In our glossary, statistics students and interested parties will find numerous instructions, tips and interesting facts from the world of statistics. This section is still growing, with new articles being added every week. If you have any comments or suggestions, we look forward to receiving your e-mail!
Everything about SPSS
General information about SPSS
Here you will find information on the history and versions of SPSS as well as information on the areas in which SPSS can provide assistance, the user interface and also differences between the SPSS software and Excel and R.
Bivariate statistics
This section describes methods for the statistical analysis of relationships between two variables. Specifically, the selection and interpretation of key figures for cross-tabulations and correlations in SPSS are presented.
Chi square test
The two chi-square test procedures in SPSS are explained here: The distribution test checks the equal distribution of the available data and the independence test examines the extent to which two variables are independent of each other.
Cluster analysis
The SPSS cluster analysis is presented here as an example of a structure-detecting multuvariate method.
Discriminant analysis
Discriminant analysis in SPSS is a multivariate method for classification analysis and is similar to the logistic regression method.
Graphical representations in SPSS
In addition to analytical functions, SPSS offers an extensive editor for graphical representations. The following article focuses on the following chart types: Bar chart, histogram, scatter plot, box plot in SPSS – these functions of SPSS are a great statistical aid in the analysis and presentation of data.
Correlation SPSS
This section explains the correlation analysis in SPSS using the Pearson and Spearman correlation coefficients.
Non-parametric tests for differences
If no normal distribution of the data can be assumed, non-parametric median tests are used as an alternative: specifically, the Wilcoxon rank test and the Mann-Whitney U-test are presented.
Parametric tests for differences
This section is dedicated to the comparison of mean values, specifically the SPSS t-test for independent samples and the t-test for related samples.
Regression SPSS
The text discusses regression in general in the mathematical model. In addition to linear regression SPSS, multiple regression SPSS and logistic regression SPSS are discussed.
Significance tests
Here you will find a general overview of the terms null hypothesis and alternative hypothesis as well as the concept of significance / p-value. The chapter also provides an overview of different parametric and non-parametric test procedures.
Test for normal distribution
Statistical test procedures in SPSS for checking the normal distribution in SPSS are presented here: Kolmogorov-Smirnov test (KS test) and Shapiro-Wilk test.
Univariate statistics
Univariate analysis describes methods for SPSS analysis of a singularly considered variable. Specifically, you will find information on the scaling of data and the common position and scattering parameters as well as an example for calculating and interpreting the mean value.
Analysis of variance SPSS
The analysis of variance, also known as ANOVA in SPSS, is explained here. The ANOVA tests possible differences between a dependent characteristic and an independent characteristic.
Statistics
Business analytics
Business analytics, also known as “business intelligence” (BI), refers to the evaluation of business data using statistical methods. The aim is to support companies in their decisions based on the evaluations, e.g. with statistical consulting.
Descriptive statistics
Descriptive statistics describe a data set and its characteristics. The aim of descriptive (descriptive) statistics is usually to describe a sample of empirical data.
Experimental economic research
Experimental economic research conducts controlled trials, experiments, real-life field trials and the evaluation of surveys and questionnaires in order to investigate economic issues, develop or test theories – e.g. through statistical consulting.
Hypothesis test
A hypothesis test is a frequently used statistical procedure to test the certainty with which a fact can be assumed to be true or false on the basis of data.
Multivariate statistics
Multivariate statistics are a family of statistical methods for analyzing dependencies between more than two variables. Some multivariate methods are presented here and where they are used.
Predictive analytics
Predictive analytics refers to data analysis methods that use statistical methods to predict future events. Smart statistical data analysis is used to create forecasts that are as reliable as possible.
P value statistics
In statistics, the P-value is the test variable that measures the statistical significance of an observation. The P-value is a probability value that indicates the probability of finding the values of the sample, assuming that the null hypothesis is true.
Regression statistics
With the help of regression analysis, data can be used to determine which variables influence each other strongly or less strongly. This section describes how to calculate the statistical regression between two or more variables and the quality of the regression.
R statistics program
The R statistics program is free, but has very extensive and constantly growing functionality. Here we describe what R is, what it can do, what its strengths are and how you can learn to use it.
Significance statistics
Statistical significance measures whether an observation could actually have a meaning or could have arisen by chance. Here we explain: What does statistical significance depend on and how is it calculated?
Statistical data
The right choice of statistical data is essential for the success of statistical evaluations of all kinds. There are typical formats and scales that are usually used.
Statistical tests
Statistical tests check whether an observation is statistically significant. Depending on the objective and data set, various statistical test methods are suitable for checking significance.
