Abhishek Singh’s Post

A reminder that effect sizes, confidence intervals, and other sources of information (when possible) should ALWAYS be obtained by default, regardless of what a p-value has to say about the matter. A discussion on the matter can be found below: https://lnkd.in/gYkMzwv9 #stats #statistics #statisticalthinking #research #researchmethods

Unlocking the Mystery of Standard Deviations: What is Considered Normal? Learn how 2 standard deviations can define what is considered within the normal range for 95% of the population. Discover how this calculation can determine norms and gain a deeper understanding of statistical significance. #StandardDeviationsExplained #WhatIsNormal #StatisticalSignificance #DataAnalysis101 #Mathematics #StatisticsExplained #NormalDistribution #UnderstandingStandardDeviations #MathNerd #DataScience Full discussion - https://lnkd.in/e5EzZae9

Orange friends! Not long ago, many of you took part in our survival analysis challenge and were incredibly successful! 🙌 We are now presenting the first video from our Survival Analysis playlist to the rest of our community 🎬   Survival analysis, also known as time-to-event analysis, is a crucial statistical method for analysing data on time to an event such as hospital readmission, treatment failure, death, etc. In medical research, understanding the time to critical outcomes like death or failure can drive significant advancements and inform better decision making. 🏥 ✅ 🍡 🥟 In our bite-sized tutorial Data and Survival Curve, we try to estimate, based on a dataset of 10 people and inspired by a real-life event of one of our colleagues, how long a dental filling will “survive” 🤔 We cover basic concepts of #survivalanalysis and show you how to prepare survival data and visualise survival curves in Orange.   In the video, you will get familiar with the following Orange widgets: Dataset, File, Data Table, As Survival Data, Kaplan-Meier 🍊 Watch the tutorial here 👉 https://lnkd.in/drMvSmKN We hope you like it, and look forward to your questions or comments 😊 #dataanalytics #machinelearning #datascience #analysis #education

🌟 Happy to present my latest data science project for Main Flow Services and Technologies . 🌟 Analyzing data from a heart disease dataset, I outlined crucial insights and predictive trends. I developed a state-of-the-art model to aptly predict the risk of heart diseases using advanced machine learning techniques. Probably it will help in early diagnosis and prevention. I am so proud of the work I have done and the impact it is creating in this important world of health care. #DataScience #MachineLearning #Bigdata #HeartDisease #MainflowServices #Mainflow

There are three main ways to show patient measurements between groups over time. In this post, I will explain the three options, with advantages and disadvantages. The examples are shown in the enclosed plot. 📊Option A Widely used are boxplots per group and timepoint. The adventage is that crude single data points are partly covered by the boxplot. The disadvantage is that individual patient effects are not visible. 📉Option B Individual patient courses are sometimes called spaghetti plots. It clearly shows the individual course from baseline to end for every patient and patient group, which provides the full picture. Unfavorable, sigle outliers could skew the whole scale so that the effect is hardly visible. 📈Option C Plot of mean or median with standard deviation, standard error, or interquartile range. This plot could give a glimpse of the significance if the standard deviations of two groups do not cross. But if the standard deviations or even standard errors of groups do cross, it shows a weak between-group effect. 💻 All three plots are crated with R. If you would like me to send you the template so that you can create the plots yourself, comment with "template" below this post. 🌎 If you want to become a professional in medical statistics, data science, and clinical research, leave me a message! #datascience #medicine #statistics #clinicalresearch #R  

Have you ever wondered how researchers make conclusions about large populations without surveying everyone? Inferential statistics allows us to draw conclusions about a larger group (called a population) based on a smaller subset (called a sample). Since collecting data from an entire population can be impractical or impossible, we rely on samples to gain valuable insights. By using techniques like; 1. hypothesis testing, 2. confidence intervals, 3. regression analysis, etc. Inferential statistics helps us make informed predictions and decisions across various fields—from healthcare to marketing. How have you used inferential statistics in your work or studies? Share your experiences below!

