My research consultant services cover all aspects of the research cycle, ranging from research conceptualisation and design, to data collection, data analysis, interpretation of results, and report writing. My clientele in this regard are public and private sector organisations, as well as not-for-profit entities.
When it comes to working with post graduate students, my services are generally limited to the data analysis aspect of research projects and the bulk of this type of consulting has been in the field of biostatistics. Biostatistics is a branch of statistics that applies statistical methods to biological, health, and medical research. A biostatistician uses statistical techniques to identify and quantify patterns in biological or medical data, assess the effectiveness of medical treatments and health interventions, and identify risk factors for diseases, among other things.
In the field of biostatistics, I assist about three to four Master of Medicine (MMed) students every month. These have mostly been doctors studying at either the University of the Witwatersrand or the University of Pretoria Health Sciences Faculties.
Medical Disciplines
I assist with biostatistical analysis across all specialisations, and have worked with data from the following disciplines:
· Anaesthesia
· Diagnostic Radiology
· Emergency Medicine
· Family Medicine
· Genetics
· Internal Medicine
· Mammography
· Medical Microbiology
· Neurology
· Obstetrics & Gynaecology
· Ophthalmology
· Optometry
· Orthopaedic Surgery
· Otorhinolaryngology
· Paediatrics
· Pathology
· Psychiatry
· Urology
Methods
Most of the data that I have worked with has been collected using retrospective designs, but occassionally from prospective crossectional designs too. In relation to the statistical analysis, over and above basic descriptive statistics that profile patients’ demographic and clinical characteristics, the most applied statistical methods have been the following:
Hypotheses testing
· Independent samples t-test (or Mann Whitney U test for non-parametric data)
· Paired samples t-test (or Wilcoxon sign rank test for non-parametric data)
· One way ANOVA (or Kruskal-Wallis test for non parametric data)
· Pearson’s chi-square test
· Fisher’s exact test
· Binomial test
Correlational Analysis
· Pearson’s correlation
· Spearman’s correlation
· Point-Biserial correlation
Regression Analysis
· Linear regression (Simple & Multiple)
· Logistic regression (Binary & Multinomial)
· Reciever Operating Characteristics (ROC) curve analysis
Survival Analysis
· Cox regression modelling
· Kaplan-Meier curve analysis
Diagnostic Reliability
· Accuracy metrics (Sensitivity, Specificity, NPV, PPV)
· Cohen’s kappa
· Fleiss kappa
· Intraclass correlation coefficient
· Bland-Altman plots
Altogether, the biostatistics arm of my services continues to grow, and I attribute this to the fact that most clients go on to recommend my services to others, for which I am truly grateful.
Comments