Statistical methods
Different statistical methods are used depending on the objectives and/or hypotheses of the statistical analysis.
Significance test
In a hypothesis test, a hypothesis is tested and either the null hypothesis or the counter-hypothesis is assumed to be correct. The question then arises as to the significance level at which the hypothesis was accepted.
Data Mining
Data mining definition
Collection of methods to generate knowledge from data. Why and how is data mining carried out? What are the steps involved? What makes data mining difficult?
Data Mining Software
The types of software tools that support data mining are explained here. In addition, some exemplary tools are briefly described and the most common data mining tools are linked.
Medical statistics
Continuous Outcome Superiority Trial
Discover how continuous outcome data is used to provide statistical evidence of the superiority of a new therapy over an established treatment. This article introduces the principles and methodology of a continuous outcome superiority design.
Continuous Outcome Non-Inferiority Trial
Learn how continuous outcome measures are used in non-inferiority trials to prove that a new intervention is not significantly worse than the standard. The article explains the specific requirements and challenges of this study design.
Binary Outcome Non-Inferiority Trial
This article shows how to use binary endpoints to statistically prove that a new treatment strategy is non-inferior. It provides a practical overview of the design and analysis of non-inferiority trials with discrete outcomes.
Binary Outcome Superiority Trial
Learn how binary outcome data are used to demonstrate the benefit of a new therapy compared to a conventional treatment. The article presents the methodological features and interpretation approaches of a superiority design with binary endpoints.
Binary Outcome Equivalence Trial
This article explains how studies with binary endpoints are designed to demonstrate the equivalence of two treatment approaches. It offers insights into the statistical requirements and the practical implementation of equivalence tests.
Continuous Outcome Equivalence Trial
Learn how continuous outcome data are used to prove the equivalence of interventions. The article highlights the statistical methods and challenges that need to be considered when planning and conducting continuous outcome equivalence trials.
Amendment
In clinical trials, an “amendment” refers to a change or addition to the protocol or other trial-related documents. It is a formal written communication that is passed on to all parties involved, such as the investigators, trial centers, ethics committees and regulatory authorities.
AMG
The AMG stands for the German Medicinal Products Act. It is a law that regulates the authorization, manufacture, dispensing, monitoring and handling
of medicinal products in Germany. The Medicinal Products Act contains provisions and regulations intended to ensure that medicinal products are effective, of high quality
and safe. Among other things, the AMG specifies the requirements that a medicinal product must fulfill in order to be approved and distributed on the market
. It also regulates the approval procedures for new medicinal products, the conduct of clinical trials, the labeling and
package leaflet of medicinal products as well as pharmacovigilance, i.e. the recording, assessment and monitoring of drug risks.
Observational study
An observational study, also known as an “observational study” or “post-marketing study”, is a study conducted after the approval of a medicinal product or medical device. In contrast to clinical trials, which are conducted prior to approval, an observational study aims to gather further information on the use, efficacy and safety of a product in real clinical
practice.
Number of treatments needed
The Number Needed to Treat (NNT) is a concept in medical statistics that indicates how many patients need to be treated to achieve success. A low NNT means that the treatment is more effective, while a high
NNT indicates that the treatment is less effective. The NNT is often used to evaluate the effectiveness of drugs or other therapies and can help inform medical decisions and allocate resources more effectively.
Audit
An audit is a systematic and independent examination process that aims to evaluate the conformity, effectiveness and efficiency of processes, systems,
organizations or products. It is a methodical examination and review aimed at gaining information and knowledge about the audited areas.
Notified body
A notified body is an institution or organization authorized by the legislator or a state authority to perform certain tasks in the
context of certification, conformity assessment or surveillance of products, services or systems. It is endowed with specific
powers and responsibilities and has the competence to check conformity with certain legal requirements, standards or technical regulations.
Observational study
An observational study is a type of study design in medical statistics in which no intervention or experimental manipulation is performed. Instead, data is collected from an existing population or cohort and analyzed to identify possible relationships or associations between variables. Observational studies can help answer questions about the frequency and distribution of diseases, causes and risk factors. However, observational studies are prone to bias and error, and it can be difficult to establish causality between variables.
BfArm
The BfArM stands for the Federal Institute for Drugs and Medical Devices. It is a German higher federal authority that reports to the Federal Ministry
of Health. The BfArM is responsible for the approval, monitoring and control of medicinal products, medical devices and blood products in Germany.
