Project Based Learning

Project work in the CBSE IP syllabus

The CBSE 2025-26 syllabus makes the project part of the Class XII practical examination. Key facts a teacher must share with students at the start of the session:

• Weightage: Project Work carries 5 marks out of the 30 practical marks (the others are Pandas/Matplotlib programs 8, SQL queries 7, practical file 5 and viva-voce 5).
• Aim: to create a tangible and useful IT application by identifying a real-world problem from the student's surroundings, such as a shop, a business, a community group or the school itself.
• Content: students take data stored in a CSV file or a database, analyse it with Python libraries (Pandas), and present findings with suitable charts (Matplotlib). Concepts from both Class XI and Class XII may be used.
• Group size: individually or in groups of 2 to 3 students.
• Start date: at least 6 months before the submission deadline.
• Ethics: students must avoid plagiarism and copyright violation, and every resource used (data, images, code) must be properly referenced.

What is project-based learning?

In project-based learning, students learn by working on a real problem over an extended period instead of solving short textbook exercises. They must break the problem into parts, gather data, choose methods, build a solution and report the result.

For IP students the benefits are concrete:

• Pandas, Matplotlib and SQL are used together on one meaningful dataset, so students see why each tool exists.
• Students practise skills that board exams cannot test easily: planning, teamwork, decision-making, communication and time management.
• The finished project gives students something to discuss confidently in the viva and, later, in college or job interviews.

The teacher acts as a guide: approving topics, setting milestones, reviewing progress and checking originality, but not writing the code.

Steps in project-based learning

A project moves through five broad stages. Students should be able to name and explain them.

1. Identification of the project: find a real, useful problem. Interdisciplinary topics (IP with Economics, Geography, Biology or Physical Education) are encouraged.
2. Defining a plan: choose a team leader, define each member's role, list the tools needed, and think about extreme or unusual cases.
3. Fixing a time frame: divide the work into activities, estimate the time each needs, and fix milestone dates while leaving some flexibility.
4. Guidance and monitoring: when stuck, consult books, websites and experts; the team leader and guide teacher monitor progress regularly.
5. Outcome of the project: produce and present the result. The outcome may be single or multiple, and it should be reviewed by peers and the teacher and improved on the basis of feedback.

Approaches to solving a project

Large problems are easier to solve when divided into smaller parts. Three common approaches:

• Modular approach: the project is divided into independent modules, each with a well-defined task, inputs and outputs. The outputs are integrated at the end. In a canteen project, the modules could be data entry, sales analysis, stock report and charts.
• Top-down approach: start with the overall goal and keep breaking it into smaller sub-tasks until each is simple enough to code. Example: "Analyse results" breaks into "compute totals", "compute grades", "compare sections", "plot charts".
• Bottom-up approach: first build small, working pieces (a function to compute a grade, a function to draw a chart) and then combine them into the complete system.

In practice, teams plan top-down and build module by module, testing each piece before integration.

Step 1: identifying a real problem

The syllabus suggests that students explore their surroundings: visit shops, businesses, offices or community groups, and ask how data is generated, stored and used there. Good sources of ideas:

• The school: results, attendance, library, canteen, transport, sports day, fee collection.
• The neighbourhood: a kirana store, a pharmacy, a tuition centre, a resident welfare association.
• Public open data: weather, rainfall, census, crop production, tourism, from data.gov.in.
• Interests: cricket, films, music charts, fitness tracking.

A good project problem is specific, has data that can actually be collected, and leads to questions that Pandas and charts can answer. "Study pollution" is too vague; "Compare monthly PM2.5 levels of Delhi and Bengaluru for 2023 using CPCB data" is a workable project.

Step 2: collecting and organising data

Data can be primary (collected by the students themselves) or secondary (collected by someone else and reused).

• Primary: a survey made with an online form tool such as Google Forms, a register copied from the school canteen or library, readings taken by students.
• Secondary: open government datasets from data.gov.in, published sports statistics, reports from government departments.

Whatever the source, store the data as a CSV file with one row per record and one column per attribute. Keep a short data dictionary: column name, meaning, data type and units. Keep the raw file unchanged and save cleaned data in a separate file, so that the original can always be checked.

Step 3: designing the solution

Design means deciding, on paper, how each question will be answered before any code is written. For every question, students should note:

1. Which columns are needed.
2. Which operation answers it: filtering, sorting, a new computed column, groupby() with sum()/mean()/count(), or an SQL query with GROUP BY.
3. Which chart best shows the answer, with its title and axis labels.

If the project uses MySQL, the design also includes the table structure (columns, data types and primary key) and, where two tables are used, the common column for the equi-join.

Step 4: implementing with Pandas, Matplotlib and SQL

A typical IP project program follows the same pattern:

1. import pandas as pd and import matplotlib.pyplot as plt.
2. Import data with pd.read_csv() (or create it and save with to_csv()).
3. Inspect it with shape, head(), columns, dtypes.
4. Clean it (see the next section).
5. Process: add computed columns, filter with Boolean indexing, group and aggregate.
6. Visualise: line, bar and histogram charts with title, axis labels and legend; save each with plt.savefig().
7. Export the processed data with to_csv().

SQL can be used for storage and querying (for example, keeping records in MySQL tables and summarising them with GROUP BY). Moving data between MySQL and Pandas needs an extra connector library and is optional for the project.