⏳ What is Survival Analysis? ⏳ 🤔How long does it take to get a job after graduation? Or for a patient to recover from a disease? Questions like these are answered with #SurvivalAnalysis, a statistical approach that makes sense of time-to-event data, even when some pieces are missing! The trickiest part? #Censored data—when we don’t have the full story for every subject. Here’s a quick look at the types: 🚦 Types of Censoring: 📍 Left Censoring: What happened before a certain point is unclear.   Example: Students who join a class with prior knowledge. 📍 Right Censoring: We lose track of what happens after a certain point.   Example: A participant drops out of a study or can’t be followed up. 📍 Interval Censoring: An event occurs between two moments, but the exact time is unknown.   Example: A disease is detected between routine health check-ups. That’s just the beginning! In my next post, I’ll explore how we uncover insights from these challenges and make predictions from incomplete data.🚀 💬 Have you dealt with censored data before? Share your #experience—I’d love to hear about it! #SurvivalAnalysis #CensoredData #DataScience #Biostatistics #Analytics #PHARMASTATS #rstat

👇#Post25 𝐊𝐞𝐲 𝐎𝐛𝐬𝐞𝐫𝐯𝐚𝐭𝐢𝐨𝐧𝐬 𝐨𝐧 𝐇𝐲𝐩𝐨𝐭𝐡𝐞𝐬𝐢𝐬 𝐓𝐞𝐬𝐭𝐢𝐧𝐠 If we reject the null hypothesis even when it is true, it is considered a Type 1 error, also known as a False Positive. 𝐄𝐱𝐚𝐦𝐩𝐥𝐞 𝐨𝐟 𝐓𝐲𝐩𝐞 𝟏 𝐄𝐫𝐫𝐨𝐫: Diagnosing a disease in a person who is not actually suffering from it. If we accept the null hypothesis even when it is false, it is considered a Type 2 error, also known as a False Negative. 𝐄𝐱𝐚𝐦𝐩𝐥𝐞 𝐨𝐟 𝐓𝐲𝐩𝐞 𝟐 𝐄𝐫𝐫𝐨𝐫: Failing to diagnose a disease in a person who is actually suffering from it. 𝐄𝐟𝐟𝐞𝐜𝐭 𝐨𝐟 𝐒𝐢𝐠𝐧𝐢𝐟𝐢𝐜𝐚𝐧𝐜𝐞 𝐕𝐚𝐥𝐮𝐞 𝐨𝐧 𝐓𝐲𝐩𝐞 𝟏 𝐚𝐧𝐝 𝐓𝐲𝐩𝐞 𝟐 𝐄𝐫𝐫𝐨𝐫𝐬: 👉 When the significance value(alpha) is increased, such as setting it at 20% or 30%, the rejection region expands. This increases the likelihood of rejecting the null hypothesis, leading to a higher chance of Type 1 error. 👉 Conversely, decreasing the significance level to very low values like 1% or 0.5% reduces the rejection region. This lowers the chances of rejecting the null hypothesis, increasing the likelihood of Type 2 error in the sample. 👉 To mitigate Type 1 and Type 2 errors, the significance value should be chosen based on the specific use case. If even small anomalies in the sample are significant, a lower significance value should be selected. Feel free to drop any advice or resources in the comment below #learnings #datascience #statistics #InferentialStatistics #Hypothesistesting

Unlock the power of data without assumptions! Explore various non-parametric tests like the Mann-Whitney U, Kruskal-Wallis, Wilcoxon Signed-Rank, and more for your research needs. 📊🔍 #Statistics #Research #ConcernNepal #NonParametricTests #DataAnalysis #ResearchMethods #knowstatistics

Hypothesis testing: One-sample test for means. A one-sample test for means is a statistical procedure used to determine if the mean of a single sample is significantly different from a known or hypothesized population mean. There are two common types of one-sample tests for means: the one-sample t-test and the one-sample z-test. 1. One-Sample t-Test This test is used when the population standard deviation is unknown and the sample size is relatively small (typically n < 30). 2. One-Sample z-Test This test is used when the population standard deviation is known or the sample size is large (typically n ≥ 30). #statistics #datascience #dataanalyst #datascientist