QCE Vault / Specialist Maths Statistical inference practice QCE Specialist Maths · Original practice questions with worked solutions
Browse all questions Complex numbers practice Vectors practice Proof and induction practice Matrices and models practice Integration and its applications practice Differential equations and related rates practice Motion and mechanics practice Statistical inference practice Original practice exam Practise sampling distributions, confidence intervals and exponential models. Identify whether the variable is an individual observation or a sample statistic.
Key ideas For independent observations with population standard deviation σ \sigma σ , the sample mean has standard deviation σ / n \sigma/\sqrt n σ / n . Under the stated normal approximation, a confidence interval for a population mean is x ˉ ± z σ / n \bar x\pm z\sigma/\sqrt n x ˉ ± z σ / n . Check the assumptions for the model used. An exponential density with rate λ > 0 \lambda>0 λ > 0 is f ( x ) = λ e − λ x f(x)=\lambda e^{-\lambda x} f ( x ) = λ e − λ x for x ≥ 0 x\ge0 x ≥ 0 , and its mean is 1 / λ 1/\lambda 1/ λ . Worked example If σ = 12 \sigma=12 σ = 12 and n = 36 n=36 n = 36 , the standard error of the mean is 12 / 36 = 2 12/\sqrt{36}=2 12/ 36 = 2 . A common mistake A 95% confidence procedure captures the fixed population mean in about 95% of repeated samples; it is not a probability statement about a fixed interval after sampling.
Try these questions Attempt each question before revealing the worked solution. Saved questions and checked answers also appear in the main bank on this device. Exam marks are recorded after marking a practice exam.
Q5 · Practice question Simple familiar 1 mark
QUESTION 5 A confidence interval for a population mean is calculated from a sample of size 64.
The population standard deviation and the z z z -value are unchanged. To halve the width of the confidence interval, the required sample size is
Confidence interval width is proportional to
1 / n 1/\sqrt n 1/ n . Halving the width requires four times the sample size:
4 ( 64 ) = 256 4(64)=256 4 ( 64 ) = 256 .
Selects the correct option. [1 mark]
Q10 · Practice question Complex unfamiliar 6 marks
QUESTION 10 (6 marks) A sensor records a quantity X X X . A random sample of 25 readings has sample mean 102.4 and sample standard deviation 3.5.
Use a z z z -value of 2 throughout.
a) Determine an approximate confidence interval for the population mean of X X X .
[2 marks] A calibration converts every reading using Y = 0.8 X + 18 Y=0.8X+18 Y = 0.8 X + 18 .
b) Determine the sample mean and sample standard deviation of the calibrated readings.
[2 marks] c) Determine an approximate confidence interval for the population mean of Y Y Y .
[1 mark] d) A target value for the calibrated population mean is 100. Comment on whether the sample provides evidence that this target is reasonable.
[1 mark] ANSWER Calibrated interval
( 98.80 , 101.04 ) (98.80,101.04) ( 98.80 , 101.04 ) ; the target 100 is reasonable.
2 3.5 25 = 1.4 2\frac{3.5}{\sqrt{25}}=1.4 2 25 3.5 = 1.4 .
Determines the original margin of error. [1 mark]
102.4 ± 1.4 = ( 101.0 , 103.8 ) 102.4\pm1.4=(101.0,103.8) 102.4 ± 1.4 = ( 101.0 , 103.8 ) .
Determines the original confidence interval. [1 mark]
y ˉ = 0.8 ( 102.4 ) + 18 = 99.92 \bar y=0.8(102.4)+18=99.92 y ˉ = 0.8 ( 102.4 ) + 18 = 99.92 .
Transforms the sample mean. [1 mark]
Adding 18 does not change spread, so
s y = 0.8 ( 3.5 ) = 2.8 s_y=0.8(3.5)=2.8 s y = 0.8 ( 3.5 ) = 2.8 .
Transforms the sample standard deviation. [1 mark]
99.92 ± 2 2.8 5 = 99.92 ± 1.12 = ( 98.80 , 101.04 ) 99.92\pm2\frac{2.8}{5}=99.92\pm1.12=(98.80,101.04) 99.92 ± 2 5 2.8 = 99.92 ± 1.12 = ( 98.80 , 101.04 ) .
Determines the calibrated confidence interval. [1 mark]
The target 100 lies within the interval, so it is reasonable based on this sample.
Evaluates the target value. [1 mark]
Q60 · Practice question Complex familiar 4 marks
QUESTION 60 (4 marks) The lifetime T T T (hours) of a component is exponentially distributed with mean 12 hours. a) Determine P ( T > 12 ) P(T>12) P ( T > 12 ) exactly. [2 marks] b) Determine the median lifetime exactly.
[2 marks] ANSWER (b) 12 ln 2 12\ln2 12 ln 2 hours. (a) The rate is λ = 1 / 12 \lambda=1/12 λ = 1/12 . Thus P ( T > 12 ) = e − ( 1 / 12 ) ( 12 ) = e − 1 P(T>12)=e^{-(1/12)(12)}=e^{-1} P ( T > 12 ) = e − ( 1/12 ) ( 12 ) = e − 1 . (b) The median m m m satisfies P ( T ≤ m ) = 1 / 2 P(T\le m)=1/2 P ( T ≤ m ) = 1/2 . Hence 1 − e − m / 12 = 1 / 2 1-e^{-m/12}=1/2 1 − e − m /12 = 1/2 , so e − m / 12 = 1 / 2 e^{-m/12}=1/2 e − m /12 = 1/2 and m = 12 ln 2 m=12\ln2 m = 12 ln 2 hours. Determines the parameter and survival probability.
[2 marks] Forms the median equation and solves exactly.
