GMAT Data Insights lesson · Determinable
GMAT Data Insights: what a table can and can't determine
Deciding on GMAT Data Insights whether a quantity can be found from a table, and whether new information would help.
The core idea
List the numbers a quantity needs. If every one is in the table (or follows from columns that are), it can be determined. New information helps only if it supplies the one missing number.
Core ideas
Write the quantity as a formula in words: "parcels late = parcels delivered × (100% − on-time %)". Then tick off each ingredient against the column headers.
These questions don't ask you to compute the answer, only whether it could be computed. So the work is a checklist, not arithmetic. Two delivery depots:
| Depot | Parcels delivered | On time |
|---|---|---|
| North | 2,400 | 85% |
| South | 1,600 | 90% |
| Quantity | Formula | Ingredients in the table? | Determined? |
|---|---|---|---|
| Late parcels at North | 2,400 × (100% − 85%) | both | yes: 360 |
| On-time rate for both depots | (2,400 × 85% + 1,600 × 90%) ÷ 4,000 | all | yes: 87% |
| Average weight of North's parcels | total weight ÷ 2,400 | weight missing | no |
- Write the quantity as a formula in words.
- List every number the formula needs.
- Tick each one against the column headers. An ingredient also counts if other columns give it (a count from a total and a percent).
- If every ingredient is ticked, the quantity can be determined. If even one is missing, it can't.
Note that you never need the final value. The two "yes" answers above (360 late parcels, 87%) are worked out only to show the formula really closes.
Can it be determined?
Why it works. A quantity is a formula of other numbers. If every input is known, the formula gives exactly one value. If one input is missing, different values of it fit the same table and give different answers, so no single answer exists.
How to spot it: a table and statements labelled "Can be determined" / "Cannot be determined".
Example. A factory table:
| Factory | Units made | Workers | Energy used (MWh) | Price per unit ($) |
|---|---|---|---|---|
| Ardmore | 12,000 | 40 | 300 | 25 |
| Bellton | 8,000 | 32 | 240 | 30 |
| Corrin | 10,000 | 25 | 260 | 28 |
Which of these can be determined from the table?
| Quantity | Needs | Missing? | Verdict |
|---|---|---|---|
| Defective units at Ardmore | units made × defect % | defect % | cannot |
| Units per worker at Corrin | units ÷ workers | nothing | can (400) |
| Average energy per unit, all factories | total energy ÷ total units | nothing | can |
- Defective units need a defect percent. No column gives it, and no columns combine to give it: cannot be determined.
- Units per worker at Corrin: 10,000 ÷ 25 = 400. Both are columns: can.
- Average energy per unit across all factories needs total energy (300 + 240 + 260 = 800 MWh) and total units (30,000). Both follow from the columns: can (800 MWh ÷ 30,000 ≈ 26.7 kWh per unit).
Trap: assuming a number you'd expect to exist. A factory table "feels" like it has defects, but it doesn't. The opposite slip is missing a derivation: a quantity needn't be a column, so "total units" is available even though no row shows it.
A second slip on the last line: the mean of the three factories' own rates (25, 30 and 26 kWh per unit) is 27, not 26.7. The table can determine both numbers, but they answer different questions.
Try it. Three bookshops last month:
| Shop | Books sold | Revenue ($) | Staff |
|---|---|---|---|
| Alder | 1,200 | 18,000 | 4 |
| Birch | 900 | 15,300 | 3 |
| Cedar | 1,500 | 21,000 | 5 |
Can each be determined? (a) The average price per book at Birch. (b) Cedar's profit. (c) Revenue per staff member for the three shops together.
Show the answer
- (a) Price per book = revenue ÷ books = 15,300 ÷ 900 = $17. Both are columns: can.
- (b) Profit = revenue − costs. No column gives costs. With costs of $15,000 the profit is $6,000; with $18,000 it is $3,000. Both fit the table: cannot.
- (c) Total revenue ÷ total staff = 54,300 ÷ 12 = $4,525. Both totals follow from the columns: can.
Edge cases.
- A comparison can be determined when the values can't. If two shops pay the same unknown cost per book, the one that sold more books spent more, even though neither amount is known.
- A range is not a value. Knowing a number lies between 70% and 90% doesn't determine it, unless the question only asks whether it is above 60%.
- Watch the unit of each column. "Revenue per book" and "revenue" sound alike, but only one can be added across rows.
GMAT-level example. Three clinics last week:
| Clinic | Visits | New patients (% of visits) | Average visit (min) |
|---|---|---|---|
| Holt | 1,500 | 40% | 20 |
| Ivy | 1,200 | 25% | 24 |
| Juno | 800 | 50% | 15 |
Which can be determined? (1) Visits by returning patients at Ivy. (2) Total visit minutes at the three clinics. (3) The average length of a new patient's visit at Holt. (4) The percent of all visits that were by new patients.
