Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
Ever wondered how many participants you need to prove your point? GPower to the rescue! Just crunched some numbers and realized the smaller the effect size, the more people you need. I met a medical doctor discussing about his research. This newsletter is the result of the discussion. In this newsletter,I shared about running some power analyses using GPower to determine the required sample sizes for a t-test comparing two independent means. The calculations and visualizations indicate the relationship between effect size and sample size needed to achieve a power of 0.8 with a 0.05 alpha level. It's fascinating to see how increasing the effect size significantly reduces the required sample size. This analysis is critical for designing efficient and effective studies. Anyone else tackled sample size calculations with GPower and lived to tell the tale?#Research #Statistics #GPower #SampleSize #DataAnalysis #StudyDesign #AcademicResearch #QuantitativeResearch #ResearchMethods #ScientificStudy
2
To view or add a comment, sign in
More Relevant Posts
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
The first time I encountered SQL was while using MS Access in my office to track all the paper-based submissions. I needed to query a database to extract specific application information, but I thought the graphical interface (GUI) would make things easier. Instead, it turned out to be more confusing than I expected.I remember trying to build a query using the drag-and-drop features in MS Access, but the GUI seemed to have a mind. It was frustrating trying to recall my memory what I have done. After a few attempts and a lot of trial and error, I decided to look under the hood—and that's when I discovered SQL.Suddenly, everything made sense. Instead of wrestling with the GUI, I could write clear, straightforward commands to get the needed data.For example, I could write:SELECT applicant, ID, productFROM applicationWHERE day > 50 AND status = 'complete';to retrieve information on applications where the review is completed and has taken more than 50 days, which unfortunately always led to our boss reprimanding us for the delay.Since then, SQL has been a constant companion in my data journey, giving me the flexibility and control I need to manage and analyze data in a way that makes sense to me.#SQL #DataScience #MSAccess #ClinicalResearch #DataManagement #FirstExperience
2
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
📈 **Discovering R: My Journey from Bioequivalence to Data Science** 🌱The first time I heard about **R**, I was sitting in the office listening to my colleague telling me how a professor used it to perform bioequivalence analysis. To be honest, it sounded like magic at first—a tool that could handle complex statistical analyses with just a few lines of code. But as I delved deeper, I realized R is not just a tool, it's like a Swiss Army knife for data scientists, designed to tackle any data-related challenge with precision.So, what exactly is R? Imagine R as the language spoken by data. It's an open-source programming language specifically built for statistical computing and graphics. Whether you're analyzing clinical trial data or exploring patterns in large datasets, R provides a rich environment to handle it all. It's a bit like a chef's kitchen—equipped with all the ingredients and tools you need.My favorite way to use R? Through **Google Colab**. It’s a powerful platform that lets me code, analyze, and share my work seamlessly online. You can share it easily through Github.#DataScience #RProgramming #Bioequivalence #PharmacistJourney #GoogleColab #ContinuousLearning
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
As a pharmacist, I've always been passionate about improving patient outcomes through accurate and evidence-based practices. But in today’s rapidly evolving healthcare landscape, I realized that staying ahead means not just understanding medications, but also the vast amounts of data that can inform better decisions.That’s why I decided to dive into the world of data science.This journey has been eye-opening. I learned how to:📈 Manipulate and analyze data to uncover trends and insights that can enhance patient care.🧠 Apply machine learning techniques to predict outcomes and optimize treatment plans.🎨 Visualize data effectively using R to communicate complex information clearly to both healthcare professionals and patients.🛠️ Clean and manage data efficiently, ensuring that the data I work with is accurate and reliable.By merging my pharmaceutical knowledge with these new data skills, I’m better equipped to contribute to the future of evidence-based healthcare. The transition hasn’t been easy, but it’s incredibly rewarding to see how data can make a real difference.To my fellow healthcare professionals: Don’t be afraid to step out of your comfort zone. The intersection of healthcare and data science is where the future lies, and it’s never too late to start learning something new. 🚀#DataScience #Pharmacist #RProgramming #MachineLearning #ContinuousLearning #HealthcareInnovation
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
Helping a colleague in the Reviewer Office to recalculate the bioequivalence (BE) 90% confidence interval (CI) using SAS code from a paper was both challenging and rewarding. The applicant was asked to provide statistical output to support the results in the synopsis, but instead, they submitted a 100-page PDF with no annotations or explanations. To make matters worse, the evaluator found discrepancies in the reported 90% CI calculations.Using the EMA fixed model in SAS, we successfully recalculated the 90% CI and presented the results in a clear table format. This process underscored the importance of accurate and transparent reporting.We are also considering developing a simple system that allows evaluators to input results from clinical study reports (CSR) into R using Google Colab. This approach would enhance shareability and include a user guide as part of a community project. This system would be particularly beneficial since we do not require applicants to submit CDISC.https://lnkd.in/g-2dAhvc
