannot compensate for a seminar where students attend unprepared or disengaged. This 3-item dimension operationalizes the social climate of learning by evaluating whether peers were adequately prepared for individual sessions (Item 17), actively contributed to seminar discussions (Item 18), and followed proceedings with attention and intellectual curiosity (Item 19). High scores indicate a vibrant, collaborative, and academically rigorous peer culture.<\/p>\n The Instructional Materials<\/em> subscale evaluates the pedagogical utility, visual quality, and clarity of the physical and digital media integrated into the seminar. Grounded in multimodal learning theory, high-quality artifacts reduce extraneous cognitive load and reinforce deeper conceptual integration. This 3-item scale evaluates whether in-class media (presentation slides, instructional video clips, diagrams, schematic sketches) meaningfully contributed to comprehension (Item 20), whether the intrinsic visual and technical quality of these media was high (Item 21), and whether supplementary materials (e.g., reading packets, handouts, digital repository texts) met high academic standards (Item 22).<\/p>\n Beyond these four core latent factors, the MFE-Sr embeds 13 specialized operational items. Items 1 to 4 assess attendance fidelity, weekly study hours, course choice priority, and specific enrollment motives (interest, time constraints, mandatory course status, lack of alternatives). Items 5 to 7 evaluate structural prerequisites: spatial suitability (seminar room, breakout rooms), acoustic intelligibility, and schedule compatibility. Items 23 and 24 track the specific categories of supplementary materials utilized and perceived material volume. Finally, Items 25 to 28 capture summative outcome indicators: perceived subjective learning gain (binary yes\/no), course recommendation willingness (binary yes\/no), an overall holistic grade mapped to the German upper-secondary 0–15 points grading scale, and qualitative open-ended narrative feedback.<\/p>\n The conceptual architecture of the MFE-Sr is firmly anchored in the Multimodal Conditional Model of Instructional Success<\/em> (Multimodales Bedingungsmodell des Lehrerfolgs<\/em>), formulated and expanded by educational psychologist Heiner Rindermann (1996, 2003, 2009). Rindermann’s model synthesizes decades of instructional effectiveness research, positing that academic learning outcomes are not determined by instructor behaviors alone, but emerge from complex, dynamic interactions among four primary clusters of determinants:<\/p>\n By mapping its measurement dimensions directly onto this systemic model, the MFE-Sr achieves robust ecological validity. Traditional single-score evaluations committed the fundamental attribution error of educational measurement: assigning total responsibility for course success or failure exclusively to the instructor. The MFE-Sr remedies this fallacy by decomposing the seminar experience into instructor-led didactics, student-driven engagement, material utility, and cognitive overload, while simultaneously controlling for contextual confounders (e.g., room acoustics, compulsory course assignment). Furthermore, the theoretical grounding draws heavily upon Self-Determination Theory<\/a> (Deci & Ryan, 2000), which highlights that classroom autonomy, perceived competence, and peer relatedness drive sustained learning motivation. In an academic seminar, the interactive dynamic captured by the Participants<\/em> subscale reflects social relatedness and collective academic competence, which directly modulates instructional success.<\/p>\n The construct, convergent, discriminant, and criterion validity of the MFE-Sr have been empirically substantiated through multi-wave psychometric investigations conducted at the University of M\u00fcnster (Hirschfeld & Thielsch, 2009; Thielsch & Weltzin, 2012):<\/p>\n The content validity of the instrument originates from a multi-stage iterative item generation protocol (Grabbe, 2003; Haaser, 2006). Based on systematic reviews of higher education literature regarding the observable characteristics of high-quality university seminars, initial pools of 17 and 14 items were refined, modified, and expanded to represent all essential facets of seminar quality without redundancy. Aligning with Rindermann’s multimodal framework, the primary operational focus remains on instructor behavioral didactics, while giving proportional representation to participant dynamics, cognitive overload, and instructional media.