- (1) Returning = 1,200 × (100% − 25%) = 900: can.
- (2) Each clinic's minutes = visits × average: 1,500 × 20 = 30,000; 1,200 × 24 = 28,800; 800 × 15 = 12,000. Total 70,800: can.
- (3) Holt had 600 new and 900 returning visits, and 30,000 minutes in all. But the table doesn't split the minutes. New visits averaging 20 minutes (returning 20) fits; so does 26 minutes (returning 16), since 600 × 26 + 900 × 16 = 30,000. Two answers: cannot.
- (4) New visits: 600 + 300 + 400 = 1,300, out of 3,500. 1,300 ÷ 3,500 ≈ 37.1%: can.
✓ Check: (2) and (4) use only columns and their products; (3) needs a column the table lacks.
Check yourself. From the clinic table, can the average visit length across all three clinics be determined?
Check your answer
Yes. Total minutes ÷ total visits = 70,800 ÷ 3,500 ≈ 20.2 minutes. (The plain mean of 20, 24 and 15 is about 19.7, which answers a different question.)
Would it help?
Why it works. New information helps only if it changes the answer from "can't" to "can". If the quantity is already fixed, nothing changes. If another ingredient is still missing, different values still fit, so the answer is still "can't".
How to spot it: statements labelled "Would help" / "Would not help", each adding one piece of information.
New data helps only if the quantity can't be found now and can be found with it. Three outcomes: it supplies the missing number (helps), the quantity was already determined (doesn't help), or another number is still missing (doesn't help).
Example. Use the depot table from Core ideas. Would each piece of information help find the quantity?
| Quantity | New information | Outcome |
|---|---|---|
| Late parcels at North | North's late parcels were counted by hand | already determined (360): would not help |
| Average weight of North's parcels | total weight of North's parcels | supplies the missing number: would help |
| Fuel cost per parcel at South | litres of fuel South used | price per litre still missing: would not help |
- Ask first: can the quantity be found now? Late parcels at North can (2,400 × 15% = 360), so nothing new helps.
- If not, does the new information supply every missing ingredient? Total weight ÷ 2,400 closes the formula.
- Fuel cost needs litres and price per litre. Litres alone leave a gap.
Trap: thinking any relevant data helps. Extra data on a quantity the table already determines adds nothing, and data that fills one gap of two still leaves the quantity unknown.
Try it. Two farms:
| Farm | Area (hectares) | Wheat yield (tonnes per hectare) |
|---|---|---|
| Kerr | 120 | 3.5 |
| Lyle | 80 | 4.0 |
Would each piece of information help? (a) Kerr's wheat revenue, given the price per tonne Kerr was paid. (b) The two farms' combined wheat output, given that Lyle's harvest weighed 320 tonnes at the mill. (c) Wheat per worker at Lyle, given the number of workers at Kerr.
Show the answer
- (a) Revenue = tonnes × price. Tonnes = 120 × 3.5 = 420, but the price is missing: at $200 a tonne revenue is $84,000, at $250 it is $105,000. The price fills the only gap: would help.
- (b) Combined output is already 420 + 80 × 4.0 = 420 + 320 = 740 tonnes: would not help.
- (c) Wheat per worker at Lyle needs Lyle's workers. Kerr's workers leave that gap open: would not help.
Edge cases.
- A link can fill a gap. "Kerr was paid $40 a tonne more than Lyle" helps find Kerr's revenue only if Lyle's price is already known.
- Narrowing is not enough. Information that only squeezes the quantity into a range doesn't help; it must fix one value.
- Judge each statement alone unless the question says to use them together.
GMAT-level example. An online shop's sales last month:
| Product | Units sold | Price ($) |
|---|---|---|
| Lamp | 240 | 45 |
| Rug | 150 | 85 |
| Vase | 320 | 30 |
The quantity is the profit on rugs. Would each statement help? (1) Each rug cost the shop 60% of its price. (2) The cost of all the goods sold was $20,000. (3) The shop made $34 on each rug.
- Write the formula: rug profit = units × (price − cost per rug). Units and price are in the table; cost per rug is the gap.
- (1) Cost = 60% × 85 = $51, so each rug makes 85 − 51 = $34, and profit = 150 × 34 = $5,100: would help.
- (2) Rug revenue is 150 × 85 = $12,750. If rugs cost $7,650 of the $20,000, the profit is $5,100; if they cost $9,000, it is $3,750. Both splits add to $20,000: would not help.
- (3) Profit = 150 × 34 = $5,100: would help.