[2 marks] All statistical inference practice questions 48 original questions · Page 1 of 2
Q5 · Original practice · 1 mark Original Specialist Mathematics practice question 5: confidence interval sample size
Statistical inference Q10 · Original practice · 6 marks Original Specialist Mathematics practice question 10: transformed confidence interval
Statistical inference Q60 · Original practice · 4 marks The lifetime T T T (hours) of a component is exponentially distributed with mean 12 hours. Exponential distributions Q67 · Original practice · 1 mark An exponential random variable has mean 5. The probability that it exceeds 8 is approximately
Exponential distributions Q69 · Original practice · 1 mark If T ∼ Exp ( 0.3 ) T\sim\operatorname{Exp}(0.3) T ∼ Exp ( 0.3 ) , the 90th percentile of T T T is approximately Exponential distributions Q91 · Original practice · 1 mark A population has mean 80 and standard deviation 15. For samples of size 25, the sample mean has mean and standard deviation
Statistical inference Q93 · Original practice · 1 mark A population has μ = 50 \mu=50 μ = 50 , σ = 10 \sigma=10 σ = 10 , and n = 100 n=100 n = 100 . Using the normal model, P ( X ˉ > 52 ) P(\bar X>52) P ( X ˉ > 52 ) is approximately Statistical inference Q95 · Original practice · 1 mark For n = 100 n=100 n = 100 , x ˉ = 20.4 \bar x=20.4 x ˉ = 20.4 , s = 5 s=5 s = 5 , and z = 1.96 z=1.96 z = 1.96 , the approximate 95% confidence interval for μ \mu μ is Statistical inference Q96 · Original practice · 6 marks A waiting time T T T (minutes) is exponentially distributed with parameter λ = 0.12 \lambda=0.12 λ = 0.12 . Exponential distributions Q97 · Original practice · 1 mark Using z = 1.96 z=1.96 z = 1.96 and planning with s = 12 s=12 s = 12 , the smallest sample size that gives margin of error at most 1.5 1.5 1.5 is Statistical inference Q98 · Original practice · 5 marks A population has mean 18 and standard deviation 4.5. Random samples of size 81 are taken.
Statistical inference Q99 · Original practice · 1 mark An approximate 95% confidence interval for a population mean is ( 48.2 , 51.8 ) (48.2,51.8) ( 48.2 , 51.8 ) . Which statement is best supported? Statistical inference Q100 · Original practice · 6 marks A sample of size 64 has x ˉ = 42.6 \bar x=42.6 x ˉ = 42.6 and s = 5.2 s=5.2 s = 5.2 . Statistical inference Q102 · Original practice · 6 marks Two independent samples from the same type of population give the following summaries. Sample n x ˉ s A 100 75.2 10 B 400 74.8 10 \begin{array}{c|ccc}\text{Sample}&n&\bar x&s\\\hline A&100&75.2&10\\B&400&74.8&10\end{array} Sample A B n 100 400 x ˉ 75.2 74.8 s 10 10 Using z = 1.96 z=1.96 z = 1.96 : Statistical inference Q104 · Original practice · 6 marks A machine-filling process is studied using a sample of 144 packets. The sample has x ˉ = 502 g \bar x=502\ \mathrm g x ˉ = 502 g and s = 12 g s=12\ \mathrm g s = 12 g . Statistical inference Q117 · Original practice · 1 mark If all else is unchanged, increasing sample size by a factor of 9 9 9 changes an approximate confidence-interval margin of error by a factor of Statistical inference Q124 · Original practice · 6 marks A 99 % 99\% 99% confidence interval for a mean is centred at 40 40 40 and has total width 5.152 5.152 5.152 . It uses z = 2.576 z=2.576 z = 2.576 and sample standard deviation s = 10 s=10 s = 10 . Statistical inference Q130 · Original practice · 1 mark A confidence interval ( 54.2 , 59.8 ) (54.2,59.8) ( 54.2 , 59.8 ) has centre and margin of error Statistical inference Q135 · Original practice · 1 mark If sample size is multiplied by 16 16 16 , confidence-interval width is approximately Statistical inference Q139 · Original practice · 1 mark For z = 1.96 , s = 8 z=1.96, s=8 z = 1.96 , s = 8 and n = 64 n=64 n = 64 , the approximate margin of error for a mean is Statistical inference Q140 · Original practice · 5 marks Two groups of sizes 40 40 40 and 60 60 60 have sample means 84 84 84 and 78 78 78 respectively. Treat the combined data as one sample with standard deviation 10 10 10 and use z = 1.96 z=1.96 z = 1.96 . Statistical inference Q147 · Original practice · 7 marks A sample of n = 144 n=144 n = 144 measurements has mean 25.6 25.6 25.6 and standard deviation 3.6 3.6 3.6 . Use z = 2.576 z=2.576 z = 2.576 for a 99 % 99\% 99% interval. Statistical inference Q150 · Original practice · 1 mark An instrument records Y = 1.2 X + 0.6 Y=1.2X+0.6 Y = 1.2 X + 0.6 , where X X X is the true measurement and Y Y Y is the recorded measurement, both in millimetres. This calibration relationship is exact. A random sample of 64 independent measurements has recorded sample mean y ‾ = 18.6 \overline y=18.6 y = 18.6 and recorded sample standard deviation s Y = 2.4 s_Y=2.4 s Y = 2.4 . Assume the conditions for an approximate normal co… Statistical inference Q171 · Original practice · 6 marks Measurements X X X are normally distributed with unknown mean μ \mu μ and known standard deviation 12 12 12 mm. Independent measurements are selected at random. A sample of size n n n has mean X ‾ \overline X X . Statistical inference: sampling precision