Analysis of an E-learning platform Foresight -BI and Analytics.

7 min readDec 23, 2022

Introduction:

Foresight BI & Analytics Global Solutions (Foresight BI) is a business intelligence and data analytics firm focused on delivering world-class training and consulting services.

Foresight BI consulting help unlock the immense potential within the “most experienced staff in any organization” — DATA, to drive strategy and decision-making.

Foresight BI E-learning specializes in developing Microsoft Power BI and reporting training for individuals and organizations.

The firm has some self-paced courses hosted on a Learning Management System (LMS). The analytics provided by the LMS application does not provide enough information required by management.

Here, this project is birthed.

Project Statement /Objective

The following project will be achieved with the help of Microsoft Excel. Tools in Microsoft Excel like Power Query, Power Pivot, and Pivot tables will be used.

Power Query will be used to clean the dataset, check for errors, detect anomalies in data, and change data types.
Power Pivot will be used to create relationships between data tables to allow for filtering tables together for analysis
Pivot tables will be used to analyze

This Project seeks to identify the following

Total number of students enrolled

2. Total number of courses available

3. Total number of enrolled courses

4. Types of courses(Learning Style)

5. What type of learning style did students must enroll in

6. Completion status of courses

7. Under which learning style did students complete their course the most

8. Student’s Performance on the quiz.

Data sourcing

Data sourcing is a process in which data is extracted and integrated from multiple internal and sometimes external sources. This process brings about authenticity that is used for handling day-to-day workflow and activities for achieving business goals.

This dataset is gotten from the tweets of Ahmed Oyelowo about project portfolios and it is downloaded from the Foresight BI website.

Data Cleaning

Data cleaning is the process of examining data and transforming it into a more usable and readable form.

The dataset was imported into Power Query for cleaning. There are seven tables in the E-learning dataset. The Field Description table was removed as it is not needed in this analysis. The Field Description table only describes the content of the other tables in this dataset.

E-learning tables getting imported into Power Query

The next step in the cleaning phase is to change the column headers, the column headers are inconsistent, unreadable and some were non-understandable

A consistent pattern with the tables are

Full stops in between headers
Inconsistent header column
Unreadable header

Power Query Editor(Before changing the headers.)

Next, I changed each column header to make them understandable and clear. As it would come in handy in the next step of the analysis.

Power Query Editor (After changing the headers.)

Checked for and removed duplicates in each table with the unique keys.

The above image is the Power Query editor, on the right are some of the steps carried out in cleaning the data.

After making column headers consistent and changing data types to a better fit for analysis, some of the data types were not in the format I desired. For example in the enrollment table, the % percentage column shows values in decimal points.

0.3125
I would prefer the percentage to display values in percentages instead. 
The newly formatted column now displays the value as
3.13
which equals the same as the previous value before formatting

After all, the tables have been cleaned, the next step is to load them into tables and create connection queries. This would enable us to load data into the data model.

Data Modelling

Data Modelling is visualized as a collection of tables. Having cleaned my data, I have to connect them. This is the only way I can relate the tables, aside from being tables in the same excel file. The dataset has no relationship with one another. Creating a relationship would allow for filtering between the tables. Without connecting the tables no analysis can be done.

Tables are connected with columns, these columns must have

must have the same data format, where the relationship would be created. For example, a numeric data type will only connect to another column that has a numeric data type
One of the columns must have unique values.

Relationships are created by selecting columns that contain the data that is related to a Lookup table(A lookup table is a table we wish to connect with another table). This is why it is important to have the columns of interest for joining tables in the same headers.

Creating a relationship opens a gateway to limitless possibilities in how these tables can be filtered against each other for analysis.

Data Exploration

Data Exploration is a crucial step in data analytics used to explore and visualize data to uncover insights. This includes a graphical representation of data.

After cleaning and modeling the data, the next step is to analyze the data to gain insights.

1. Total number of Students enrolled

Using the COUNT() function, to count each row, The total number of students enrolled is 1850 from 2020 to 2022.

=COUNT(Enrollments[User ID])
 1850

2. Total number of courses available

Here, I used COUNT()function to count the total number of courses available. There are 27 courses in total

=COUNT(Courses[Course ID])
 27

3. Total number of courses enrolled

The course with the most enrolled students is the Introduction to Power BI course with 809 enrollments and the least course enrolled students is Excel BI: Analyzing & Visualizing Data With Microsoft Excel with 4 student enrollments(replace picture) include a chart picture too

After grouping the Course Name by the Count of Enrollment ID, I noticed the rows became shorter which means all 27 courses were not enrolled. I counted the number of rows and found out about five courses were not enrolled.

DA 100 practice Questions
Data Analytics Essentials
Excel Business Intelligence
Intro to Power BI
Power BI Bootcamp 8are the courses that are not enrolled. Which makes a total of 22 courses enrolled

4. Types of courses

There are four types of courses available in the E-learning platform

4 
Types of learning

5. What type of learning did students enroll in?

To get the total number of students who enrolled in the different learning styles (Types of learning). I used a pivot table which was then analyzed with a bar chart.

The learning style most used by students is Online Self Paced and the least learning style adopted by students is the Online test

Online Self-Paced 1385Online Live 385Survey 83Online test 23

6. Completion Status of courses

Here the completion Status of the Courses is calculated with a Pivot table and further analyzed with a Pie chart.

A high number of the students did not complete their course

1390 Students did not complete their course460 Students completed their course

7. Under which learning style did students complete their course the most

Pivot chart alongside a Pivot table showing the category of the number of students based on their learning mode that completed the quiz

The above chart represents the number of students who didn't complete their course grouped by the types of learning. A higher number of students didn't complete their courses. The ratio of students who completed their courses to those who didn't complete their course is greater.

8. Student’s performance on the quiz.

The course with the most enrollments is the Introduction to Power BI course which has 809 numbers of enrollments. In the Quiz Detailed table, the questions column was centered on the PowerBI course which means quizzes were only created for the PowerBI courses.