Mains PYQs
UPSC Mains Statistics paper 2 2018 questions
- If a linear programming problem has an optimal solution, then prove that it attains its optimum solution at an extreme point of the convex set generated by the set of all feasible solutions to the problem.
- Let $\{X_n, n \ge 0\}$ be a Markov chain having a sample space $S = \{1, 2, 3, 4\}$ and transition probability matrix
$$P = \begin{bmatrix} \frac{1}{3} & \frac{2}{3} & 0 & 0 \\ 1 & 0 & 0 & 0 \\ \frac{1}{2} & 0 & \frac{1}{2} & 0 \\ 0 & 0 & \frac{1}{2} & \frac{1}{2} \end{bmatrix}$$
Show that states 3 and 4 are transient states.
- Explain the theoretical basis for control charts.
- Describe the behaviour of the hazard rate in a bathtub curve.
- Find the range of values for $p$ and $q$ that will render the cell (2, 2) a saddle point in the game with the following pay-off matrix :
Player B (columns), Player A (rows)
$$\begin{bmatrix} 1 & q & 6 \\ p & 5 & 10 \\ 6 & 2 & 3 \end{bmatrix}$$
- Define a series system and a parallel system. Compute the reliability of the following configuration whose components work independently :
[Figure]
It is given that the components $c_1$, $c_2$, $c_3$, $c_4$ and $c_5$ have component reliabilities 0.6, 0.4, 0.7, 0.8 and 0.5 respectively.
- Define failure rate and reliability function of a random variable $T$. Obtain the failure rate and reliability function of the Weibull failure model. State the conditions under which the failure rate is increasing, decreasing and constant.
- Discuss a transportation problem (TP) and show that it can be formulated as a linear programming problem. Obtain a necessary and sufficient condition under which a TP has always a feasible solution. Find the first feasible solution of the following TP using Vogel's approximation method (VAM) :
| Port \ Destination | D1 | D2 | D3 | D4 | D5 | D6 | Requirement |
|---|---|---|---|---|---|---|---|
| 1 | 5 | 3 | 7 | 3 | 8 | 5 | 3 |
| 2 | 5 | 6 | 12 | 5 | 7 | 11 | 4 |
| 3 | 2 | 8 | 3 | 4 | 8 | 2 | 2 |
| 4 | 9 | 6 | 10 | 5 | 10 | 9 | 8 |
| Availability | 3 | 3 | 6 | 2 | 1 | 2 | |
Discuss whether the solution is degenerate or non-degenerate.
- Stating the underlying assumptions, find the expression of the probability of happening $n$ events in the time interval $(0, t)$ in Poisson process. Hence obtain the distribution of inter-arrival times.
- Describe Monte Carlo simulation technique and mention its different steps. Also describe how you would generate random numbers from a Weibull density with parameters $(\theta, \beta)$.
- Control charts for $\bar{x}$ and $R$ are maintained for an important quality characteristic. The sample size is $n = 7$; $\bar{x}$ and $R$ are computed for each sample. After 35 samples, we have found that
$$\sum_{i=1}^{35} \bar{x}_i = 7805 \quad \text{and} \quad \sum_{i=1}^{35} R_i = 1200$$
(i) Compute the central line and control limits for $\bar{x}$ and $R$ charts.
(ii) Assuming both charts exhibit control, estimate the process mean and standard deviation.
(iii) If the quality characteristic is normally distributed and if the specification limits are $220 \pm 35$, estimate $C_p$ and $C_{pk}$.
(Given : $D_3 = 0.76$, $D_4 = 1.924$, $d_2 = 2.704$, $A_2 = 0.419$)
- The following figures give the number of defectives in 20 samples each containing 2000 items :
305, 425, 430, 216, 341, 225, 322, 280, 306, 337, 356, 402, 216, 264, 126, 409, 193, 326, 280, 390
Can we conclude that the process is in control by setting up an appropriate control chart in a graph sheet?
- What is the importance of sensitivity analysis? Mention what type of problems can be resolved using it.
