Statistical Consultant
Expert Consultation for Your Statistical Studies and Data
I’m Justin Belair – the expert you’ve been looking for on data science and statistics. My statistical consulting and freelance work has seen my collaborator’s methods and papers being published in top peer reviewed journals.
Save Yourself Scientific Investigation Stress
Simplified data handling
Streamline the process of cleaning and managing data
Boost data quality
Enhance the reliability and validity of your results
Tailored study design
Create a research plan that aligns with your questions
Statistical Tool Selection
Choose the best statistical tools to meet your goals
Statistical Modeling
Derive meaningful inferences from your data
Publication-Quality Results
Figures ready to copy-paste for submission to journals
Methodology Documentation
Receive professionally written methodology sections
Results Interpretation
Obtain clear explanations of statistical results Bibliographic
Bibliographic References
Document your work for reliability and reproducibility
The Online Statistics Expert Who’s Here to Help
Statistical Consulting
It’s no secret that simple manipulation of large amounts of scientific data can be complex and confusing at the best of times. Scientific investigations use traditional approaches when it comes to statistical estimation and inference. But recently, machine learning and big data techniques are becoming more prevalent in making these estimations. Take for example bioinformatics which references and comprises both biology and computer science fields of thought.
Traditional Approaches
- ANOVA
- T-Tests
- Bayesian Models
- Bootstrap
- Simulations
- Regression
- Models
- …and more
Machine Learning
- Dimensionality Reduction
- (PCA, UMAP, PHATE, etc.)
- Unsupervised Clustering
- (kNN, k-means, etc.)
- Data as Strings of Letters
- (Nucleotides ACGT in Genomic Data, etc.)
Fortunately, thanks to my deep understanding of these methods – combined with my R Studio and Python programming expertise – I can help you present your findings with clarity and conciseness for both traditional and machine learning based approaches.
Biostatistics & Bioinformatics
These can include:
- Clinical Research Planning
- Correct Number of Patients to Study
- Therapy Design
- Genome-Wide
- Association Studies (GWAS)
- Impacts of Epidemics on Public Health
- Vaccination response models across time and segments of population
- …and more
Common failures of batteries of non-parametric tests include:
- Test Multiplicity
- Require corrections such as Holm-Bonferroni, Benjamini Hochberg false discovery rate, etc.
- Insufficient Power
- Compounded by common and relatively small datasets.
Previous work example
- Using linear mixed-effects models to assess robustness of COVID-19 vaccine immune responses in pre-infected patients. This study showed that delayed booster shots still conferred robust immune responses (B and T cells) (DOI :10.1016/j.celrep.2022.111013)
- A third SAS-CoV-2 mRNA vaccine dose in people receiving hemodialysis overcomes B cell defects but elicits a skewed CD4+ T cell profile (DOI: 10.1101/2022.09.05.506622)
- For more : https://www.researchgate.net/profile/Justin-Belair
RStudio
In the world of statistics, RStudio’s R programming language is the standard tool for analyzing statistical data.
R is specifically designed to deal with statistical applications and data analysis when compared to other open-source programming languages like Python.
R programming is best suited to handle complex statistical analysis, as the code serves as a robust documentation of how the analysis proceeded. This allows for more overall clarity, as it provides an efficient data-integrity check.
Not only am I able to offer RStudio expertise (in both English and French), but I can also educate you on the program to prevent the need to hire an external expert every time.
"He has a deep understanding of statistics, yet is able to skillfully explain complex concepts to people with little-to-no background in statistics both in French and English."
Elsa Brunet-Ratnasginham
Post-doc, UCSF
"I had never before worked with a biostatistician who speaks the same language as me! [...] Personally, our collaboration has allowed me to reconcile with statistics."
Manon Nayrac
Post-doc, CRCHUM
"I highly recommend Justin and his team for your data processing needs in order to have a more robust statistical model than the tests we are accustomed to conducting as biologists!"
Gérémy Sannier
Ph.D candidate, CRCHUM
"Thanks to him, statistics appeared to me as solutions rather than obstacles."
Mathieu Dubé
Research Associate, CRCHUM
Why Work With Me?
Experience
I’ve worked with dozens of academic and industry research professionals, leading to impactful work and many scientific publications in prestigious peer-reviewed journals.
Great Communicator
Not only do I understand statistics deeply, but I’ve been honing my communication and presentation skills for over a decade as a teacher and lecturer at 2 universities. My student evaluations are always very positive with many students highlighting my openness and the comfort they feel when learning with me.
Extensive Academic Background
I have an extensive background in Mathematics and Statistics, where I’ve excelled in all of my degrees.
Thought-Leader
I’m a thought leader in the space of applied statistics, highlighted by my dedicated following on Linkedin and my web platform biostatistics[.]ca which unites a community of budding biostatisticians. Upon request, I can share our work with my broad audience of science enthusiasts.