DEV Community: Benard Otieno

Parametric vs Non-parametric tests

Benard Otieno — Sun, 13 Sep 2026 18:55:40 +0000

Introduction
Statistical test is one of the first road maps when decieding whether to use parametric or non-parametric tests. The choice affects how valid your results are, that is the power of your test and the confidence you place in your conclusions. Choosing whether to use a parametric or a no-parametric test has a great impact on your prediction either negatively or positively.

Parametric Test
Parametric test assumes that a polpulation data follows a specific probability distribution, commonly known as normal distribution. This distribution can be described by a set of parameters namely; mean and variance. Since the distribution of parametric tests is assumed to be known, the test tends to be be statistically powerful and can extract more information from the data.

Core Assumptions of Parametric Tests

Normality: The data is normally distributed or approximately normally distributed.
Homogeneity of variance: Groups that are bieng compared have similar or approximately similar variance.
Interval or ratio scale data: Meanignful spacing of data.
Independence of observations: One data point does not have an influence on the others.

Types of Parametric tests

One sample t-test
Used to compare one sample mean to a known value. For instance, the standard weight of bred is 400grams, if we have 10 loaves of bread, we add their weights and get the mean of the 10 loaves of bread and compare it with the standard 400grams.
Independant t-test
Compares the mean of two independant groups. The independant variable is always a categorical variable while the dependant variable is a continous variable. For instance a drug test on two sets of five people, the first five people are given 300mg of panadol while the second group is given 500mg of panadol at the same time to see the of both groups after the test.

3.Paired t-test
Compares the mean of two related samples. For instance a drug test on five people, for the first they are given 300mg of panadol while for the second test they are given 500mg of panadol at a different time to see the results before and after. The independant variable is always a categorical variable while the dependant variable is a continous variable.

4.One-way ANOVA
Compare means of more than two groups, Compare average satisfaction scores across three product tiers. Similarly, the independant variable is always a categorical variable while the dependant variable is a continous variable.

5.Pearson correlation
Measure linear relationship between two continuous variables. Used for numerical variables.

6.Linear regression
Model a continuous outcome using predictors

Non-parametric test
Non-parametric test has little to no assumption about the distribution. Many non-parametric tests convert data into ranks instead of directly working with raw data. This makes them resistant to skewed distribution, non-normal distributions and outliers. Non-parametric tests are suitable when:

You have an ordinal data instead of continous.
Your data has extreme outliers or heavily skewed.
You have a categorical data and you're testing for association instead of means.
The sample size of your data is small, making it difficult to validate normality.

Types Non-Parametric Tests

Mann-Whitney U test This is the alternative for Independant t-test that does not qualify the parametric tests assumptions.

Wilcoxon signed-rank test This is the alternative for paired t-test that does not meet the parametric tests assumptions.

Kruskal-Wallis test This is the alternative for One-way ANOVA that does not meet the parametric tests assumptions.

Spearman's rank correlation This is the alternative for Pearson correlation that does not meet the parametric tests assumptions.

Chi-square test of independence This is a non-parametric test used to check for relationship between categorical variables.

The Table below shows side by side comparison of parametric and non-parametric tests.

Conclusion
Choosing between parametric and non-parametric tests does not really matter whether you chose the correct on or not, It solely relies on the nature of data. When assumption are held, parametric tests offer more power in terms of statistical analysis. However, they can produce results that are misleading when applied to data that is ordinal, small and has a skewed distribution. Therefore non-parametric tests offer alternatives when data does not meet the assumptions. In conclusion, before running any tests, ensure to check the assumptions in-order to produce correct results.

Data Distribution and their impact in Data Science.

Benard Otieno — Sun, 13 Sep 2026 08:41:45 +0000

Introduction
Datasets show distribution and it's shape. Distribution is the pattern that explains hows values are spread from a variable. Data distribution is important in data science since it influences most of the decisions that follows. For instance, the valid statistical test to use, the machine learning algorithm that will perform well, hoe to handle extreme outliers and how to honestly communicate results.

What is Data Distribution
Data distribution is a graphical represenattion that explains the frequency of occurennces of each valuein a dataset. It is mostlyb visualised using tools like boxplots, histograms and density plots.

Why Distributions Matter in Data Science

Choosing the right statistical tests
**
Many hypothesis tests assume normal distribution of data ansd therefore applying heavily skewed datawithout checking the assumptions leads to false conclusion. Alternatively, Non-parametric tests exists for the cases where there is no normal distribution.
**
Outliers Detection
**
Outliers are data points that have statistical anomaly, in simple terms outliers are data points that significantly deviates from other data points in the dataset. They can be brought about by natural variation or measurement errors and they can have a negative impact on the statistical analysis and machine learning model.

