Profiling Job Interview Difficulties among Management Information Systems Graduates Using K-Means Clustering


Creative Commons License

Kaldırımcı Ş. G., Özen Ü., Gökdağ M. B., Seven S., İpkin M.

IConMIS 2026 International Congress on Management Information Systems, Antalya, Türkiye, 15 - 17 Ekim 2026, ss.141-142, (Özet Bildiri)

  • Yayın Türü: Bildiri / Özet Bildiri
  • Basıldığı Şehir: Antalya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.141-142
  • Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
  • Atatürk Üniversitesi Adresli: Evet

Özet

Background/Aims: Management Information Systems (MIS) graduates may experience job interview difficulties arising from both individual/professional and contextual conditions. This study identifies the dimensions of these difficulties, derives profiles, and examines their relationships with demographic, academic, and employment characteristics.

Materials/Method: Survey data from 364 graduates of the MIS Department at Atatürk University included ten Likert-type items measuring job interview difficulties. Internal consistency was assessed using Cronbach’s alpha, and the factor structure was examined through principal axis factoring with Direct Oblimin rotation. After screening, 304 participants who answered at least six of the seven retained items were included; six missing responses were median-imputed. Individual/professional and contextual difficulty scores were calculated and standardized. K-means solutions with two to six clusters were compared using the Silhouette, Calinski-Harabasz, and Davies-Bouldin indices. Stability was evaluated through 1,000 bootstrap replications. Associations with gender, work experience, postgraduate education, age, graduation year, and grade point average were tested using chi-square and Kruskal-Wallis tests with Holm correction.

Results: Internal consistency was satisfactory (Cronbach’s α = 0.854). The Kaiser-Meyer-Olkin value was 0.834, Bartlett’s test of sphericity was significant (p < 0.001), and the two-factor structure explained 53.261% of the variance. Considering validation indices, parsimony, and interpretability, a three-cluster solution was selected (Silhouette = 0.412; Calinski-Harabasz = 265.752; Davies-Bouldin = 0.863). The profiles were Low Difficulty (n = 154; 50.66%), Individual/Professional Difficulty-Dominant (n = 96; 31.58%), and Multidimensional High Difficulty (n = 54; 17.76%). The mean bootstrap Adjusted Rand Index was 0.843, and profile-specific Jaccard coefficients ranged from 0.863 to 0.931. No significant associations with external variables remained after Holm correction.

Conclusion/Implications: The findings provide an exploratory basis for understanding heterogeneity in graduates’ job interview difficulties and may inform the future development of profile-sensitive career support. Further research should validate these profiles and assess their practical relevance before profile-based interventions are implemented.