The School of Medicine, University of Zagreb, often searched as MEF Zagreb or Medicinski fakultet Zagreb, is the oldest and most recognized medical faculty in Croatia. Because it combines basic medical sciences, clinical medicine, public health, biomedicine, laboratory research and doctoral projects, statistical data analysis plays a central role in academic and scientific work.
Students, doctoral candidates and researchers often work with clinical data, laboratory measurements, questionnaires, scales, retrospective patient databases, prospective studies, epidemiological data, diagnostic tests, treatment outcomes and biomedical parameters. This article belongs to the F2 hub: statistical help for clinical and scientific research at medical faculties in Croatia.
Statistical data analysis for the School of Medicine, University of Zagreb
The School of Medicine, University of Zagreb, often searched as MEF Zagreb, Medicinski fakultet Zagreb or University of Zagreb School of Medicine, is the oldest medical faculty in Croatia and one of the most important institutions in biomedicine, clinical medicine, public health and scientific research. Because of its broad clinical, laboratory and doctoral activities, statistical data analysis plays a central role in serious academic and scientific work.
Medical research differs from simple student surveys because it often includes clinical outcomes, laboratory values, diagnostic indicators, treatment protocols, longitudinal measurements, retrospective patient databases, prospective studies, questionnaire validation, epidemiological indicators, survival, sensitivity and specificity of tests, and complex relationships between predictors and outcomes.
At StatistischeBeratung.de, professional support is provided throughout the research process — from study design and sample size planning to dataset structuring, method selection, SPSS analysis, interpretation and table preparation. The author remains responsible for the final text and interpretation. This article belongs to the F2 hub: statistical help for clinical and scientific research at medical faculties in Croatia.
Specific characteristics of research at MEF Zagreb
Research at the School of Medicine in Zagreb may include basic medical sciences, clinical medicine, public health, epidemiology, laboratory medicine, biomedicine, pharmacology, pathophysiology, surgery, internal medicine, pediatrics, gynecology, neurology, psychiatry, oncology, radiology, diagnostics and rehabilitation.
Medical data are often not normally distributed, include missing values, different variable types, small subgroups, unequal group sizes, repeated measurements and clinically important covariates. Analysis must therefore consider not only statistical significance but also clinical relevance.
Data types in medical papers
Projects may include demographic data, clinical outcomes, laboratory parameters, diagnostic tests, treatment outcomes, symptom scales, quality-of-life questionnaires, comorbidities, complications, hospitalization, mortality, survival and public health indicators.
Clinical databases often include age, sex, diagnosis, disease stage, disease duration, treatment, laboratory values, imaging findings, surgical data, complications and treatment outcomes. Questionnaire studies may include instrument validation, scale reliability, factor structure, correlations between psychological and clinical variables, and differences between patient groups. For psychological and questionnaire topics, the F3 hub may be relevant: statistical help for humanities and social sciences, and for validation of instruments and rehabilitation assessments the F6 hub: questionnaires, assessments and instrument validation.
Medical research design
Retrospective clinical studies
Retrospective studies often use existing hospital databases, medical records or laboratory registries. The advantage is data availability; challenges include missing values, heterogeneous documentation and limited control over data collection.
Prospective studies
Prospective studies allow better control over measurements, follow-up and outcome definition. Primary outcome, secondary outcomes, sample size and statistical analysis plan should be defined in advance.
Cross-sectional studies
Cross-sectional studies assess prevalence, risk factors, knowledge, attitudes, quality of life or relationships between clinical and sociodemographic indicators at a single time point.
Longitudinal studies and repeated measurements
Longitudinal studies follow patients over time, for example before and after therapy, during rehabilitation or across several follow-up visits. Such data require methods that account for dependence between repeated measurements.
Clinical trials and intervention studies
Intervention studies require careful planning, clear inclusion and exclusion criteria, defined outcomes, control of possible confounders and transparent reporting.
Important statistical methods
Descriptive statistics describe sample structure: age, sex, diagnoses, clinical characteristics, laboratory values and outcomes. Depending on distribution, mean and standard deviation or median and interquartile range are reported.
Group comparisons use t-test, Mann-Whitney U test, ANOVA, Kruskal-Wallis test, chi-square test and Fisher’s exact test. Method selection depends on variable type, distribution and group size.
Correlation analysis examines relationships between laboratory values, clinical scales, disease duration, symptom intensity, functional status and other continuous or ordinal indicators.
Linear, logistic and ordinal regression are used to evaluate predictors of outcomes. Logistic regression is especially important when the outcome is binary, such as complication, positive finding, treatment response or mortality.
ROC analysis evaluates diagnostic accuracy of biomarkers, laboratory parameters and clinical tests. It presents sensitivity, specificity, AUC and possible cut-off values.
Kaplan-Meier curves and Cox regression are used when time-to-event outcomes are important, such as time to relapse, death, complication, hospitalization or treatment failure.
Reliability analysis and instrument validation include Cronbach’s alpha, item analysis, factor analysis, construct validity and correlations.
SPSS analysis and interpretation
SPSS is frequently used in medical papers because it supports group comparisons, correlations, regression, ROC analysis, reliability analysis and graphical presentation. Technical execution alone is not enough.
In medicine, results must be interpreted statistically and clinically. A p-value shows statistical significance, but not automatically effect size, clinical importance or applicability. Confidence intervals, effect sizes, odds ratios, relative risks, AUC values and other indicators are therefore important.
Sample size and statistical power
Medical studies often require sample size planning. Too small a sample may fail to confirm a real effect, while very large samples may show statistically significant but clinically small differences. Sample size should be planned according to the primary outcome, expected effect size, significance level and desired power.
Typical topics for Zagreb
Typical topics include clinical studies, retrospective patient analyses, prospective studies, laboratory biomarkers, diagnostic accuracy, treatment outcomes, epidemiology, public health, quality of life, questionnaire validation, medical rehabilitation, neurological and psychiatric scales, oncology outcomes, cardiovascular risk factors, pediatric data and survival analysis. For rehabilitation and functional outcomes the F4 hub may be useful: sports and kinesiology research, and for experimental biomedical approaches the F8 hub: experimental and biotechnical statistics.
Workflow
The workflow includes reviewing the topic and hypotheses, checking dataset structure, defining primary and secondary outcomes, selecting statistical methods, SPSS analysis, table preparation and interpretation. I prepare statistical results and suggest how to present them in the academic paper. The author remains responsible for the final text and interpretation.
FAQ
Can support be provided before a medical study starts?
Yes. This is the best time to plan study design, sample size, outcomes and database structure.
Can a retrospective hospital database be analyzed?
Yes. Data structure, missing values, variable types and possible confounders must be checked.
Is SPSS sufficient for a medical paper?
For many student papers, master theses and doctoral dissertations, SPSS is sufficient. More complex analyses may require additional tools.
Can ROC analysis and logistic regression be performed?
Yes. ROC analysis and logistic regression are often very useful in diagnostic and clinical research.
Keywords
statistics School of Medicine Zagreb, Medicinski fakultet Zagreb statistics, MEF Zagreb SPSS, clinical research statistics, medical statistics Zagreb, logistic regression medicine, ROC analysis medicine, survival analysis medicine, biomedicine statistics, public health statistics
Metatext
Statistical data analysis for the School of Medicine, University of Zagreb: SPSS, clinical research, ROC analysis, logistic regression, survival analysis, public health and biomedicine.
