The distribution function (df) is also known as cumulative distribution function (cdf).
- AYes
- BNo
F(x) = P(X ≤ x) is called the distribution function or cumulative distribution function.
Syllabus: lectures 22–45. Past-paper MCQs are shown first.Syllabus: lecture 22–45. Past paper ki MCQs pehle dikhai gayi hain.
384 MCQs · 130 short · 35 long questions · 384 MCQs from past papers384 MCQs · 130 short · 35 long sawal · 384 MCQs past papers se
The distribution function (df) is also known as cumulative distribution function (cdf).
F(x) = P(X ≤ x) is called the distribution function or cumulative distribution function.
The function F(x) gives the probability of the event that X takes a value ......
F(x) = P(X ≤ x).
The distribution function (df) is also known as
F(x) = P(X ≤ x) is the cumulative distribution function.
In a discrete distribution function, F(23) can be stated as:
F(23) = P(X ≤ 23), i.e. at most 23 successes.
In a discrete distribution function, F(10) can be stated as:
F(10) = P(X ≤ 10), at most 10 successes.
The distribution function F(x) is equal to
The (cumulative) distribution function is F(x) = P(X ≤ x).
When the random variable X and Y are independent, its co-variance is:
Cov(X, Y) = E(XY) − E(X)E(Y) = 0 when X and Y are independent.
When the random variables X and Y are independent, then its co-variance must be:
Independence gives E(XY) = E(X)E(Y), so Cov(X, Y) = 0.
The covariance of a random variable with itself is:
Cov(X, X) = E[(X − μ)²] = Var(X).
Which of the following is true for the binomial distribution b(x: n, p):
Mean np exceeds variance npq because q < 1.
The parameters of the binomial distribution b(x; n, p) are:
The binomial distribution is fully determined by the number of trials n and success probability p.
The range of the binomial distribution is:
In n trials the number of successes can be 0, 1, 2, …, n.
Let X be a random variable with binomial distribution, that is (X=0,1,…, n). The expected value E[X] is
For a binomial variable E(X) = np.
Suppose 60% of a herd of cattle is infected with a particular disease. Let Y = the number of non-diseased cattle in a sample of size 5. The distribution of Y is:
Success = non-diseased with p = 1 − 0.6 = 0.4, in n = 5 trials.
Binomial distribution is skewed to the right if:
When p < q (p < 0.5) the binomial distribution is positively (right) skewed.
If b(x, 7, 0.30), the variance of this distribution is:
npq = 7 × 0.30 × 0.70 = 1.47.
Which of the following is a characteristic of a binomial probability experiment?
Binomial trials are independent with constant p.
The binomial distribution is negatively skewed when:
When p > q the binomial distribution is negatively (left) skewed.
If b(x, 15, 0.70), the variance of this distribution is
Var = npq = 15 × 0.70 × 0.30 = 3.15.
Let X be a random variable with binomial distribution, that is (x=0,1,…, n). The Var[X] is:
Var(X) = npq = np(1 − p).
The probability that a certain machine will produce a defective item is 1/4. If a random sample of 6 items is taken from the output of this machine, what is the probability that there will be 5 or more defective in the sample.
P(X ≥ 5) = 6C5(1/4)⁵(3/4) + (1/4)⁶ = 18/4096 + 1/4096 = 19/4096.
For a binomial distribution, n = 10 & q = 0.6, the mean of the distribution is:
p = 1 − 0.6 = 0.4, mean = np = 10 × 0.4 = 4.
In the FA examination, 24 candidates offered Statistics. If the probability of passing the subject be 1/3, what will be the mean of the distribution?
Mean = np = 24 × 1/3 = 8.
In a binomial experiment the total number of trials are:
A binomial experiment has a fixed number n of independent trials decided in advance.
Which of the following is true for the Poisson distribution:
For the Poisson distribution E(X) = Var(X) = μ.
A uniform distribution is defined by:
The continuous uniform distribution on [a, b] is fixed by its smallest value a and largest value b.
The number of telephone calls that pass through a switchboard has a Poisson distribution with mean equal to 2 per minute. The probability that no telephone calls pass through the switchboard in two consecutive minutes is:
For two minutes μ = 2 × 2 = 4, so P(X = 0) = e^(−4) = 0.0183.
For good approximation of Poisson distribution to the binomial distribution, which of the following condition(s) is/are required:
The Poisson approximates the binomial when n is large and p is small.
If p is very small and n is considerably large then we shall apply the:
For small p and large n the binomial is approximated by the Poisson with μ = np.
Which of the following is NOT applicable to a Poisson distribution?
Poisson needs small p and large n; p = 0.5 with n = 19 fails both.