Model Selection ** Some models are built around certain distributional assumptions:

Linear Regresssion; This model assumes normal distribution.
Poisson Regression; Designed to count data that follows a poisson distribution.
Naive Bayes classifiers; Mostly assumes data following Gaussian distribution

## Common types of Data Distributions

1. Normal Distribution
This is the most common type of distribution also known as Bell-Curve. It is symmetric around the mean where most of the values cluster near the center.

2. Uniform Distribution
In uniform distribution every outcome within a range is distributed equally.

3. Right Skewed Distribution
In Right skewed distribution, the tail stretches to the right, in most ocassion the mean is higher than the median. This is brought about by outliers that dragging the mean to the right. It is also known as a positive distribution.

4. Left Skewed Distribution
In Left skewed distribution, the tail stretches to the left, in most ocassion the median is higher than the mean. This is brought about by outliers that dragging the mean to the left. It is also known as a negative distribution.

5. Binomial Distribution
Explains the number of successes in a fixed number of yes or no trials. Mostly used in quality control and A/B testing conversion rate analysis.

Practical Tips for Working with Distributions

Visualize your data before modelling.
Test for normality
Do not fix normality where it does not belong
Check for multimodal distribution
After pre-processing, re-check distribution

Visualizing Distributions
Visualization is one of the easiest ways to check how your data is distributed. Some of the ways we can visualise to check for distribution include:

Histogram Visualizes countinous data into intervals known as bins and in each bin, frequency of occurence is counted. Hence showing the overal distribution of data.

Box Plot Visualizes distribution of numeric variables using five key summary statistics.

The box spans the middle 50% of the data, from Q1 which is 25th percentile to Q3 which is 75th percentile.
The 50th percentile is marked by the line inside the box which shows the median and not mean, hence making boxplot resistant to skew and outliers
The whiskers extend to the smallest and largest values that are still within 1.5 × IQR of the box edges.
Points that are beyond the whiskers are outliers

Conlusion
Data distributions is a practical lens that shapes nearly every step of the data science workflow, from exploratory data analysis(EDA) to feature engineering to model selection and evaluation. A data scientist who understands the distribution of their data will make better decisions about which tools to use, avoid misleading conclusions, and build models that generalize more reliably to the real world. Before running a single test or training a single model, the first and most valuable question to ask is often the simplest one: _what does this data actually look like _

How to publish a Power BI report and embed it in a website.

Benard Otieno — Sat, 04 Apr 2026 12:07:08 +0000

Introduction to Power BI
Power BI is a data analytics tool used to turn raw data into interactive reports and dashboard. It enables users to obtain insights, visualize data and make decisions that are data driven.

Publishing and Embeding Process
In Power BI, this process involves workspace creation in a Power BI service, thereafter you upload and publish the Power BI report, before you generate and embed code and lastly embedding the report to the website.

1. Creating a Workspace
In Power BI, a worksapce is a collaborative shared enviroment for hosting dashboards, reports and managing datasets. Below are steps and examples of how to create a workspace.

Step 1: Open your browser and click this website https://app.powerbi.com/home?experience=power-bi
Step 2: Sign in to your account.
Step 3: Once you're logged in, go to the left navigation paneand click workspaces.

Step 4: After opening the workspaces, click +New Workspace
Step 5: Lastly enter the workspace name and description to the pop up then click apply to save.

2. Uploading and Publishing a Power BI report
Open the Power BI report in your computer then click Publish which is located at the top right.

Thereafter select your workspace and the report will be uploaded and the and becomes accessible online.

3. Generating the embed code
Power BI enables us to embed reports via iframe code. To begin you need to open the report in the Power BI service, then click file, hover down to embed report and lastly click website or portal to create the embed code.

4. Embedding the Report on a Website
Since we now have the code, we can easily intergrate it in the HTML code. Open this website in your browser https://www.w3schools.com/html/ and copy the html code.

Open a text editor and go back to the desktop a create a folder with a name of your choice. Go back to the text editor and and open a new folder and select the folder you created.

Now in the text editor go to the folder you created and create a new file with a name of your choice, then paste the HTML code that was earlier copied. Finally, in the text editor delete line nine of the html code and replace it with the embed code and save then open the folder you created in the desktop and open the file in the folder which leads to this page below.

Lastly sign in to the page and your report will be visible.

Conclusion
Publishing a power BI report and embedding it enables the sharing of insights not only on the Power BI platform. If there is proper optimization and security , embedded Power BI reports has the potential to become a powerful feature of modern data-driven websites.

Understanding Data Modeling in Power BI.

Benard Otieno — Mon, 30 Mar 2026 14:36:50 +0000

Data Modelling
Data modelling involves creating a visual diagram of data to show how it can be organized, structered and analysed within a database. In Power BI, data modelling enables ease in report creation, accuracy in calculations and efficient performance.

