Skip to main content

WHEN TO USE WHAT STATISTICAL TEST IN RESEARCH

There are several statistical test types for analyzing Research Data. When to use what is often the challenge. This piece provides a simplification 

1️⃣t-test:

- Use when: You want to compare the means of two groups to determine if there's a significant difference.
- Example: You want to compare the average score of students who received traditional teaching vs. those who received innovative teaching.

2️⃣ANOVA (Analysis of Variance):

- Use when: You want to compare the means of three or more groups to determine if there are significant differences.
- Example: You want to compare the average score of students from different schools to determine if there are significant differences in their performance.

3️⃣Regression (Simple and Multiple):

- Use when: You want to examine the relationship between a dependent variable and one or more independent variables.
- Example: You want to examine the relationship between hours studied and exam scores (simple regression), or the relationship between hours studied, exam scores, and student motivation (multiple regression).

4️⃣Chi-squared test:

- Use when: You want to determine if there's a significant association between two categorical variables.
- Example: You want to determine if there's a significant association between smoking and lung cancer.

5️⃣Wilcoxon rank-sum test (Mann-Whitney U test):

- Use when: You want to compare the distributions of two independent groups.
- Example: You want to compare the distribution of scores between students who received traditional teaching and those who received innovative teaching.

6️⃣Kruskal-Wallis H test:

- Use when: You want to compare the distributions of three or more independent groups.
- Example: You want to compare the distribution of scores among students from different schools.

7️⃣Friedman test:

- Use when: You want to compare the distributions of three or more related groups.
- Example: You want to compare the distribution of scores among students at different time points.

8️⃣Pearson correlation coefficient:

- Use when: You want to examine the linear relationship between two continuous variables.
- Example: You want to examine the relationship between hours studied and exam scores.

9️⃣Spearman rank correlation coefficient:

- Use when: You want to examine the relationship between two variables when data is not normally distributed.
- Example: You want to examine the relationship between ranking of favorite foods and ranking of nutritional value.

🔟Kendall's tau correlation coefficient:

- Use when: You want to examine the relationship between two variables when data is ordinal or categorical.
- Example: You want to examine the relationship between socioeconomic status and education level.

1️⃣1️⃣ARIMA models:

- Use when: You want to forecast future values in a time series data.
- Example: You want to predict stock prices based on past trends.

1️⃣2️⃣Exponential smoothing (ES):

- Use when: You want to forecast future values in a time series data with a simple exponential smoothing method.
- Example: You want to predict sales based on past trends.

1️⃣3️⃣Seasonal decomposition:

- Use when: You want to decompose time series data into trend, seasonality, and residuals.
- Example: You want to analyze website traffic data to identify seasonal patterns.

1️⃣4️⃣Kaplan-Meier estimator:

- Use when: You want to estimate the survival function of a population.
- Example: You want to analyze the survival rate of patients with a specific disease.

1️⃣5️⃣Cox proportional hazards model:

- Use when: You want to examine the relationship between covariates and survival time.
- Example: You want to investigate the effect of treatment on survival time.

1️⃣6️⃣Log-rank test:

- Use when: You want to compare the survival curves of two or more groups.
- Example: You want to compare the survival rates of patients with different treatments.

1️⃣7️⃣K-means clustering:

- Use when: You want to group similar observations into clusters based on features.
- Example: You want to segment customers based on buying behavior.

1️⃣8️⃣Hierarchical clustering:

- Use when: You want to group similar observations into clusters based on features, with a hierarchical structure.
- Example: You want to analyze gene expression data to identify clusters of genes.

1️⃣9️⃣DBSCAN (density-based spatial clustering of applications with noise):

- Use when: You want to group similar observations into clusters based on features, with noise handling.
- Example: You want to analyze spatial data to identify clusters of high density.

2️⃣0️⃣Principal component analysis (PCA):

- Use when: You want to reduce the dimensionality of a dataset by identifying principal components.
- Example: You want to analyze stock prices to identify principal components of variation.

2️⃣1️⃣Discriminant analysis:

- Use when: You want to predict group membership based on multivariate data.
- Example: You want to predict customer churn based on usage patterns.

2️⃣2️⃣Canonical correlation analysis:

- Use when: You want to examine the relationship between two sets of multivariate data.
- Example: You want to investigate the relationship between personality traits and behavior.

2️⃣3️⃣Bayesian inference:

- Use when: You want to update probabilities based on new data.
- Example: You want to update the probability of a hypothesis based on new evidence.

2️⃣4️⃣Bayesian regression:

- Use when: You want to model the relationship between variables using Bayesian methods.
- Example:

2️⃣5️⃣Bayesian networks:

- Use when: You want to model complex relationships between variables using Bayesian methods.
- Example: You want to model the relationship between genes and diseases.

2️⃣6️⃣Decision trees:

- Use when: You want to classify observations based on a tree-like model.
- Example: You want to predict customer churn based on usage patterns.

2️⃣7️⃣Random forests:

- Use when: You want to classify observations based on an ensemble of decision trees.
- Example: You want to predict disease diagnosis based on symptoms.

2️⃣8️⃣Support vector machines (SVMs):

- Use when: You want to classify observations based on a hyperplane.
- Example: You want to predict customer churn based on usage patterns.

2️⃣9️⃣Cluster analysis:

- Use when: You want to group similar observations into clusters based on features.
- Example: You want to segment customers based on buying behavior.

3️⃣0️⃣Factor analysis:

- Use when: You want to reduce the dimensionality of a dataset by identifying underlying factors.
- Example: You want to analyze survey data to identify underlying factors of satisfaction.

