Biostatistics Platform Resources
This is intended as a “living document”, with individual resources being added or modified from time to time – please revisit frequently.
Training resources provided by University of Melbourne and MACH partners
| Methods and Implementation Support for Clinical and Health (MISCH) research hub | MISCH provides a range of training courses to researchers in clinical and health sciences. These courses are designed to equip researchers with the knowledge and skills they need to design and conduct high-quality research studies and implement evidence-based practices in clinical and health settings. From beginners and advanced REDcap workshops to seminars on a variety of topics relating to biostatistics, grants, health economics and co-design: https://clinicalresearch.mdhs.unimelb.edu.au/ |
| Melbourne Academic Centre for Health (MACH) | MACH is a collaborative partnership of clinical and research partners. Those of particular relevance to Department of O & G include The Royal Women's Hospital, Women's, Mercy Hospital for Women, Joan Kirner Hospital, Western Health, the Northern Hospital. MACH's training courses cover a range of topics related to health research, including research design, statistical analysis, data management, and research translation. The courses are designed to provide practical and hands-on training that enables researchers to develop the skills and knowledge they need to conduct high-quality research: https://machaustralia.org/ |
| Statistical consulting centre (SCC) | SCC at the University of Melbourne offers training and seminars on various statistical software packages, including R, Stata and SAS. These seminars cover topics such as data management, data analysis and visualization, and are designed to help researchers and students become proficient in using these software packages: https://scc.ms.unimelb.edu.au/#courses |
| The Clinical Epidemiology & Biostatistics Unit (CEBU) | CEBU provides research methods leadership and support on the Melbourne Children’s campus, jointly supported by the Murdoch Children’s Research Institute (MCRI) and University of Melbourne Department of Paediatrics. CEBU delivers a comprehensive range of short courses in the methods of clinical and population health research, including conceptualisation, design and statistical analysis, as well as training in statistical software: https://www.mcri.edu.au/research/research-areas/population-health/clinical-epidemiology-biostatistics-cebu |
Study design and analysis resources provided by the leading journals
| Study analysis and reporting guidelines (some of these guidelines, e.g. EQUATOR, New England Journal of Medicine or JAMA, are applicable to all clinical areas, while other guidelines, despite coming from clinical areas different to obstetrics and gynaecology, still provide useful and important information to consider) | |
|---|---|
| Enhancing the Quality and Transparency of Health Research (EQUATOR) Network guidelines | EQUATOR Network provides a range of resources and tools to support researchers, editors, peer reviewers such as reporting guidelines across various study types (e.g. randomised trials, observational studies, systematic reviews, study protocols, etc.), database of reporting guidelines, online resources including training and educational materials, webinars, and podcasts: https://www.equator-network.org/ |
| New England Journal of Medicine | https://www.nejm.org/author-center/new-manuscripts |
| Journal of Clinical Oncology | Statistical guidelines: https://ascopubs.org/jco/authors/manuscript-guidelines |
| Annals of Internal Medicine | Information for authors – general statistical guidance: https://annals.org/aim/pages/author-information-statistics-only |
The Journal of the American Medical Association (JAMA): | - Statistical methods and data presentation: https://jamanetwork.com/journals/jama/pages/instructions-for-authors#SecStatisticalMethodsandDataPresentation - The Journal also publishes JAMA Guide to Statistics and Methods - essay series that explains the basics of statistical techniques used in clinical research, to help clinicians interpret and critically appraise the medical literature: https://jamanetwork.com/collections/44042/jama-guide-to-statistics-and-methods |
| Study design and analysis, notes | |
|---|---|
| British Medical Journal (BMJ) statistics notes | BMJ publishes original research, clinical reviews, practice guidelines, and educational articles across a wide range of medical specialties. BMJ Statistics Notes covers a wide range of topics, including hypothesis testing, sample size calculation, regression analysis, survival analysis, and meta-analysis. Each article is designed to provide a clear explanation of the statistical method or technique being discussed, with examples and practical advice for its application in medical research: https://www.bmj.com/specialties/statistics-notes |
| Journal of Physiotherapy | The Journal of Physiotherapy is a peer-reviewed medical journal that covers research related to physiotherapy, which is a branch of healthcare that involves the treatment of physical conditions through exercise, manual therapy, and other modalities. Since 2010, the Journal of Physiotherapy published a collection of Research Notes which cover a broad range of topics including meta-analysis, causal inference, single-case experimental designs, prognostic model research, and others: https://www.sciencedirect.com/journal/journal-of-physiotherapy/special-issue/10J93TWVB4T |
| University of Oxford, Centre for Evidence-Based Medicine | https://www.cebm.ox.ac.uk/resources/ebm-tools/study-designs |
| OBGYN specific resources | |
|---|---|
| The Cochrane Pregnancy and Childbirth Group | The Cochrane Collaboration is an international organization that produces systematic reviews of healthcare interventions. The Pregnancy and Childbirth Group provides reviews of interventions related to pregnancy and childbirth, including study design and analysis principles. Statistical analysis checklist outlined in the Cochrane Protocol: https://pregnancy.cochrane.org/sites/pregnancy.cochrane.org/files/public/uploads/MethodologicalChecklistProtocols.pdf |
| American Journal of Obstetrics and Gynecology (AJOG) | This journal publishes original research, reviews, case reports, and commentary on a broad range of topics within obstetrics and gynaecology. Its Guide for Authors includes the outline of Trial and Research Guidelines to support study design and reporting, i.e.: randomized control trial, data sharing, systematic review or meta-analysis, systematic review or meta-analysis of randomized controlled trials, meta-analysis or systematic review of observational studies, other systematic review reporting guidelines, observational study in epidemiology, health economics, and additional guidelines/standards: https://www.elsevier.com/journals/american-journal-of-obstetrics-and-gynecology/0002-9378/guide-for-authors |