Bland-Altman analysis
The Bland-Altman analysis is a method for evaluating the agreement between two diagnostic or measurement procedures in medical statistics. The differences between the results of the two procedures are plotted against the mean value of these results. The analysis makes it possible to identify systematic differences and random fluctuations between the procedures. By using Bland-Altman plots, outliers and incorrect measurements can also be identified. Bland-Altman analysis is a useful method for assessing the validity and reproducibility of measurement procedures in medical research.
Higher federal authority
A higher federal authority is an authority at federal level in Germany that reports directly to the Federal Ministry. It has the status of a supreme
federal authority and is responsible for specific areas of responsibility. As a rule, higher federal authorities perform higher-level tasks, coordinate the work of subordinate authorities and are responsible for the development and implementation of specialist and technical supervisory matters.
Case Report Form
A Case Report Form (CRF) is a standardized document or electronic data collection form used in clinical trials to collect information about individual trial participants. It is used to record relevant data and information that is important for evaluating the safety and efficacy of a medical intervention or drug.
Clinical Research Associate
A Clinical Research Associate (CRA) is a professional who works in clinical research. The CRA plays an important role in the conduct of clinical trials and is responsible for the supervision and coordination of trial activities. The primary role of a Clinical Research Associate is to ensure that a clinical trial is conducted in accordance with applicable regulations, study protocols and ethical principles. This includes monitoring study sites to ensure that study protocols are followed, data is recorded correctly and patient rights are upheld.
Consort Statement
The CONSORT Statement (Consolidated Standards of Reporting Trials) is a guideline for the transparent reporting of randomized controlled trials (RCTs). It was developed to support researchers in the publication of study reports and to improve the quality and
comparability of study results.
CRO
A CRO (Contract Research Organization) is a company or organization contracted by pharmaceutical companies, biotechnology companies or
medical device manufacturers to perform clinical research services. CROs offer a wide range of
services
Declaration of Helsinki
The Declaration of Helsinki is an ethical guideline for medical research involving human subjects. It was first adopted by the World Medical Association (WMA) in 1964 and has been updated several times since then. The Declaration of Helsinki contains
fundamental principles and guidelines that physicians and researchers must observe when conducting clinical trials and experiments with human
subjects in order to protect their safety, dignity and rights.
Dichotomy
In medical statistics, dichotomy refers to the categorization of data into only two possible forms. A well-known example is the distinction between sick and healthy or between a positive and a negative diagnosis. Dichotomous variables are often used to describe outcomes in clinical trials. However, the analysis of such variables requires special statistical procedures in order to ensure a
meaningful interpretation of the results.
Difference-in-differences method
The difference-in-differences (DiD) method is a statistical method often used in medical statistics to evaluate the impact of interventions or treatments on a particular disease or condition. It compares the difference in changes in outcomes between a treated and untreated group over a period of time. The DiD method can help to minimize bias in the evaluation of treatment effects and enable a more precise evaluation of interventions in medical practice.
DiGA
A DiGA stands for “digital health application”. It is a medical app or software that is used for the diagnosis, treatment,
monitoring or prevention of diseases. DiGAs are digital applications that can be used on smartphones, tablets, computers or other
electronic devices.
DIMDI
DIMDI stands for “German Institute of Medical Documentation and Information”. It is a higher federal authority in Germany that is responsible for the provision of medical information and data. DIMDI is subordinate to the Federal Ministry of Health and is based in Cologne.
Ethics Committee
The ethics committee is an independent institution that plays a central role in medical research and clinical practice. Its task is to examine and evaluate the ethical justifiability of research projects, in particular clinical studies. In doing so, it monitors the protection of the rights, welfare and safety of the subjects or patients involved.
EudraCT number
The EudraCT number (European Union Drug Regulating Authorities Clinical Trials) is a unique identification number for clinical trials in the European Union (EU). It is used to register and track studies, particularly in connection with drug trials.
FDA
The FDA stands for Food and Drug Administration. The FDA is a federal agency in the United States that is responsible for regulating and monitoring food, drugs, medical devices, biological products and tobacco products.
Gold standard (medicine)
The gold standard in medical statistics refers to the best available test or diagnostic method that serves as a reference for the evaluation of other tests or methods. The gold standard is often used to assess the validity and reliability of new or experimental tests. In medical practice, the gold standard often refers to a diagnostic method or treatment that is generally considered to be the most effective or appropriate. The use of gold standards is an important aspect of evidence-based medicine and data-based decision making.
Good Clinical Practice
Good Clinical Practice (GCP) is an internationally recognized standard for the planning, conduct, documentation and reporting of clinical
trials. GCP ensures that studies are conducted ethically and scientifically correct and that the rights, safety and well-being of
study participants are protected.