✓ Check: (1) and (3) both give $5,100, because 60% of $85 is $51 and $85 − $51 = $34.
Check yourself. Would the shop's total revenue from all three products help you find the revenue from rugs?
Check your answer
Would not help. Rug revenue is already 150 × 85 = $12,750 from the table.
Combined quantities
Why it works. A combined rate is a total part divided by a total whole, and each row's part is its rate times its size. Without the sizes, the big rows could be the high-rate rows or the low-rate rows. So the combined rate can land anywhere between the lowest and the highest rate.
How to spot it: a table of rates only, and a question about a rate for all the rows together.
A combined rate across rows ("percent of all the depots' parcels delivered on time") needs each row's count as well as its rate, because the rows are weighted by size. Knowing each depot's parcel count would help; knowing only the rates wouldn't.
Example. A second company's table shows only on-time rates: East 90%, West 70%. Can the on-time rate for both depots together be determined?
Try two sets of counts that fit the same rates:
| East parcels | West parcels | On time | Combined rate | |
|---|---|---|---|---|
| Case 1 | 300 | 100 | 270 + 70 = 340 | 340 ÷ 400 = 85% |
| Case 2 | 100 | 300 | 90 + 210 = 300 | 300 ÷ 400 = 75% |
- Same rates, two different answers: the combined rate cannot be determined.
- Each depot's parcel count would fix the weights: that information would help.
- The ratio of the counts alone would help too, because it fixes the weights.
Trap: answering 80%, the plain average of the rates. That's right only if the two depots handled the same number of parcels (200 each gives 80%). Knowing only another rate, or the overall total of parcels, still leaves the weights unknown.
Try it. School A has a 60% pass rate and School B an 80% pass rate. Would knowing that School A has 3 times as many students as School B help find the combined pass rate?
Show the answer
- The ratio fixes the weights: 3 parts A to 1 part B.
- Combined rate = (3 × 60 + 1 × 80) ÷ 4 = 260 ÷ 4 = 65%.
- Check with 300 and 100 students: 180 + 80 = 260 passes out of 400, which is 65%. So it would help.
Edge cases.
- Equal rates need no counts. If both rows have a 75% rate, the combined rate is 75% whatever the sizes.
- Bounds can answer a yes/no question. The combined rate always lies between the lowest and highest row rates, so "is it above 65%?" with rates of 70% and 90% is yes without any counts.
- One count alone helps only with the total. One row's count plus the overall count gives the other row's count by subtraction.
GMAT-level example. A retailer's return rates at three stores. Ames sold 2,000 items and Cray 1,000; Bow's sales are not given.
Would each statement help find the return rate for all three stores together? (1) Bow sold 3,000 items. (2) The three stores sold 6,000 items in all. (3) Bow sold more items than Cray.
- Known returns: Ames 4% of 2,000 = 80; Cray 12% of 1,000 = 120.
- (1) Bow returns 6% of 3,000 = 180. Combined = (80 + 180 + 120) ÷ 6,000 = 380 ÷ 6,000 ≈ 6.3%: would help.
- (2) Bow = 6,000 − 2,000 − 1,000 = 3,000, the same as (1): would help.
- (3) Bow could have sold 3,000 (≈ 6.3%) or 1,500: then returns are 80 + 90 + 120 = 290 out of 4,500 ≈ 6.4%. Both fit: would not help.
✓ Check: both possible answers lie between the lowest rate (4%) and the highest (12%), as they must.
Check yourself. Two warehouses ship 92% and 96% of their orders on time. Without any order counts, can you tell whether the combined on-time rate is above 90%?
Check your answer
Yes. The combined rate must lie between 92% and 96%, so it is above 90% whatever the sizes. The exact rate can't be determined, but the yes/no question can.
How hard questions disguise it
For harder questions (L4–L5)
At L4 and L5, the missing ingredient is hidden. A table of rates looks complete until you see that the question needs a weighted average, which needs sizes. A part-to-whole or ratio statement can quietly supply a size you thought was missing, and a percent change needs the starting value as well as the percent. Hard "would it help" items mix all three outcomes: one statement fills the gap, one repeats what the table already gives, and one fills only half of it. Write the formula, mark each gap, and test a doubtful case with two made-up values that fit the table.
Practice
Practice "can it be determined" and "would it help" questions in timed sets on the determinable practice page.
Common traps
Assuming a missing number
Revenue needs a price; late deliveries need a count and an on-time percent. If the table lacks one, the quantity can't be found.
Missing a derivation
A quantity needn't be a column: parcels delivered late follow from parcels delivered and the on-time percent.
Thinking extra data helps when it doesn't
If the table already determines the quantity, more data adds nothing: 'Would not help'.