Statistical Aid: A School of Statistics

World Statistics Day - 2025.

𝐒𝐞𝐧𝐬𝐢𝐭𝐢𝐯𝐢𝐭𝐲 𝐯𝐬 𝐒𝐩𝐞𝐜𝐢𝐟𝐢𝐜𝐢𝐭𝐲

✅𝐒𝐞𝐧𝐬𝐢𝐭𝐢𝐯𝐢𝐭𝐲: Sensitivity is the measure of a test's ability to correctly identify individuals who have a disease or condition. It is often called the true positive rate.

Mathematically, sensitivity is calculated as:

Sensitivity=True Positives (TP) / True Positives (TP)+False Negatives (FN)

👉For example, if 100 people have a disease and a test correctly identifies 90 of them as positive, the sensitivity is 90%. This means the test misses 10 people who actually have the disease (false negatives).

✅𝐒𝐩𝐞𝐜𝐢𝐟𝐢𝐜𝐢𝐭𝐲: Specificity measures a test's ability to correctly identify individuals who do not have the disease. It is also known as the true negative rate.

Specificity is calculated as:

Specificity=True Negatives (TN) / True Negatives (TN)+False Positives (FP)

👉For instance, if 100 healthy individuals are tested and 95 are correctly identified as negative, the specificity is 95%. This means 5 healthy people were incorrectly diagnosed as having the disease.

For more: https://www.statisticalaid.com/sensitivity-vs-specificity/

𝐖𝐡𝐚𝐭 𝐢𝐬 𝐭𝐡𝐞 𝐖𝐞𝐢𝐛𝐮𝐥𝐥 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐢𝐨𝐧?

The Weibull distribution is a continuous probability distribution used to model the time until a specified event occurs, such as the failure of a component or the time to complete a process.

𝐀𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 𝐨𝐟 𝐭𝐡𝐞 𝐖𝐞𝐢𝐛𝐮𝐥𝐥 𝐃𝐢𝐬𝐭𝐫𝐢𝐛𝐮𝐭𝐢𝐨𝐧

1. Reliability Engineering

✅Product Lifetimes: Modeling how long products or components last before failure.
✅Maintenance Scheduling: Predicting optimal maintenance intervals.
✅Warranty Analysis: Estimating the probability of failure within a warranty period.

2. Survival Analysis

✅Medical Studies: Modeling time until an event, such as death or relapse.
✅Biological Research: Analyzing lifespans of organisms.

3. Weather Forecasting

✅Wind Speed Modeling: The Weibull distribution often fits natural wind speed data, aiding in wind energy assessments.

For more: https://www.statisticalaid.com/weibull-distribution/

𝐖𝐡𝐚𝐭 𝐈𝐬 𝐚 𝐏𝐚𝐢𝐫𝐞𝐝 𝐒𝐚𝐦𝐩𝐥𝐞 𝐭-𝐓𝐞𝐬𝐭?

The paired sample t-test is designed to compare the means of two related groups to determine whether there is a statistically significant difference between these means. Paired samples occur when the observations in one sample are linked to observations in the other sample. The most common example is before-and-after measurements on the same subjects, but it can also involve matched subjects or pairs.

𝐄𝐱𝐚𝐦𝐩𝐥𝐞𝐬 𝐨𝐟 𝐏𝐚𝐢𝐫𝐞𝐝 𝐒𝐚𝐦𝐩𝐥𝐞𝐬

✅Measuring the blood pressure of patients before and after a treatment.
✅Testing the reading ability of children before and after a specialized training program.
✅Comparing the test scores of students on two different exams where the same students took both.

For more: https://www.statisticalaid.com/paired-sample-t-test/

𝐖𝐡𝐚𝐭 𝐈𝐬 𝐚𝐧 𝐈𝐧𝐝𝐞𝐩𝐞𝐧𝐝𝐞𝐧𝐭 𝐒𝐚𝐦𝐩𝐥𝐞 𝐭-𝐓𝐞𝐬𝐭?

The independent sample t-test, also called the two-sample t-test or unpaired t-test, is a statistical method used to compare the means of two unrelated or independent groups to determine if there is a statistically significant difference between them. The groups involved are distinct; the participants in one group do not overlap with those in the other group.

𝐖𝐡𝐞𝐧 𝐒𝐡𝐨𝐮𝐥𝐝 𝐘𝐨𝐮 𝐔𝐬𝐞 𝐚𝐧 𝐈𝐧𝐝𝐞𝐩𝐞𝐧𝐝𝐞𝐧𝐭 𝐒𝐚𝐦𝐩𝐥𝐞 𝐭-𝐓𝐞𝐬𝐭?