Interquartile range
The interquartile range is a measure used in medical statistics to describe the dispersion or variability of a group of data. It is calculated by subtracting the value of the 75th percentile from the value of the 25th percentile. The interquartile range indicates how far apart the middle 50% of the data are and is less prone to outliers than the standard deviation. The interquartile range is therefore a useful method for quantifying the variability of data in a study.
Intervention study
An intervention study is a research design in medical statistics in which the effects of a specific intervention on a target group are investigated. It is a type of clinical trial in which one or more groups of participants are exposed to a specific intervention, while a control group receives no intervention. Intervention studies are conducted to evaluate the efficacy, safety and efficiency of drugs, therapies or other medical interventions. Various statistical methods such as randomization, blinding and placebo controls are used to ensure the validity and accuracy of the results.
Investigator Initiated Study IIT
An Investigator Initiated Study (IIT) is a clinical trial that is initiated and conducted by an independent researcher or research group
. In contrast to sponsored clinical trials, in which a pharmaceutical company or other external organization finances and manages the study
, the initiative and responsibility for IITs lies with the researchers themselves.
Karnofsky Index
The Karnofsky Index is a measure of patients’ overall functional ability that is used by healthcare providers to assess disease progression and plan treatments. The index assesses the patient’s ability to perform daily activities such as eating, personal hygiene and exercise and assigns a score from 0 to 100 points, with higher scores indicating better functioning. The
Karnofsky Index is often used in oncology to assess the progression of cancer and the impact of treatments on patients’
quality of life.
Clinical relevance
In medical statistics, clinical relevance refers to whether the results of a study are useful in clinical practice. A statistically significant result is not automatically clinically relevant if, for example, it only has a very small effect. Clinical relevance is often determined on the basis of threshold values that indicate the extent to which an effect or change makes an intervention useful or necessary in practice. The assessment of clinical relevance is an important aspect of evidence-based medicine.
Cohort study
Cohort studies are used in medical statistics to investigate the relationship between exposure (e.g. smoking) and the occurrence of a disease. Here, people who are exposed are compared with people who are not exposed. Data is collected and analyzed over a longer period of time in order to determine the influence of exposure on the occurrence of the disease. Cohort studies are considered one of the strongest study designs in medical research.
McNemar test
The McNemar test is a statistical procedure in medical statistics that is used to analyze binary data. It is used in particular to determine differences between two correlated samples. The test compares the number of matches and non-matches in one sample with another sample and determines whether the differences are significant. The McNemar test is particularly useful in examining changes or improvements in treatments when only a limited number of patients are available.
MedDEV
MedDEV stands for “Medical Device Development” and refers to a series of directives and guidelines developed by the European Commission
to support companies in the development and approval of medical devices. These guidelines are intended to ensure
that medical devices comply with the applicable quality standards and that the health and safety of patients is guaranteed.
MedDRA
MedDRA stands for “Medical Dictionary for Regulatory Activities”. It is an internationally recognized medical vocabulary used to classify, code and report adverse events (side effects) of drugs and other
medical products.
Medical Devices Act
The Medical Devices Act (MPG) is a German law that regulates the legal requirements for medical devices. It is the national implementation
of the European Medical Device Directives and, since May 26, 2021, the European Medical Device Regulation (MDR).
The risk-benefit assessment is a complex process that takes various factors into account in order to evaluate the relationship between the expected benefits of a medical intervention and the associated risks. Statistical methods are used to collect and analyze objective data and evidence.
Placebo test
A placebo test is a statistical test used in medical statistics to evaluate the effect of a treatment compared to a placebo. It involves treating one group of patients with a placebo while another group receives the actual treatment. The difference in results between the two groups can help to evaluate the effect of the treatment on the disease or condition being studied. Placebo tests are an important tool for evaluating the effectiveness of drugs and therapies and contribute to evidence-based medicine.
PMCF
PMCF stands for “Post-Market Clinical Follow-up” and is part of the process of monitoring the safety and performance of medical devices in real clinical use after their market launch. It is used to collect additional information about the product that may not have been collected during the clinical trial prior to approval.
Positive predictive value
The positive predictive value (PPV) is a statistical measure that indicates the probability that a positive test result actually indicates the presence of a certain disease or condition. The PPW depends both on the prevalence (frequency) of the disease in the population being tested and on the sensitivity and specificity of the test. A high PPW means that the test result is probably correct and that there is a high probability of the disease or condition being present. The PPW is an important measure for clinical decision making and the interpretation of diagnostic tests.
