- Foundation
- About 12 min
- 5 practice questions
Building time sense at home: estimate first, then measure
Train your feeling for how long a question takes by writing an estimate before solving, comparing it with the stopwatch, and turning your average into a pace card for the hall.
If something looks wrong, please tell us.
Most students feel they have a good sense of time until they measure it. Time sense, the feeling for how long a task will take, is a trainable skill. In an exam it lets you decide in a second whether a question fits your budget. The training is simple: estimate first, measure second, compare. This lesson shows the routine and what to do with the numbers.
The estimate-then-measure routine
- Before solving a practice question, write an estimate of the seconds it will take.
- Start a timer, solve, stop the timer; write the actual time.
- Note the difference. A positive difference means you underestimated.
- After ten questions, add up the estimates and the actual times and compare the totals.
Do this for a week or two, ten questions a day, using practice that you would do anyway. The extra cost is a few seconds per question to write the numbers.
Worked example 1: Isha's estimates
| Question | Estimate (s) | Actual (s) | Difference |
|---|---|---|---|
| 1 | 30 | 38 | +8 |
| 2 | 45 | 41 | -4 |
| 3 | 60 | 95 | +35 |
| 4 | 40 | 52 | +12 |
| 5 | 90 | 120 | +30 |
| Total | 265 | 346 | +81 |
Her estimates add up to 265 seconds, her actual times to 346 seconds. She underestimated on four of the five questions. The shortfall is 346 - 265 = 81 seconds, which is 81 / 346 = about 23% of the actual time. The ratio is 346 / 265 = 1.31, so a rule of thumb for her is to multiply her gut estimate by about 1.3. Applying it to the estimates gives 39, 58, 78, 52 and 117 seconds, which are closer to the real numbers, though still about 17 seconds off on two of them (questions 2 and 3). Isha now knows that a question she feels will take a minute will probably take about 78 seconds.
The estimate matters most for the questions that run long. In question 3 she was 35 seconds out. These are the ones to watch: when a question looks as if it will take longer than her budget, her real time may be longer still.
Building a pace card
A pace card tells you how many questions to finish in a fixed block of time. Measure your normal pace on a calm set, repeat it, and take the average.
Worked example 2: a pace card
Rohan times three sets of 10 questions of mixed types: 6 minutes 5 seconds, 5 minutes 40 seconds and 5 minutes 55 seconds. In seconds these are 365, 340 and 355, and their total is 1,060. The average is 1,060 / 3 = 353 seconds per set, or about 35 seconds a question. He rounds to 35 and writes his pace card: 10 questions in about 6 minutes (350 seconds is 5 minutes 50 seconds). In the hall he can check himself against the card every 10 questions. If the budget for the paper is lower than his pace, he knows from the card that he cannot attempt every question at his accurate pace and must choose.
Estimate, measure, compare. Correct your gut feeling with a number from your own data, and make a pace card from your average.
Making it stick
- Keep the estimates honest: write them before you start, not after.
- Repeat after a month; your pace and your feeling both change.
- Use calm practice for the pace card, not a rushed set.
Key takeaways
- Time sense is a skill: estimate first, then measure.
- Compare totals to find your correction factor.
- A pace card turns your average into a check for the hall.
- Re-measure after a few weeks.
Check your understanding
-
1. Isha estimated 265 seconds in all and the actual time was 346. By how many seconds did she underestimate?
- 46
- 265
- 611
- 81
Show answer
Answer: D. 346 - 265 = 81 seconds.
-
2. A student's gut estimates average 40 seconds but the real average is 52. What is the approximate correction factor?
- 0.77
- 1.3
- 1.5
- 12
Show answer
Answer: B. 52 / 40 = 1.3, so multiply gut feelings by about 1.3.
-
3. Three sets of 10 questions take 360, 330 and 390 seconds. What is the average seconds per question?
- 33
- 39
- 36
- 360
Show answer
Answer: C. The total is 1,080 seconds over 30 questions: 1,080 / 30 = 36.
-
4. When should the time estimate be written?
- Before you start solving the question
- Straight after you have solved it
- Only on the days when the answer is wrong
- At the end of the week, from memory
Show answer
Answer: A. An estimate written afterwards is influenced by what you know, so it does not train your time sense.
-
5. At 35 seconds a question, about how many questions can you finish in 7 minutes?
- 10
- 14
- 12
- 20
Show answer
Answer: C. 7 minutes = 420 seconds; 420 / 35 = 12.
Sign in to keep track of the lessons you finish.