The number of parameters in uniform distribution is (are):
The uniform distribution on [a, b] has two parameters a and b.
The relationship between the mean and variance of Poisson distribution is:
For the Poisson distribution mean = variance = μ.
We use the Poisson approximation to the binomial when:
The usual rule is p ≤ 0.05 with n ≥ 20.
A uniform distribution is defined by:
The uniform distribution on [a, b] is fixed by its smallest and largest values.
For the given poisson distribution P(X = 1) = e^(−0.135) 0.135¹ / 1! the mean value is :
In P(X = x) = e^(−λ)λˣ/x! the mean is λ, here 0.135.
In normal distribution Q.D =
For a normal distribution Q.D ≈ 0.6745σ (Q1, Q3 = μ ∓ 0.6745σ).
The area under a normal curve between 0 and -1.75 is
By symmetry the area from −1.75 to 0 equals the table area from 0 to 1.75 = 0.4599.
In normal distribution M.D. =
For a normal distribution M.D = √(2/π)σ ≈ 0.7979σ.
Which of the following is NOT CORRECT about a standard normal distribution?
P(Z ≥ −2.5) = 0.5 + 0.4938 = 0.9938, not 0.4938; the other three values are correct.
The total area under the normal curve is:
The normal curve is a probability density, so its total area is 1.
The moment ratios of normal distribution come out to be:
For the normal distribution β1 = 0 and β2 = 3.
Suppose the test scores of 600 students are normally distributed with a mean of 76 and standard deviation of 8. The number of students scoring between 70 and 82 is:
z = (70 − 76)/8 = −0.75 and (82 − 76)/8 = 0.75; P = 2 × 0.2734 = 0.5468; 0.5468 × 600 ≈ 328.
The mean deviation of the normal distribution is approximately:
M.D = 0.7979σ ≈ (4/5)σ.
Suppose the test scores of 600 students are normally distributed with a mean 76 and standard deviation 8. The no. of students scoring between 70 and 82 is
z = ±0.75; P = 2 × 0.2734 = 0.5468; 0.5468 × 600 ≈ 328.
Which of the following is a characteristic of the normal distribution:
The normal curve is symmetric and bell-shaped (mesokurtic, asymptotic to the x-axis).
In normal distribution β₁ = ........ :
The normal distribution is symmetric, so β₁ = μ₃²/μ₂³ = 0.
In normal distribution, the quartile deviation Q.D = .................... :
For a normal distribution Q.D = 0.6745σ (≈ 2σ/3).
As a rule of thumb, when n >= 30, then we can assume that………is normally distributed:
By the central limit theorem the sampling distribution of x̄ is approximately normal for n ≥ 30.
The Central Limit Theorem is important in Statistics because it allows us to use the normal distribution to make inferences concerning the population mean:
The CLT gives an approximately normal x̄ for a reasonably large n, whatever the population shape.
The Maximum Likelihood Estimators (MLE) are ………… and …………… but not necessarily ………
ML estimators are consistent and (asymptotically) efficient but may be biased.
In moments method, how many equations are needed to find the 2 unknown parameters?
One moment equation is needed per unknown parameter, so 2 equations.
Which of the following is most important and most widely used method in point estimation?
Maximum likelihood is the most important and widely used method of point estimation.
"To consider every possible value that the parameter might have, and for each value, compute the probability and THAT value of the parameter for which the probability of a given sample is greatest, is chosen as an estimate." This procedure is known as:
Choosing the parameter value that maximizes the probability (likelihood) of the observed sample is maximum likelihood.
In moments method, how many equations are needed to find the 2 unknown parameters?
One moment equation per unknown parameter: 2 equations.
ML estimators may not be ………
Maximum likelihood estimators are consistent and efficient but not necessarily unbiased.
Which one of the formula will be used to find out the confidence interval for μ, when population variance unknown and sample size is large?
For large n, s replaces σ and the normal z value is used: x̄ ± z_(α/2) s/√n.
Which one is the correct formula to find the desired sample size?
From e = Zα/2·σ/√n we get n = (Zα/2·σ/e)².
For testing of hypothesis about population proportion, we use:
For large samples Z = (p̂ − p0)/√(p0q0/n) is used to test a proportion.
The test statistic to test the μ1 = μ2 (μ represents the mean of population) for normal population for n > 30.
For large samples the difference of means is tested with Z.
As the degree of freedom increases, the t-distribution tends to coincide with:
As ν → ∞ the t-distribution approaches the standard normal distribution.
The Chi-Square distribution is continuous distribution ranging from:
χ² is a sum of squares, so it takes values from 0 to ∞.