2
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
It all started when I was updating my CV after ages. I opened DataCamp to find a certificate and stumbled upon my XP—over 100,000! That was my cue to get serious about data and programming.So, while I’ve been office-bound, I turned my solitude into an opportunity. With DataCamp as my sidekick, I dove into the world of data. 23 courses, one track, and 112,000+ XP later, I’m giving myself a well-deserved pat on the shoulder! As a pharmacist with zero data background, I can confidently say this is a fantastic place to start if you’re thinking about making the leap.Plus, I’m open to any collaboration with medical doctors in Malaysia using R—let’s turn our shared knowledge into something impactful! This “isolation innovation” was totally worth it. Here’s to turning downtime into data time! 🚀
3
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
In clinical trials, unscheduled visits are any participant visits that happen outside the planned schedule. These visits can occur due to unexpected adverse events, additional medical assessments, or participant concerns.Dealing with unscheduled visits can be challenging, but with the right strategies, they can be managed efficiently. My latest slide deck covers crucial key questions to help inspectors and teams handle these visits effectively. #ClinicalTrials #UnscheduledVisits #ParticipantSafety #MedicalResearch #ClinicalResearch #TrialManagement #DataIntegrity #HealthCare #ResearchCompliance #StudyProtocol #PatientCare
6
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
Efficacy data in a clinical trial refers to the information gathered to determine whether a treatment, drug, or intervention produces the desired therapeutic effect in participants. This data is crucial for evaluating the benefit of the treatment compared to a placebo or another treatment.Efficacy data must be systematically collected and meticulously documented. This involves standardized procedures to ensure consistency, accuracy, and reliability of the data across all trial participants.Here is my checklist of essential questions for this aspect, which are detailed on the following slides.#ClinicalTrials #EfficacyData #Research #DataIntegrity #ClinicalResearch #RegulatoryApproval #PatientSafety #DataCollection #Documentation #Healthcare #MedicalResearch
4
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
I’m thrilled to share an essential guide on the process of selecting pivotal studies for inspection, crafted by myself. This guide outlines the systematic approach used to ensure a balanced and unbiased selection, which is crucial for maintaining the integrity and quality of inspections. Whether you're new to the field or a seasoned professional, I hope this tool is helpful for you. #ClinicalResearch #GCP #InspectionProcess #QualityAssurance #PharmaceuticalStudies
12
Like CommentTo view or add a comment, sign in
-
Chen Wei Oh
Bio-Statistician |Pharmacist (x-BCAPS) | Regulatory & GCP Inspector | use Python and R
- Report this post
🎯 Nailing the Perfect Sample Size for Fisher's Exact Test – No Guesswork Required! 🎯Hey data enthusiasts! Ever wondered how to get the perfect sample size for comparing two groups with Fisher's exact test? Well, I’ve got you covered in this newsletter!Why This Matters: Getting your sample size right is like hitting the bullseye 🎯. It makes sure your results are spot-on and reliable. So, next time you’re planning a study, remember: size does matter!Check out my slide for all the juicy details and make your next study a slam dunk! 🏀📈Services Offered by My Data Clinic SessionUnlock the Full Potential of Your Research with Expert GuidanceSample Size CalculationEnsure the validity and reliability of your research findings with precise sample size calculations. I assist researchers in determining the optimal number of participants required for their studies, guaranteeing statistically significant results and minimizing the risk of errors.Feasibility GuidanceMaximize your study's success with tailored feasibility advice. I provide expert guidance on the number of subjects that can realistically be enrolled during your study period. This helps you design a feasible and efficient study, saving time and resources.Research Potential ExplorationDiscover the true potential of your research topics with comprehensive exploration services. I guide researchers in evaluating expected differences, helping to develop impactful and feasible study designs. Unlock innovative insights and achieve groundbreaking results with expert support.Transform Your Research with Professional Data Clinic SessionsDon't leave the success of your study to chance. Partner with a data expert to navigate the complexities of research design and execution. Book a session today and elevate the quality and impact of your research.#ResearchMethodology #StatisticalAnalysis #FisherExactTest #SampleSizeCalculation #DataScience #Biostatistics #StudyDesign #GPower #ClinicalResearch #DataAnalysis #StatisticalPower #ResearchTips #DataDriven #ScientificResearch #HealthResearch
2
Like CommentTo view or add a comment, sign in
1,197 followers
- 451 Posts
- 69 Articles
View Profile
FollowMore from this author
- Case Study: Leveraging G*Power for Optimal Sample Size Calculation in Research using Two Mean Formula. Chen Wei Oh 5d
- How I use Sample Size Determination for Two Independent Groups Using Fisher's Exact Test Chen Wei Oh 1w
- The Importance of Correctly Using Standard Deviation in Sample Size Calculations Chen Wei Oh 1mo
Explore topics
- Sales
- Marketing
- Business Administration
- HR Management
- Content Management
- Engineering
- Soft Skills
- See All