The short program below answers two questions for a school library project.

import pandas as pd

# Question: Which genres are borrowed most, and how much late fine is due?
log = pd.DataFrame({
    "BookID": ["B101", "B205", "B118", "B330", "B205", "B412", "B101", "B330"],
    "Genre":  ["Fiction", "Science", "Fiction", "History",
               "Science", "Comics", "Fiction", "History"],
    "Class":  ["IX", "XII", "X", "XI", "XI", "IX", "XII", "X"],
    "Days":   [10, 18, 7, 21, 12, 3, 16, 9],
})
log.to_csv("library_issues.csv", index=False)      # export

df = pd.read_csv("library_issues.csv")              # import
print(df.groupby("Genre")["BookID"].count().sort_values(ascending=False).to_string())

late = df[df["Days"] > 14]                          # Boolean indexing
fine = ((late["Days"] - 14) * 2).sum()              # Rs 2 per extra day
print("Late returns:", len(late), "| Total fine: Rs", fine)

Genre
Fiction    3
History    2
Science    2
Comics     1
Late returns: 3 | Total fine: Rs 26

Choosing the right chart

The syllabus covers three chart types. Students must justify their choice in the report.

• Line plot (plt.plot()): change over time or an ordered sequence, such as daily sales or monthly temperature.
• Bar chart (plt.bar()): comparison of categories, such as revenue by item or average marks by subject. Side-by-side bars with a legend compare two groups.
• Histogram (plt.hist()): the distribution of one numeric variable, such as how many students fall in each percentage band.

Every chart needs a title, axis labels with units, and a legend whenever more than one series is drawn. Save charts with plt.savefig('name.png') so that they can be pasted into the report.

Testing and validating results

Wrong conclusions from dirty data are the most common weakness in student projects. Build these checks into every project:

• Missing values: df.isnull().sum() shows how many blanks each column has. Decide whether to drop them (dropna()) or fill them (fillna()), and record the decision in the report.
• Duplicates: df.duplicated().sum(); remove with drop_duplicates() (common when a survey form is submitted twice).
• Range checks: marks between 0 and 100, quantities not negative, dates within the study period.
• Hand check: compute one total or average by hand or in a spreadsheet and match it with the program's answer.

The example reads a small survey file screen_time.csv that has one blank answer and one duplicate response.

import pandas as pd

survey = pd.read_csv("screen_time.csv")
print(survey)
print("Missing values:", survey["Hours"].isnull().sum())
print("Duplicate rows:", survey.duplicated().sum())

clean = survey.drop_duplicates().dropna()
print("Rows after cleaning:", len(clean))
print("Mean hours:", round(clean["Hours"].mean(), 2),
      "| Median hours:", clean["Hours"].median())

   RespID Class  Hours
0       1    XI    3.5
1       2   XII    5.0
2       3    XI    NaN
3       4   XII    2.0
4       5    XI    6.5
5       6   XII    4.0
6       6   XII    4.0
Missing values: 1
Duplicate rows: 1
Rows after cleaning: 5
Mean hours: 4.2 | Median hours: 4.0

Teamwork in a project

Most IP projects are done in groups of 2 or 3. Teamwork is the combined effort of members working towards one goal; just as a cricket wicket needs both a good ball and a safe catch, a project needs every member's contribution.

Components of good teamwork:

• Communicate: meet regularly and use a shared group or document to record decisions.
• Listen: hear every member's ideas before deciding.
• Share: share data, code and knowledge; a member who is strong in Matplotlib helps others learn it.
• Respect: value every opinion; ignoring a member reduces their effort.
• Help: support a member who is stuck; take expert help when needed.
• Participate: every member must take part in coding, documentation and the viva.

Suggested roles for a team of three: Data lead (collection and cleaning), Analysis lead (Pandas/SQL), Visualisation and report lead (charts and documentation). Roles guide the work but do not limit it; every member must understand the whole project.

Plagiarism, copyright and citing data

The syllabus requires teachers to sensitise students about plagiarism and copyright, and to ensure that every resource is suitably referenced.

• Plagiarism is presenting someone else's work (code, text, charts, images, ideas) as your own. Copying a project from the internet or a senior and changing the names is plagiarism.
• Copyright protects creative work. Images, logos and datasets may be used only if their licence allows it, and the source must be credited.
• Open data from data.gov.in is published under the Government Open Data License - India, which allows reuse with attribution.
• Code adapted from a book, website or AI assistant must be acknowledged in the report, and the student must be able to explain every line.
• Data about real people (students, customers) must be used with permission and made anonymous: replace names with codes such as S01, S02.

A citation should let a reader find the same source again.

Documentation: structure of the project report

The project report (often submitted as a file) shows what was done and how. A standard structure for an IP project:

Keep the report concise and readable. Every chart should have a figure number and one or two lines explaining what it shows. Source code should be printed with comments, and outputs should be screenshots or copied text from actual runs.

Timeline: starting six months early

The syllabus asks students to start the project at least six months before the submission deadline. If practical exams are in late January or February, the project should begin in July. A month-wise plan keeps groups on track:

Assessing the project: a suggested rubric

CBSE allots 5 marks to project work but does not prescribe a detailed marking scheme. The rubric below is a suggested school-level scheme that spreads the 5 marks over the qualities the syllabus values. Share it with students at the start so that they know how they will be judged.

Twelve project ideas with an Indian context

These ideas can be adapted to any school. Each can be done with Pandas and Matplotlib; several can also use MySQL tables and SQL queries.

Preparing for the project viva

The project viva checks understanding and originality. Students should be ready to answer questions such as:

• What problem does your project solve, and who benefits?
• Where did your data come from? How many records and columns does it have? How did you cite it?
• How did you handle missing values and duplicate rows?
• Show the line of code that computes a particular result. Why did you use groupby() there?
• Why is a line plot (or bar chart, or histogram) suitable for this chart?
• Which part of the project did you do, and how did your team divide the work?
• What are the limitations of your project, and how would you improve it?

Each member should be able to run the program on the lab computer and change a small thing (a chart title, a filter value) when asked.