The value of χ2 can never be:
χ² is a sum of squared quantities, so it is never negative.
The value of chi square can never be:
χ² is a sum of squares and cannot be negative.
The parameter of the chi-square distribution is.......
The chi-square distribution has a single parameter, its degrees of freedom ν.
If mean of χ² distribution is k then variance will be:
For χ² with ν d.f., mean = ν and variance = 2ν, so mean k gives variance 2k.
What is the graphical shape of the chi-square distribution?
The χ² curve starts at 0 and has a long right tail, so it is positively skewed.
The Chi-Square distribution is continuous distribution ranging from:
χ² is a sum of squares, so it cannot be negative and ranges from 0 to ∞.
The F-distribution always ranges from:
F is a ratio of two non-negative chi-square variables, so 0 ≤ F < ∞.
To find the confidence interval for the ratio of two variances we use:
(S1²/σ1²)/(S2²/σ2²) follows the F-distribution, which gives the interval for σ1²/σ2².
The F-distribution has …… parameter.
F has two parameters, the numerator and denominator degrees of freedom ν1 and ν2.
The curve of the F-distribution depends upon:
The shape of F is fixed by its two degrees of freedom ν1 and ν2.
F-distribution is a…………….. distribution.
The F curve has a single peak (it is positively skewed and continuous).
F-distribution tends to normality, if
F approaches normality as both degrees of freedom ν1 and ν2 become large.
The curve of the F-distribution depends upon:
The F curve depends on its degrees of freedom, which are fixed by the sample sizes.
When we want to test the equality of two variances we usually use
H0: σ1² = σ2² is tested with F = S1²/S2².
The curve of the F- distribution depends upon:
The shape of the F-curve is determined by its two degrees of freedom ν1 and ν2.
Mean of the F-distribution is possible only, when
Mean of F is ν2/(ν2 − 2), which exists only when ν2 > 2.
F-distribution is ..........
The F curve is non-negative with a long right tail, so it is positively skewed.
ANOVA was introduced by:
Analysis of variance was developed by Sir R.A. Fisher.
The test statistic used in analysis of variance procedure follow the ....... distribution:
ANOVA uses F = MS(treatment)/MS(error), which follows the F-distribution.
Analysis of variance is a procedure that enables us to test the equality of several:
ANOVA tests H0: μ1 = μ2 = … = μk.
The analysis of variance technique is a method for:
ANOVA tests the equality of three or more population means.
In an ANOVA test there are 5 observations in each of three treatments. The degrees of freedom in the numerator and denominator respectively are.......
k = 3, N = 15: numerator df = k − 1 = 2, denominator df = N − k = 12.
Analysis of Variance (ANOVA) is a test for equality of:
ANOVA tests H0: μ1 = μ2 = … = μk.
In a one-way ANOVA:
One-way ANOVA has only one factor, so there is no interaction term.
The test statistic used in analysis of variance procedure follow the:
ANOVA uses F = MS(between)/MS(within), which follows the F-distribution.
In a one way ANOVA test there are 5 observations in each of three treatments. The degrees of freedom for the treatments is:
Treatment df = k − 1 = 3 − 1 = 2.
The LSD test is applied only if the null hypothesis is:
LSD is used to find which means differ, which is needed only after ANOVA rejects H0.
In the test of goodness of fit, the ________ is used as a test statistic.
The goodness-of-fit test uses χ² = Σ(o − e)²/e.
If there are 8 treatments with 6 blocks in a randomized completed block design then what are the degrees of freedom for treatments?
Treatment df = k − 1 = 8 − 1 = 7.
The LSD test is applied when the null hypothesis is:
The LSD test compares pairs of treatment means only after the ANOVA F-test rejects H₀ of equal means.
When testing for independence in a contingency table with 3 rows and 4 columns, there are ________ degrees of freedom.
df = (r − 1)(c − 1) = 2 × 3 = 6.
When testing for independence in a contingency table with 2 rows and 5 columns, there are ________ degrees of freedom.
df = (r − 1)(c − 1) = (2 − 1)(5 − 1) = 4.
For a particular data the value of Pearson’s coefficient of skewness is greater than zero. What will be the shape of distribution?
Sk = (Mean − Mode)/S > 0 means mean > mode, i.e. a positively skewed distribution.
In measures of relative dispersion unit of measurement is:
Relative measures (e.g. C.V. = S/x̄ × 100) are ratios, so the units cancel and vanish.
If ŷ is the predicted value for a given x-value and b is the y-intercept then the equation of a regression line for an independent variable x and a dependent variable y is:
A straight line is ŷ = (slope)·x + (intercept) = mx + b.