Using the following linear programming problem, show that the value 6 of the resource vector $b = \mathrm{col.}\{6\ 8\ 1\ 2\}$ can be increased to 7 :
Maximize $Z = 3x_1 + 2x_2$
subject to
$x_1 + 2x_2 \le 6$
$2x_1 + x_2 \le 8$
$-x_1 + x_2 \le 1$
$x_2 \le 2$
$x_1, x_2 \ge 0$
- In a single sampling plan (1000, 89, 2), compute the probability of acceptance of the lot if the lot fraction defective is 0.01. Also compute Average Outgoing Quality (AOQ) of the lot.
- Distinguish between OLS and GLS estimators. Show that OLS estimators are BLUEs.
- Describe registration and census methods of demographic data collection.
- What is time series? Discuss various components of a time series and their significances in contributing variation in the model.
- What are the importances of index numbers? Define the different tests an index number has to satisfy. Which index numbers satisfy these tests?
- Discuss the importance of scaling in scores in psychometry. Describe any two scaling methods.
- Given the following demand-and-supply model
Demand : $Q_t = \alpha_0 + \alpha_1 P_t + \alpha_2 I_t + u_{1t}$; $\alpha_1 < 0$, $\alpha_2 > 0$
Supply : $Q_t = \beta_0 + \beta_1 P_t + \beta_2 P_{t-1} + u_{2t}$; $\beta_1 > 0$, $\beta_2 > 0$
construct the reduced-form model and justify whether the parameters can be estimated completely.
- Distinguish between wholesale price index number and consumer price index number. How are they calculated? The following data relate to the wholesale prices of six different pulses in June and December of a year :
| Pulses | Weight | Price in Rs, June | Price in Rs, December |
|---|---|---|---|
| 1 | 1.26 | 120 | 135 |
| 2 | 0.35 | 135 | 156 |
| 3 | 1.79 | 148 | 162 |
| 4 | 1.88 | 132 | 148 |
| 5 | 0.98 | 146 | 158 |
| 6 | 1.12 | 128 | 150 |
Calculate the index number for December with June as the base by weighted aggregative method.
- Discuss the concept of reliability of a test. Define reliability ratio of a test and show that it is always between 0 and 1. What is the index of reliability? Is it also bounded by 0 and 1?
- Calculate (i) crude death rate and (ii) standardised death rates of Town A from the following data :
| Age Group (in years) | Population of Town A | Deaths in Town A | Standard Population |
|---|---|---|---|
| 0-5 | 2000 | 100 | 20000 |
| 5-10 | 1600 | 48 | 16000 |
| 10-25 | 2400 | 24 | 24000 |
| 25-45 | 6000 | 30 | 60000 |
| 45 and above | 8000 | 104 | 80000 |
Are the two rates equal? If so, why?
- Suppose there are four items $A$, $B$, $C$ and $D$ respectively passed by 90%, 80%, 70% and 60% of individuals. Compare the differences in difficulty between $A$ and $B$ with the differences in difficulty between $C$ and $D$.
- Consider a two-variable regression model $Y_t = \beta_1 + \beta_2 X_t + u_t$. Suppose $u_t$ is autocorrelated as $u_t = \rho u_{t-1} + \varepsilon_t$, $-1 < \rho < 1$, where $\rho$ is the autocovariance coefficient and $\varepsilon_t$ is the stochastic disturbance term satisfying OLS assumptions. Obtain an estimate for $\beta_2$ and its corresponding variance. Also discuss the Durbin-Watson statistic for detecting autocorrelation in the model, and its interpretation.
- Describe the Box-Jenkins methodology to identify a particular time series model.
- What is a life table? Distinguish between a complete life table and an abridged life table. Complete the following life table :
| $x$ | $l_x$ | $d_x$ | $q_x$ | $L_x$ | $T_x$ | $e_x^0$ |
|---|---|---|---|---|---|---|
| 83 | 3560 | ? | 0.16 | ? | ? | ? |
| 84 | ? | ? | 0.17 | ? | 11975 | ? |
- Write a brief note on National Statistical Commission stating the components, functions and objectives of each.