<\/p>\n Convergent validity was evaluated at the course level across an aggregated dataset of 93 distinct seminar courses. Scale aggregate scores were correlated with students’ holistic summative evaluation of the overall seminar (measured via an open-ended 0–15 point academic grading metric). The observed correlations demonstrated robust, theoretically coherent relationships:<\/p>\n Divergent validity was examined by correlating MFE-Sr subscales with potential non-instructional biasing factors, particularly students’ initial enrollment preference (first-choice vs. second-choice vs. third-choice course allocation). Spearman rank correlations between enrollment priority and the subscales Excessive Demand<\/em>, Participants<\/em>, and Instructional Materials<\/em> were non-significant, demonstrating that pre-existing course allocation priorities do not artifactually bias these evaluative dimensions. Although a statistically significant rank correlation emerged between course preference and Instructor & Didactics<\/em> (r<\/em> = -.13, p<\/em> < .05), the effect size is negligible (< 2% explained variance), establishing that ratings primarily reflect actual in-class didactic performance rather than prior course selection bias<\/span>.<\/p>\n Furthermore, discriminant validity across course formats was confirmed using multivariate analysis of variance (MANOVA), with course type as the independent factor and the four MFE-Sr subscales as dependent variables. The omnibus test demonstrated significant between-course discrimination: F<\/em>(368) = 3.06, p<\/em> < .01, η² = .33. This indicates that one-third of the total multivariate variance in questionnaire ratings is directly attributable to systematic differences across courses, confirming that the MFE-Sr sensitively detects true pedagogical variation<\/span>.<\/p>\n The internal consistency of the MFE-Sr was evaluated using Cronbach’s α based on the standard validation sample of N<\/em> = 657 completed, cleaned seminar evaluations collected during the winter semester 2009\/2010. Despite the deliberate economy of the subscales (consisting of only 3 to 6 items each), all dimensions exceed conventional psychometric benchmarks for group-level diagnostic instruments<\/span>:<\/p>\n These reliability values correspond favorably with much longer, established German instruments such as the FEVOR (Staufenbiel, 2000), HILVE (Rindermann, 2009), KIEL (Gediga et al., 2000), and TRIL (Gollwitzer & Schlotz, 2003), while requiring a fraction of respondent completion time.<\/p>\n The latent structure of the MFE-Sr was verified using Confirmatory Factor Analysis<\/a> (CFA) executed in AMOS via Maximum Likelihood (ML) estimation on the N<\/em> = 657 validation dataset. The hypothesized structural model posited four correlated first-order latent factors accounting for the covariance among the 15 core items<\/span>.<\/p>\n The four-factor measurement model achieved an acceptable-to-good empirical fit according to established methodological standards (Hu & Bentler, 1999):<\/p>\n The standardized factor loadings (λ), item means (M<\/em>), standard deviations (SD<\/em>), corrected item-total correlations (T<\/em>), and Cronbach’s alpha if item deleted (αdel<\/sub>) are summarized below:<\/p>\n The inter-factor correlations (N<\/em> = 657) illustrate that Instructor & Didactics<\/em> is strongly and positively correlated with Instructional Materials<\/em> (r<\/em> = .76) and Participants<\/em> (r<\/em> = .66). The Participants<\/em> and Instructional Materials<\/em> subscales correlate at r<\/em> = .65. In contrast, Excessive Demand<\/em> correlates negatively with all three supportive learning dimensions: Instructor & Didactics<\/em> (r<\/em> = -.23), Participants<\/em> (r<\/em> = -.22), and Materials<\/em> (r<\/em> = -.19)<\/span>.<\/p>\n Descriptive statistics reveal substantial negative skewness across the positive dimensions (Instructor & Didactics: Median = 6.17, Mean = 5.86, SD<\/em> = 1.06, Skewness = -1.58, Kurtosis = 3.10; Materials: Median = 6.00, Mean = 5.88, SD<\/em> = 0.99, Skewness = -1.64, Kurtosis = 4.36; Participants: Median = 5.67, Mean = 5.56, SD<\/em> = 1.04, Skewness = -0.96, Kurtosis = 1.28). In contrast, Excessive Demand<\/em> displays positive skewness (Median = 2.00, Mean = 2.36, SD<\/em> = 1.16, Skewness = 1.09, Kurtosis = 1.06), confirming that students generally experience low-to-moderate levels of overload in standard university seminar curricula<\/span>.