Table-based DI in CAT: filling gaps, comparing fractions and avoiding base errors

A table looks like the friendliest DI format: the numbers are printed, not read off a chart. CAT makes up for that in two ways. Some cells are usually missing, so you must deduce them from totals or conditions before answering. And the questions ask for comparisons that are close, so rough estimates are not always enough.

This guide covers how to read a table, how to fill its gaps, the fraction benchmarks that make comparisons fast, and a worked set with the kind of questions CAT asks.

Reading a table before you calculate

Spend the first minute on the frame, not the numbers.

• Units. Thousands, lakhs or crores? Units or rupees? A question in rupees on a table in thousands of units needs a price from somewhere.
• What the rows and columns are. Which way do the totals run?
• What is missing. Mark every blank cell. Count the equations you have (row totals, column totals, conditions) against the unknowns.
• Time. If there are years, is the question about a change between years or a share within a year?

Filling missing cells

Every total gives you one equation. Fill any row or column that has only one blank first, then see which new rows or columns now have only one blank. Keep going. If you run out of single-blank lines, look for a condition in the set's text that links two blanks.

Fraction benchmarks

Percentages in DI are mostly fractions in disguise. These are worth knowing cold.

When two fractions are close, compare them by cross-multiplying instead of converting both. To compare 45/105 and 64/150, compare 45 × 150 = 6,750 with 64 × 105 = 6,720. The first is larger, so 45/105 is larger.

Worked set: a bookshop chain

A chain of four bookshops sells three categories of books. The table shows units sold in 2025, in thousands. Some values are missing.

Total units sold by each store in 2024 (thousands): North 100, South 120, East 90, West 100. Average selling price per unit in 2025: fiction ₹300, non-fiction ₹400, exam guides ₹250.

Step 1: fill the table

• North, exam guides: 120 − 42 − 28 = 50.
• South, fiction: 150 − 36 − 64 = 50.
• West, total: 48 + 32 + 40 = 120.
• East, non-fiction: from the column total, 121 − 28 − 36 − 32 = 25. Check against East's row: 35 + 25 + 45 = 105. It matches.
• Fiction total: 42 + 50 + 35 + 48 = 175.
• Exam guides total: 50 + 64 + 45 + 40 = 199.
• Grand total: 120 + 150 + 105 + 120 = 495. Check: 175 + 121 + 199 = 495.

Step 2: the questions

Q1. In which store did exam guides make up the largest share of that store's own sales?

• North 50/120 = 41.67%. South 64/150 = 42.67%. East 45/105 = 3/7 = 42.86%. West 40/120 = 33.33%.
• Answer: East, narrowly ahead of South. This is exactly the comparison the cross-multiplication check above settles in one step.

Q2. By approximately what percentage did the chain's fiction sales exceed its non-fiction sales?

• (175 − 121) ÷ 121 = 54 ÷ 121 ≈ 44.6%.
• The base is non-fiction, the thing being compared against. Using 175 as the base gives 30.9%, which is the answer to a different question: by what percentage non-fiction fell short of fiction.

Q3 (TITA). By how many lakh rupees did South's 2025 revenue exceed West's?

• South: 50 × 300 + 36 × 400 + 64 × 250 = 15,000 + 14,400 + 16,000 = 45,400 (₹ thousand, since units are in thousands).
• West: 48 × 300 + 32 × 400 + 40 × 250 = 14,400 + 12,800 + 10,000 = 37,200 (₹ thousand).
• Difference: 8,200 thousand rupees = ₹82,00,000 = 82 lakh.

Q4. For how many stores did total sales grow from 2024 to 2025 by a higher percentage than the chain as a whole?

• Chain: 2024 total = 100 + 120 + 90 + 100 = 410. Growth = (495 − 410) ÷ 410 = 85 ÷ 410 ≈ 20.73%.
• North: 20 ÷ 100 = 20%. South: 30 ÷ 120 = 25%. East: 15 ÷ 90 ≈ 16.67%. West: 20 ÷ 100 = 20%.
• Answer: 1 (only South). North and West at exactly 20% are the trap: they look like "about 20%", but 20% is below 20.73%.

Q5. Which store had the highest 2025 revenue?

• North: 12,600 + 11,200 + 12,500 = 36,300. East: 10,500 + 10,000 + 11,250 = 31,750. With South at 45,400 and West at 37,200 from Q3, South is highest (all in ₹ thousand).

Percentage change, share and percentage points

Common traps

• The wrong base. "A exceeds B by what percent" uses B as the base.
• Units. The revenue question above gives thousands of rupees because the units are in thousands.
• Rounding too early. Round at the end, not at each step, when answers are close.
• Answering before filling. A question that seems to need only one row may depend on a cell you have not filled.
• "More than" versus "at least". In Q4, a store at exactly the chain's growth rate would not count.

Practice set

Use the completed table above.

1. What percentage of the chain's non-fiction sales were made at South?
2. If North's fiction sales had been 25% higher, with everything else unchanged, what share of North's sales would fiction have been?
3. What is the ratio of exam guides sold at South to fiction sold at West?
4. What was the average number of exam guides sold per store (in thousands)?

Answers:

1. About 29.8%. 36 ÷ 121 ≈ 0.2975.
2. About 40.2%. New fiction = 42 × 1.25 = 52.5; new total = 120 + 10.5 = 130.5; 52.5 ÷ 130.5 ≈ 0.4023.
3. 4 : 3. 64 : 48, dividing both by 16.
4. 49.75 thousand. 199 ÷ 4.

What to do next

• Learn the fraction table until you can write it from memory, including 3/7 and 5/8.
• Do three gap-table sets this week, filling every cell before reading the questions.
• For every percentage question you get wrong, write down what the base should have been.
• Move on to bar and line charts, and use DILR set selection to judge table sets quickly in mocks.