On this page
Informatics Practices students often prepare unit by unit and only meet the full paper in the pre-boards. By then it is late to learn how the questions are spread. The CBSE sample question paper answers that early. This post reads the 2026-27 Informatics Practices (065) paper question by question, shows how its marks match the curriculum, and works two of its questions in full.
How is the paper laid out?
The 2026-27 sample question paper is a 3-hour paper for 70 marks with 37 questions in five sections:
Section A: 21 questions of 1 mark (21 marks). Multiple choice, true/false, fill in the blank and two assertion-reason questions.
Section B: 7 questions of 2 marks (14 marks).
Section C: 4 questions of 3 marks (12 marks).
Section D: 2 questions of 4 marks (8 marks).
Section E: 3 questions of 5 marks (15 marks).
There is no overall choice, but some questions have internal choices. Programming answers must be in Python. Every page notes that there is no change in the question paper design and assessment pattern for 2026-27.
Note how this differs from computer science: IP's Section C is bigger (four 3-mark questions) and its Section D smaller (two 4-mark questions), with three 5-mark questions at the end.
Do the questions follow the curriculum's unit weights?
The 2026-27 IP curriculum gives Data Handling using Pandas and Data Visualization 25 marks, Database Query using SQL 25, Introduction to Computer Networks 10 and Societal Impacts 10, with 30 for practicals. We sorted every sample question into a unit:
Pandas and Matplotlib, 25: seven 1-mark questions (1, 4, 8, 12, 15, 18, 20), questions 22, 24 and 28 (2 each), 30 (3), 33 (4, a line graph) and 36 (5, building a DataFrame).
SQL, 25: eight 1-mark questions (2, 6, 9, 11, 14, 17, 19, 21), question 26 (2), 31 and 32 (3 each), 34 (4) and 37 (5).
Networks, 10: questions 5, 10 and 16 (1 each), 25 (2) and the campus case study in 35 (5).
Societal impacts, 10: questions 3, 7 and 13 (1 each), 23 and 27 (2 each) and 29 (3).
Each unit's total matches the curriculum exactly. Societal impacts is the only unit with no 4- or 5-mark question in this paper, so its marks come from short, precise definitions.
Worked example 1: SQL functions
Question 32(B) asks for the output of three queries.
SELECT POWER(MOD(13, 5), 2); MOD(13, 5) is the remainder 3, and 3 squared is 9.
SELECT UCASE(MID("Digital India", 9, 5)); count characters from 1: "Digital" is positions 1 to 7, the space is 8, so position 9 is "I". Five characters from there give "India", which UCASE turns into INDIA.
SELECT LENGTH(TRIM(" Informatics ")); TRIM removes the spaces at both ends, leaving "Informatics", which has 11 characters.
All three match the official marking scheme. The common slip is counting MID positions from 0, as Python does. SQL string positions start at 1.
Worked example 2: GROUP BY with HAVING
Question 34(B)(i) uses an Employee table with five rows: Aarav (D1, 5200), Riya (D2, 7500), Kabir (D2, 6800), Ananya (D3, 4000) and Ishaan (D1, 4900).
SELECT DeptID, COUNT(*), MAX(Salary) FROM Employee GROUP BY DeptID HAVING COUNT(*) > 1 ORDER BY DeptID;
Group by department: D1 has Aarav and Ishaan, D2 has Riya and Kabir, D3 has only Ananya.
HAVING COUNT(*) > 1 removes D3.
For each remaining group, find the maximum salary: D1 gives 5200, D2 gives 7500.
Output: D1, 2, 5200 and D2, 2, 7500. HAVING filters groups after grouping; WHERE would filter rows before it. Our guide to SQL functions in Class 12 IP has more practice.
What patterns repeat across the paper?
Label versus position in Pandas. Question 1 asks whether S1['b':'d'] equals S1[1:3]. The marking scheme says False: label slicing includes the end label, position slicing does not. See our Pandas Series and DataFrames guide.
Missing values. Adding two Series with different index labels gives NaN for unmatched labels (questions 12 and 30).
NULL in SQL. Question 17 tests IS NOT NULL; comparing with != NULL does not work.
Societal impacts vocabulary. Refurbishing, the "reduce" step of e-waste management, active and passive digital footprints, cookies. Our post on e-waste and health in societal impacts covers the e-waste part.
A timing plan for 70 marks
Keep 15 minutes for checking. The remaining 165 minutes give about 2.4 minutes per mark:
Section A (21 marks): about 50 minutes.
Section B (14 marks): about 33 minutes.
Section C (12 marks): about 28 minutes.
Section D (8 marks): about 19 minutes.
Section E (15 marks): about 35 minutes.
Write SQL keywords and Pandas method names exactly; a missing bracket in plt.show() or a misspelt HAVING costs part marks.
A four-week routine
Week 1: all Pandas and Matplotlib questions from the sample paper, run in Python.
Week 2: all SQL questions, run in MySQL, with outputs written before you run them.
Week 3: networks and societal impacts, as flash cards and the two case studies.
Week 4: two full timed papers, marked with the official scheme.
If a Pandas idea such as boolean indexing will not stick, the Fermi Concept Library covers every concept in physics, chemistry, biology, mathematics and computing, so you can look it up before trying the question again.
Start today: attempt question 32(B) and 34(B)(i) without looking, then check against the working above.