✅Comparing test scores between two classes taught by different instructors.
✅Comparing average salaries between males and females.
✅Comparing clinical outcomes between patients receiving two different treatments.
✅Any scenario where you have exactly two independent groups and a numeric outcome.

𝐘𝐨𝐮 𝐜𝐚𝐧𝐧𝐨𝐭 𝐮𝐬𝐞 𝐭𝐡𝐢𝐬 𝐭𝐞𝐬𝐭 𝐢𝐟:

❌The groups are related or paired (use a paired sample t-test instead).
❌The dependent variable is categorical.
❌There are more than two groups involved in your analysis (consider ANOVA).

For Details: https://www.statisticalaid.com/independent-sample-t-test/

𝐖𝐡𝐚𝐭 𝐈𝐬 𝐚 𝐎𝐧𝐞-𝐒𝐚𝐦𝐩𝐥𝐞 𝐭-𝐓𝐞𝐬𝐭?

A one-sample t-test is a statistical method used to determine whether the mean of a single sample differs significantly from a known or hypothesized population mean. It is especially useful when the population standard deviation is unknown (and must be estimated from the sample), a common scenario in real-world data analysis.

𝐖𝐡𝐞𝐧 𝐭𝐨 𝐔𝐬𝐞 𝐚 𝐎𝐧𝐞-𝐒𝐚𝐦𝐩𝐥𝐞 𝐭-𝐓𝐞𝐬𝐭?

✅You have one sample of data.
✅You know or want to test against a specific value (the population mean).
✅The population standard deviation is unknown.
✅The data are approximately normally distributed (or the sample size is large enough for the Central Limit Theorem to apply).

For More: https://www.statisticalaid.com/one-sample-t-test/

𝐖𝐡𝐚𝐭 𝐢𝐬 𝐂𝐨𝐡𝐞𝐧'𝐬 𝐝?

Cohen’s d is a statistical measure used to express the size of the difference between two group means relative to the variability observed in the data. It is a standardized effect size, meaning it is unit-free and allows comparison across different studies or variables measured on different scales.

Cohen’s d is commonly used in:

✅Comparing experimental and control groups in psychology and medicine.
✅Reporting alongside t-tests and ANOVA results.
✅Meta-analyses to synthesize findings across studies.

More Details: https://www.statisticalaid.com/cohens-d/

𝐖𝐡𝐲 𝐌𝐞𝐚𝐬𝐮𝐫𝐞𝐬 𝐨𝐟 𝐕𝐚𝐫𝐢𝐚𝐛𝐢𝐥𝐢𝐭𝐲 𝐈𝐦𝐩𝐨𝐫𝐭𝐚𝐧𝐭?

Measures of variability quantify the degree to which data points differ from each other and from the central value (mean or median). Understanding variability is crucial for several reasons:

✅Data Interpretation: Variability helps interpret what the average represents. A high variability means the average might not be representative of most data points.
✅Comparing Groups: When comparing two or more groups, variability indicates whether differences in means are meaningful or if there is too much overlap.
✅Risk Assessment: In finance or quality control, variability measures risk or uncertainty. Lower variability often implies more predictability.
✅Statistical Inference: Many statistical tests and models rely on assumptions about variability (e.g., homogeneity of variance).
✅Decision Making: Knowing variability helps in making informed decisions, such as setting tolerance limits or evaluating consistency.

For more details: https://www.statisticalaid.com/measures-of-variability/

Root Mean Square Error (RMSE) is a statistical measure that quantifies the average magnitude of the errors between predicted values and actual observed values in a dataset. More specifically, it measures the standard deviation of the residuals (prediction errors), which are the differences between predicted and true values.
https://www.statisticalaid.com/root-mean-square-error-rmse/

Root Mean Square Error (RMSE) Root Mean Square Error (RMSE) is a statistical measure that quantifies the average magnitude of the errors between predicted values and

𝐇𝐨𝐰 𝐭𝐨 𝐂𝐡𝐨𝐨𝐬𝐞 𝐭𝐡𝐞 𝐑𝐢𝐠𝐡𝐭 𝐒𝐭𝐚𝐭𝐢𝐬𝐭𝐢𝐜𝐚𝐥 𝐓𝐞𝐬𝐭?

The key to unlocking the power of statistical analysis lies in selecting the appropriate test for a given research question and data structure. An incorrect choice can lead to erroneous conclusions, misinterpretation of results, and ultimately, flawed research findings.

𝟏. 𝐒𝐭𝐚𝐫𝐭 𝐰𝐢𝐭𝐡 𝐲𝐨𝐮𝐫 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐨𝐫 𝐕𝐚𝐫𝐢𝐚𝐛𝐥𝐞:

✅Categorical? Proceed to determine your Outcome Variable.
✅Quantitative? Proceed to determine your Outcome Variable.