3️⃣1️⃣Survival analysis:

- Use when: You want to analyze the time-to-event data.
- Example: You want to analyze the survival rate of patients with a specific disease.

3️⃣2️⃣Time-series analysis:

- Use when: You want to analyze data that is ordered in time.
- Example: You want to analyze stock prices to identify patterns and trends.

3️⃣3️⃣Non-parametric tests:

- Use when: You want to analyze data without assuming a specific distribution.
- Example: You want to compare the median scores of students who received traditional teaching vs. those who received innovative teaching.

3️⃣4️⃣Machine learning algorithms:

- Use when: You want to predict outcomes or classify observations based on large datasets.
- Example: You want to predict customer churn based on usage patterns.

The specific test or technique used depends on the research question, data type, and study design.




Comments

Popular posts from this blog

Evaluation and Characteristics of Himalayas

Time Period Event / Process Geological Evidence Key Terms & Concepts Late Precambrian – Palaeozoic (>541 Ma – ~250 Ma) India part of Gondwana , north bordered by Cimmerian Superterranes, separated from Eurasia by Paleo-Tethys Ocean . Pan-African granitic intrusions (~500 Ma), unconformity between Ordovician conglomerates & Cambrian sediments. Gondwana, Paleo-Tethys Ocean, Pan-African orogeny, unconformity, granitic intrusions, Cimmerian Superterranes. Early Carboniferous – Early Permian (~359 – 272 Ma) Rifting between India & Cimmerian Superterranes → Neotethys Ocean formation. Rift-related sediments, passive margin sequences. Rifting, Neotethys Ocean, passive continental margin. Norian (210 Ma) – Callovian (160–155 Ma) Gondwana split into East & West; India part of East Gondwana with Australia & Antarctica. Rift basins, oceanic crust formation. Continental breakup, East Gondwana, West Gondwana, oceanic crust. Early Cretaceous (130–125 Ma) India broke fr...

Seismicity and Earthquakes, Isostasy and Gravity

1. Seismicity and Earthquakes in the Indian Subcontinent Key Concept: Seismicity Definition : The occurrence, frequency, and magnitude of earthquakes in a region. In India, seismicity is high due to active tectonic processes . Plate Tectonics 🌏 Indian Plate : Moves northward at about 5 cm/year. Collision with Eurasian Plate : Causes intense crustal deformation , mountain building (Himalayas), and earthquakes. This is an example of a continental-continental collision zone . Seismic Zones of India Classified into Zone II, III, IV, V (Bureau of Indian Standards, BIS). Zone V = highest hazard (e.g., Himalayas, Northeast India). Zone II = lowest hazard (e.g., parts of peninsular India). Earthquake Hazards ⚠️ Himalayas: prone to large shallow-focus earthquakes due to active thrust faulting. Northeast India: complex subduction and strike-slip faults . Examples: 1897 Shillong Earthquake (Magnitude ~8.1) 1950 Assam–Tib...

geostationary and sun-synchronous

Orbital characteristics of Remote sensing satellite geostationary and sun-synchronous  Orbits in Remote Sensing Orbit = the path a satellite follows around the Earth. The orbit determines what part of Earth the satellite can see , how often it revisits , and what applications it is good for . Remote sensing satellites mainly use two standard orbits : Geostationary Orbit (GEO) Sun-Synchronous Orbit (SSO)  Geostationary Satellites (GEO) Characteristics Altitude : ~35,786 km above the equator. Period : 24 hours → same as Earth's rotation. Orbit type : Circular, directly above the equator . Appears "stationary" over one fixed point on Earth. Concepts & Terminologies Geosynchronous = orbit period matches Earth's rotation (24h). Geostationary = special type of geosynchronous orbit directly above equator → looks fixed. Continuous coverage : Can monitor the same area all the time. Applications Weather...

Network data model

GIS, a network data model is used to represent and study things that are connected like a web — for example, roads, rivers, railway tracks, water pipes, or electric lines . It focuses on how things are connected and helps us solve problems like finding the best route, the nearest hospital, or where water will flow. Nodes → Points where things meet or end (e.g., road intersections, railway stations, pumping stations). Edges → Lines connecting the nodes (e.g., roads, pipelines, cables). Topology → The "rules" of connection — which node is linked to which edge. Attributes → Extra details about each part (e.g., road speed limit, pipe size, traffic volume). How It Works 🔍 Make the Network Model Start with a map of lines (roads, pipes, rivers) and mark how they connect. Run Analyses Routing → Find the shortest or fastest path. Closest Facility → Find the nearest hospital, petrol station, etc. Service Area → Find how far y...

Pre During and Post Disaster

Disaster management is a structured approach aimed at reducing risks, responding effectively, and ensuring a swift recovery from disasters. It consists of three main phases: Pre-Disaster (Mitigation & Preparedness), During Disaster (Response), and Post-Disaster (Recovery). These phases involve various strategies, policies, and actions to protect lives, property, and the environment. Below is a breakdown of each phase with key concepts, terminologies, and examples. 1. Pre-Disaster Phase (Mitigation and Preparedness) Mitigation: This phase focuses on reducing the severity of a disaster by minimizing risks and vulnerabilities. It involves structural and non-structural measures. Hazard Identification: Recognizing potential natural and human-made hazards (e.g., earthquakes, floods, industrial accidents). Risk Assessment: Evaluating the probability and consequences of disasters using GIS, remote sensing, and historical data. Vulnerability Analysis: Identifying areas and p...