SQL JOINS
SQL JOINS is used to combine data in multiple tables based on the related columns between them. They are important in relational database since it enables us to retrieve data with unified view that is mostly spread across multiple tables.

INNER JOIN
INNER JOIN is an SQL operation that enables combining rows from multiple tables with respect to a related column.

LEFT JOIN
LEFT JOIN retains rows from the left side of the table, and combines rows that are matching from the right side of the table.

RIGHT JOIN
RIGHT JOIN merges two tables by retaining data from the right ide of the table and only matching data from the left side of the table.

FULL OUTER JOIN
FULL OUTER JOIN merges two tables by retaining all the data from both tables, regardless of whether they are matching.

LEFT ANTI JOIN
LEFT ANTI JOIN returns rows from the left side of the table that do not have a match on the right side of the table.

RIGHT ANTI JOIN
RIGHT ANTI JOIN returns rows from the right side of the table that do not have a match on the left side of the table.

How to access JOINS in a Power BI query
To access JOINS, go to HOME then click transform data > open power query editor > choose second table > select matching columns, then select the join type and lastly click ok.

Relationships in Power BI
Relationships connects tables, without merging them physically, enabling records from multiple places to be used jointly in reports. There are four types of relationships namely;

One-to-many
One-to-many connects a lookup table that contains values that are unique , to a table, where the values repeat.

Many-to-many
Many-to-many normally occur in a relationship when both tables contain duplicate values , enabling multiple data in one table to match multiple data in another.

One-to-one
One-to-one connects two different tables whereby two unique values from the different tables are matched.

Cardinality
Cardinality explains the kind of the relationship between table A and table B, describing how unique values in one column relate to values in another column.

Cross-filter Direction
Cross-filter Direction determines the flow of filters in related tables, with options for single direction or both direction.

Active vs Inactive Relationships
Active relationships are used in visuals by default while Inactive relationships has to be activated using DAX.

Creating Relationships in Power BI

Go to Model View
Drag a column from one table to the next
Lastly set to Cardinality or Cross-filter direction

Joins vs Relationships
In Joins, data is combined in one table, hence increasing the size of dataset, while in relationship, the performance is fast and efficient due to the schemas.

Fact and Dimension Tables
Fact tables has data that is quantitative and forms the star schema, while dimension tables hold descriptive data.

Data Modeling Schemas
A Schema is a framework and organization of data within in a model, which is significant for efficient data analysis.There are two types of schemas namely Star schema and Snowflake schema.

Star schema
Star schema organizes data into a central fact table and is surrounded by dimension tables.

Snowflake schema
Snowflake schema is an approach where dimension tables are normalized and broken into numerous tables that are related.

Flat Table
Flat Table is a table that contains all data, including descriptive and transactional facts.

Conclusion
Power BI seemed something complicated to me until I engaged in hands on learning that has made it easier for me to understand. Data medelling is a very important to learn since it is fundamental in analysing and reporting.

How Excel is Used in Real-World Data Analysis

Benard Otieno — Thu, 26 Mar 2026 20:25:17 +0000

Microsoft Excel is a tool used to collect, organize, analyse and visualize data. Various industries use excel to analyse data, for instance the businesses and marketing sector. Businesses use excel to keep track of their income, profit and losses, so as to see where they can improve on. On the other hand, sales and marketing sector use excel to analyse trend in sales and consumer data.
Some of the excel features include example:

*Formulas *
AVERAGE, Which is used to calculate the average e.g. =AVERAGE(B2:B30)

SUM, Which is used to addition of a range of numbers e.g SUM(B2:B30)

Another feature used in excel also includes pivot table which is used to analyse big data. Example of a Pivot table

Pivort Table
Pivot tables is an excel tool used to analyse and simplify data without using formulas. To access the pivot table, click anywhere in the data and go to insert, then at the top right corner click pivot table then select New worksheet.

Slicers
Slicers are used in excel to filter pivort tables and graphs. It is a easy to use since it helps display what is currently filtered by just clicking the button menus.

Graphs
Grphs are used in excel to show a visual representation of numerical data and analyse the data in more easier way to understand and intepret. Graphs enable the identifivation of trends, highlighting anomalies and comparing categories. Exaples of graphs include, line grapg and bar graph etc.

Dashboard
A dashboard is a visual report that helps to simplify Key Perfomance Indicater through tables, charts and graphs that are interactive. Dashboard enables raw data to be transformed into meaningful insights, to help in decision making by enabling users to filter data through slicers for automatic update of data.

In Summary
Microsoft Excel is a significant tool for data analysis. Excel has the ability to do simple calculations, it also enables users to turn raw data into meaningful insights. Excel features such as Pivot Tables, and visualization tools like the graphs, individuals can significantly enhance their ability to analyze and interpret data.

Excel for beginners

Benard Otieno — Thu, 26 Mar 2026 12:22:24 +0000