𝟐. 𝐃𝐞𝐭𝐞𝐫𝐦𝐢𝐧𝐞 𝐲𝐨𝐮𝐫 𝐎𝐮𝐭𝐜𝐨𝐦𝐞 𝐕𝐚𝐫𝐢𝐚𝐛𝐥𝐞: If the Predictor is Categorical:

✅Outcome Categorical? Choose a non-parametric test.
✅Outcome Quantitative? Do a comparison of means test. Then, ask:

𝟑. 𝐇𝐨𝐰 𝐦𝐚𝐧𝐲 𝐠𝐫𝐨𝐮𝐩𝐬 𝐚𝐫𝐞 𝐛𝐞𝐢𝐧𝐠 𝐜𝐨𝐦𝐩𝐚𝐫𝐞𝐝?

✅Two? Use a T-test.
✅More than two? Ask: How many outcome variables?
✅One? Use ANOVA.
✅More than one? Use MANOVA.

𝟒. 𝐈𝐟 𝐭𝐡𝐞 𝐏𝐫𝐞𝐝𝐢𝐜𝐭𝐨𝐫 𝐢𝐬 𝐐𝐮𝐚𝐧𝐭𝐢𝐭𝐚𝐭𝐢𝐯𝐞:

✅Outcome Categorical? Use Logistic Regression.
✅Outcome Quantitative? Ask: How many predictor variables?
✅One? Use Simple Regression.
✅More than one? Use Multiple Regression.

More Details: https://www.statisticalaid.com/how-to-choose-the-right-statistical-test/

𝐇𝐨𝐰 𝐭𝐨 𝐂𝐡𝐨𝐨𝐬𝐞 𝐑𝐢𝐠𝐡𝐭 𝐑𝐞𝐠𝐫𝐞𝐬𝐬𝐢𝐨𝐧 𝐀𝐧𝐚𝐥𝐲𝐬𝐢𝐬?

Regression analysis is a powerful statistical tool used to understand the relationship between a dependent variable and one or more independent variables. It's a cornerstone of data analysis, allowing us to model, predict, and gain valuable insights from data. Here are some steps to choose the right regression model.

✅Step-1: What type of dependent variable do you have?
Continuous: Go to step 2.
Categorical: Go to step 5.
Count: Go to step 6.
Time-to-Event: Go to Cox Proportional Hazards Regression.

✅Step-2: Is the relationship between the variables approximately linear?
Yes: Go to step 3.
No: Consider Polynomial Regression, Spline Regression, or more advanced machine learning models like Decision Tree Regression, Random Forest Regression, or Neural Network Regression.

✅Step-3: How many independent variables do you have?
One: Use Simple Linear Regression.
Multiple: Use Multiple Linear Regression. Go to step 4.

✅Step-4: Is there multicollinearity among the independent variables, or do you have a high number of predictors?
Yes: Consider Ridge Regression, Lasso Regression, or Elastic Net Regression to address multicollinearity and prevent overfitting. Consider PCA for dimensionality reduction.
No: Proceed with Multiple Linear Regression, ensuring the assumptions of linear regression are met.

✅Step-5: How many categories does your dependent variable have?
Two: Use Logistic Regression.
More than two, unordered: Use Multinomial Logistic Regression.
More than two, ordered: Use Ordinal Logistic Regression.

✅Step-6: Does the count data exhibit overdispersion (variance > mean)?
No: Use Poisson Regression.
Yes: Use Negative Binomial Regression.

More Details: https://www.statisticalaid.com/choosing-the-right-regression-analysis/

𝐖𝐡𝐚𝐭 𝐢𝐬 𝐏𝐨𝐥𝐲𝐧𝐨𝐦𝐢𝐚𝐥 𝐑𝐞𝐠𝐫𝐞𝐬𝐬𝐢𝐨𝐧?

Polynomial regression is a form of linear regression in which the relationship between the independent variable(s) (predictors) and the dependent variable (response) is modeled as an nth degree polynomial.

Standard linear regression uses the equation:

y = b0 + b1*x

Where:

✅y is the dependent variable

✅x is the independent variable

✅b0 is the y-intercept

✅b1 is the slope

Polynomial regression, on the other hand, extends this by adding polynomial terms of x:

y = b0 + b1*x + b2*x^2 + b3*x^3 + ... + bn*x^n

Here:

✅n is the degree of the polynomial

✅b2, b3, ..., bn are the coefficients for the polynomial terms

More Details: https://www.statisticalaid.com/polynomial-regression/