| An International Journal of Obstetrics and Gynaecology | This journal publishes original research, systematic reviews, and meta-analyses on topics related to obstetrics, gynaecology, and reproductive health. The journal provides Research Method Guidelines series that have been produced by statistical experts and clinicians to explain how to conduct and understand research, using examples from obstetrics and gynaecology: https://obgyn.onlinelibrary.wiley.com/hub/journal/14710528/collections/research_methods_guides |
| Obstetrics and Gynecology | The official publication of the American College of Obstetricians and Gynecologists (ACOG) and features original research, reviews, and case reports related to women's health. In 2022, the ACOG published a 10-page research note titled “Study Design and Statistics for the Nonstatistician": https://journals.lww.com/greenjournal/Documents/Research_methods.pdf |
Guide for selecting appropriate statistical analysis
The University of California, Los Angeles (UCLA) provides a comprehensive online resource called "Choosing the Correct Statistical Test" that assists researchers in selecting the appropriate statistical test for their research design. The resource includes a table that helps researchers identify the correct statistical test for their research question based on the variables they have collected and the type of analysis they want to perform. The table includes options for STATA, SPSS, R and SAS software: https://stats.oarc.ucla.edu/other/mult-pkg/whatstat/
Statistical software
The following resources offer instruction on the application of diverse statistical tests through widely used statistical software tools, which are provided by the IT department of the University of Melbourne
| Stata | Stata is a general-purpose statistical software package used for data analysis, data management, and graphics. STATA can handle both small and large datasets, and its data management features allow users to clean and manipulate data in a variety of ways. STATA's statistical analysis tools include regression analysis, generalized linear models, survival analysis, time series analysis, and panel data analysis. |
- Stata video tutorials: https://www.youtube.com/user/StataCorp - Various statistical tests explained (with example): https://stats.oarc.ucla.edu/stata/whatstat/what-statistical-analysis-should-i-usestatistical-analyses-using-stata/ | |
| R | R is an open-source programming language and software environment for statistical computing and graphics. R provides a wide range of statistical and graphical techniques, including linear and nonlinear modelling, time-series analysis, and machine learning algorithms. It also has a powerful suite of data manipulation and visualization tools, allowing users to explore, clean, and visualize data in a variety of ways. R has a rich library of packages available for download, which add functionality to the core software, including tools for spatial data analysis, text mining, and network analysis, among others. |
- Introductory statistics with R by Peter Dalgaard (2008): https://link.springer.com/book/10.1007/978-0-387-79054-1 - A Handbook of Statistical Analysis Using R, Third Edition by Torsten Hothorn; Brian S. Everitt (2014): http://cat.lib.unimelb.edu.au/record=b5663148~S30 | |
| SPSS | SPSS provides a wide range of statistical techniques, including descriptive statistics, inferential statistics, factor analysis, cluster analysis, and regression analysis. It also has powerful data management tools, allowing users to manipulate and transform data in a variety of ways. |
- IBM SPSS statistics documentation: https://www.ibm.com/support/pages/ibm-spss-statistics-28-documentation#en - Discovering Statistics Using IBM SPSS Statistics by Andy Field (2018): http://cat.lib.unimelb.edu.au/record=b7765963~S30 | |
| GraphPad Prism | GraphPad Prism is a statistical software package used for scientific graphing, data analysis, and presentation. GraphPad Prism provides a wide range of statistical and graphical techniques, including descriptive statistics, inferential statistics, regression analysis, ANOVA, and survival analysis. It also has powerful data management tools, allowing users to manipulate and transform data in a variety of ways. |
- GraphPad user guide on statistical tests and graphs: https://www.graphpad.com/guides/prism/latest/user-guide/index.htm |
Data structure and data handling
Data for statistical analysis expected to be Excel or CSV format and organized in a structured format that allows for easy manipulation and analysis. For most statistical analyses, data is organized into rows and columns, with each row representing an observation and each column representing a variable. The first row usually contains variable names, and the first column usually contains observation labels or IDs.
The data itself should be in a numeric or categorical format that can be interpreted by the statistical software. Numeric data should be entered as numbers, and categorical data should be entered as codes or categories. Missing data should be clearly coded with a specific value out of the range of possible values for that variable (e.g. 9999 if such a value cannot be encountered within the data range).
Depending on the statistical analysis, the data needs to be either in wide or long format. In wide format data, each variable is represented in its own column, with each observation in a separate row (for example, each row corresponds to an individual patient with values for three separate time points the patient was observed on, are placed in three respective columns). Long format data are organised by sub-observation, storing data in multiple rows (for example, each row containing information about individual time points the patient was observed on, resulting in patient information distributed across multiple rows).
| Useful resources for data handling with R, Stata and SPSS | |
|---|---|
| Data standardization | This involves scaling the data to a common scale, such as z-scores or min-max scaling. |
| Data aggregation | This involves summarizing data by collapsing multiple observations into a single value. For example, calculating the mean or median of a group of observations. |
Stata: https://kb.iu.edu/d/aqdz | |
| Data merging | This involves combining multiple datasets into a single dataset for analysis. This may involve merging datasets based on a common identifier or key variable |
| Data subsetting | This involves selecting a subset of the data for analysis based on specific criteria, such as a particular time period or a subset of variables. |