Primary endpoint – secondary endpoint
A primary endpoint (also known as a primary outcome) is a pre-specified measurable result or event that is considered the main objective in a clinical trial
or experiment. The primary endpoint is often defined to evaluate the effectiveness or success of a medical
treatment, drug or intervention.
Subject study
A subject study refers to a study in which humans act as participants or subjects. In such studies, human subjects are used to investigate various aspects of medical research, clinical trials or scientific investigations.
Qualitative characteristics in the medical field
Qualitative characteristics in the medical field are characteristics that cannot be measured numerically, such as gender, ethnicity or disease diagnosis. In medical statistics, these characteristics are referred to as categorical variables and can be analyzed using frequency distributions and cross-tabulations. Analyzing qualitative characteristics can help to identify differences and correlations between different groups of patients or between patients and treatment outcomes. The consideration of qualitative characteristics is an important aspect of data-based decision making in medical practice.
Randomization
Randomization refers to the process of randomly assigning participants in a study or experiment to different treatment groups or intervention groups. It is an important method to minimize bias or systematic differences between groups and to ensure the validity of study results.
Residuals
In medical statistics, residuals are the deviations between the observed and predicted values of a dependent variable. They are important for checking the quality of a regression analysis. If they are randomly distributed and show no systematic patterns, this indicates a good fit of the model. Otherwise, there may be fitting problems, e.g. a non-linear relationship between the variables or missing variables that should be included in the model.
Retrospective studies
Retrospective studies are a type of observational study in medical statistics in which data from the past is analyzed. These studies can be conducted through the use of medical records or questionnaires. Retrospective studies can help identify cause-and-effect relationships between different factors and diseases and evaluate risk factors. However, retrospective studies can be prone to bias and error, and the causality of relationships between variables may be difficult to establish.
Sensitivity analysis
Sensitivity analysis is a statistical method used in medical statistics to test the robustness of the results of an analysis against changes in assumptions or parameters. By changing the parameters, potential effects on the results can be investigated. This can help to increase the strength of the evidence by identifying and addressing potential weaknesses in the analysis. Sensitivity analysis is an important part of evidence-based medicine and helps to improve the accuracy and reliability of medical research results.
Serious Adverse Event
A Serious Adverse Event (SAE) is a serious adverse event that occurs during participation in a clinical trial, medical treatment or the use of a medicinal product. An SAE is defined as an event that results in patient death, is life-threatening, requires
inpatient hospitalization or prolongation of existing hospitalization, results in permanent or significant impairment of body functions, or causes a congenital anomaly or birth defect.
Standard Operating Procedure
A standard operating procedure (SOP) is a written guide or procedure that sets out specific steps and instructions for carrying out
a particular task or activity in an organization. SOPs are used to ensure uniform standards, consistency and quality assurance in
different areas of work.
Partial benefit value
In the field of medical statistics, the partial benefit value is used to assess the benefit of a diagnostic test. This can be done by assessing the sensitivity and specificity of the test as well as the prevalence of the disease in the population under investigation. The partial benefit value can help to determine the benefit of the test for different groups of patients and support informed decision-making in medical practice.
Trial Master File TMF
A Trial Master File (TMF) is a central component of a clinical trial or a clinical trial program. It is a comprehensive and
organized collection of documents that document all aspects of the planning, conduct and monitoring of the trial. The TMF serves as proof of compliance with Good Clinical Practice (GCP) and other relevant regulations and guidelines.
Visual Analog Scale
The visual analog scale (VAS) is a measurement tool often used in medical statistics to assess the intensity of symptoms or the effectiveness of treatments. The VAS consists of a horizontal line with two endpoints representing the extremes of the assessed characteristic. The person then marks a point on the line that reflects their rating. The VAS is easy to use and has good sensitivity to detect differences in intensity, which is why it is often used in clinical studies.
World Health Organization
The World Health Organization (WHO) is a specialized agency of the United Nations (UN) with headquarters in Geneva, Switzerland. It was founded on 7 April
1948 and is the leading international public health organization. The WHO’s main mission is to promote health, prevent disease and improve medical care worldwide. It is committed to promoting health standards, developing evidence-based policies and coordinating measures to combat health problems.
Youden Index
The Youden index is a measure used to evaluate the performance of a diagnostic test in medical statistics. It is calculated by the difference between the sensitivity and the complementary specificity of the test and can take values from -1 to 1. A high value of the Youden Index indicates that the test has a high diagnostic accuracy, while a low value indicates a lower accuracy. The Youden index can help to evaluate the effectiveness of diagnostic tests and determine the best threshold for interpreting test results.