<\/p>\n The MFE-Sr is structured as a multidimensional, standardized educational evaluation battery designed primarily for online computer-assisted web interviewing (CAWI) or mixed-mode administration.<\/p>\n The M\u00fcnster Questionnaire for the Evaluation of Seminars \u2013 Revised (MFE-Sr) was developed and validated in its revised format between 2009 and 2010 at the Westf\u00e4lische Wilhelms-Universit\u00e4t M\u00fcnster (Psychological Institute I), following earlier foundational iterations dating back to 2002\/2003 (Grabbe, 2003; Haaser, 2006; Hirschfeld & Thielsch, 2009). The questionnaire was released into the academic public domain for scientific research, academic teaching evaluation, and institutional quality assurance purposes.<\/p>\n There are no licensing fees or commercial purchase requirements associated with utilizing the MFE-Sr for institutional quality assurance or non-commercial scholarly research. Higher education institutions and academic investigators are permitted to implement the items within web-based learning management systems (e.g., Moodle, ILIAS, EvaSys) or paper-pencil formats, provided proper academic attribution is maintained. Inquiries concerning system integration, normative comparative data benchmarks, or collaborative research should be directed to Dr. Meinald Thielsch at thielsch@uni-muenster.de<\/a>.<\/p>\nKeywords<\/h2>\n
Authors<\/h2>\n
\n
Purpose<\/h2>\n
\n
Psychological Construct<\/h2>\n
1. Instructor & Didactics (Dozent & Didaktik<\/em>)<\/h3>\n
2. Excessive Demand \/ Overload (\u00dcberforderung<\/em>)<\/h3>\n
3. Participants \/ Peer Interaction (Teilnehmer<\/em>)<\/h3>\n
4. Instructional Materials (Materialien<\/em>)<\/h3>\n
Contextual, Environmental, and Diagnostic Bias Variables<\/h3>\n
Theoretical Framework<\/h2>\n
\n
Validity<\/h2>\n
Content Validity<\/h3>\n
Convergent and Criterion Validity<\/h3>\n
\n
Discriminant and Divergent Validity<\/h3>\n
Reliability<\/h2>\n
\n
Factor Analysis<\/h2>\n
Goodness-of-Fit Indices<\/h3>\n
\n
Standardized Factor Loadings and Item Statistics<\/h3>\n
\n\n
\n Subscale & Item Number<\/th>\n M<\/em><\/th>\n SD<\/em><\/th>\n T<\/em><\/th>\n Factor Loading (λ)<\/th>\n α if Deleted<\/th>\n<\/tr>\n<\/thead>\n \n\n Instructor & Didactics (α = .89)<\/strong><\/span><\/td>\n<\/tr>\n \n Item 8 (Overview of subject area)<\/td>\n 5.94<\/td>\n 1.18<\/td>\n .74<\/td>\n .79<\/td>\n .87<\/td>\n<\/tr>\n \n Item 9 (Use of explanatory examples)<\/td>\n 5.85<\/td>\n 1.28<\/td>\n .77<\/td>\n .82<\/td>\n .87<\/td>\n<\/tr>\n \n Item 10 (Responsiveness to student questions)<\/td>\n 6.09<\/td>\n 1.29<\/td>\n .71<\/td>\n .76<\/td>\n .87<\/td>\n<\/tr>\n \n Item 11 (Interesting presentation)<\/td>\n 5.69<\/td>\n 1.41<\/td>\n .79<\/td>\n .86<\/td>\n .86<\/td>\n<\/tr>\n \n Item 12 (Traceable structure\/outline)<\/td>\n 5.87<\/td>\n 1.31<\/td>\n .65<\/td>\n .68<\/td>\n .88<\/td>\n<\/tr>\n \n Item 13 (Effective time management)<\/td>\n 5.71<\/td>\n 1.38<\/td>\n .64<\/td>\n .67<\/td>\n .89<\/td>\n<\/tr>\n \n Excessive Demand \/ Overload (α = .77)<\/strong><\/span><\/td>\n<\/tr>\n \n Item 14 (Content too difficult)<\/td>\n 2.22<\/td>\n 1.27<\/td>\n .68<\/td>\n .86<\/td>\n .62<\/td>\n<\/tr>\n \n Item 15 (Pace too fast)<\/td>\n 2.19<\/td>\n 1.33<\/td>\n .69<\/td>\n .87<\/td>\n .60<\/td>\n<\/tr>\n \n Item 16 (Time expenditure overtaxing)<\/td>\n 2.67<\/td>\n 1.61<\/td>\n .48<\/td>\n .52<\/td>\n .86<\/td>\n<\/tr>\n \n Participants (α = .82)<\/strong><\/span><\/td>\n<\/tr>\n \n Item 17 (Preparation of participants)<\/td>\n 5.49<\/td>\n 1.26<\/td>\n .58<\/td>\n .61<\/td>\n .85<\/td>\n<\/tr>\n \n Item 18 (Active contributions of peers)<\/td>\n 5.58<\/td>\n 1.18<\/td>\n .74<\/td>\n .82<\/td>\n .68<\/td>\n<\/tr>\n \n Item 19 (Attentive\/interested participation)<\/td>\n 5.61<\/td>\n 1.20<\/td>\n .71<\/td>\n .90<\/td>\n .71<\/td>\n<\/tr>\n \n Instructional Materials (α = .83)<\/strong><\/span><\/td>\n<\/tr>\n \n Item 20 (Media aided comprehension)<\/td>\n 6.02<\/td>\n 1.08<\/td>\n .75<\/td>\n .90<\/td>\n .70<\/td>\n<\/tr>\n \n Item 21 (Visual\/technical media quality)<\/td>\n 5.96<\/td>\n 1.11<\/td>\n .75<\/td>\n .88<\/td>\n .70<\/td>\n<\/tr>\n \n Item 22 [listed in source table as Item 23] (Quality of additional materials)<\/td>\n 5.66<\/td>\n 1.25<\/td>\n .57<\/td>\n .62<\/td>\n .88<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n Inter-Factor Correlations and Descriptive Distributions<\/h3>\n
Instrument \/ Measurement Tool<\/h2>\n
\n
\n
\n
\n
\n
Permissions & Fee and Test Year<\/h2>\n
References<\/h2>\n
\n