The location of the critical region depends upon:
H1 decides whether the test is left-, right- or two-tailed, i.e. where the critical region lies.
The t-Distribution is……………. spread out than the standard normal distribution.
The t-distribution has variance ν/(ν − 2) > 1, so it is more spread out than Z (variance 1).
How many percent of values are less than 4th decile in a symmetric distribution.
D4 is the value below which 4/10 = 40% of the observations lie.
The combined distribution of more than two random variables is:
The distribution of several random variables taken together is their joint distribution (bivariate is only two).
The degrees of freedom for a T-test with sample size 14 is:
For a one-sample t-test df = n − 1 = 14 − 1 = 13.
What is fm in the formula of mode?
In Mode = l + (fm − f1)/((fm − f1) + (fm − f2)) × h, fm is the frequency of the modal (highest-frequency) class.
If a significance level of 1% is used rather than 5%, the null hypothesis is:
A smaller α makes the critical region smaller, so H0 is rejected less often.
The degrees of freedom for a t-test with sample size 10 is:
df = n − 1 = 10 − 1 = 9.
A die is rolled. What is the probability that the number rolled is greater than 2 and even:
Favourable outcomes {4, 6}: P = 2/6 = 1/3.
The probability of drawing a king of spade from a pack of 52 cards is:
There is only one king of spades among 52 cards, so P = 1/52.
An estimator T is said to be unbiased estimator of θ if
Unbiasedness means the expected value of the estimator equals the parameter: E(T) = θ.
From point estimation, we always get:
A point estimate is a single number; an interval estimate gives a range.
The best unbiased estimator for population variance σ2 is:
The sample variance S² = Σ(x − x̄)²/(n − 1) is the unbiased estimator of σ².
When c is a constant, then E(c) is:
A constant takes the value c with probability 1, so E(c) = c.
Var(4X + 5) = __________
Var(aX + b) = a²Var(X), so Var(4X + 5) = 16 Var(X); adding a constant does not change variance.
When f(x) is continuous probability function, then P(X = 1) is:
For a continuous random variable the probability at any single point is 0.
The hypergeometric random variable is a(an):
It counts the number of successes (0, 1, 2, …), so it is discrete.
From a sample of 200 people were asked whether they like a particular product. Fifty said 'yes' and remaining said 'no', assuming ‘yes’ means a success, which of the following is correct?
Sample proportion p̂ = 50/200 = 0.25.
In any data set, what percent of values fall in the interval Median ± Q.D?
For a (roughly symmetric) distribution Median ± Q.D runs from about Q1 to Q3, covering the middle 50%.
The height of a student is 60 inches. This is an example of ...........?
Height is measured and can take any value in an interval, so it is continuous data.
In Statistics, we have MSE which is abbreviation of……
MSE stands for mean square error (e.g. the error mean square in ANOVA).
The deviation of a distribution from symmetry is called:
Skewness measures lack of symmetry; kurtosis measures peakedness.
If E is an impossible event, then P(E) is:
An impossible event has no favourable outcomes, so P(E) = 0.
If a data set has the even number of observations, the median:
For even n the median is the mean of the two middle values of the ordered (arrayed) data.
In testing of hypothesis, we always begin it with assuming that:
A test starts by assuming H0 true and checks whether the sample contradicts it.
When each outcome of a sample space has equal chance to occur as any other, the outcomes are called:
Outcomes with the same chance of occurring are called equally likely.
If a random variable X denotes the number of heads when three distinct coins are tossed, the X assumed the values:
With three coins the number of heads can be 0, 1, 2 or 3.
If X and Y are independent variables, then E(XY) is:
For independent X and Y, E(XY) = E(X)·E(Y).
If P(E) is the probability that an event will occur, which of the following must be false:
Probabilities lie between 0 and 1, so P(E) = −1 is impossible.
The continuity correction factor is used when:
The ±0.5 correction is applied when a continuous (normal) curve approximates a discrete distribution such as the binomial.
Stem and leaf is more informative when data is:
A stem-and-leaf display is practical and informative for small data sets (fewer than about 100 values).
The branch of Statistics that is concerned with the procedures and methodology for obtaining valid conclusions is called:
Inferential statistics draws valid conclusions about populations from samples.
Which of the following is a systematic arrangement of data into rows and columns?
Tabulation arranges data systematically in rows and columns.
If you connect the mid-points of rectangles in a histogram by a series of lines that also touches the x-axis from both ends, what will you get?
Joining the tops of the histogram bars at class midpoints with straight lines, closed at both ends on the x-axis, gives a frequency polygon.
Which one of the following statements is true regarding a population?
A population is the whole collection of individuals, objects or measurements of interest; it need not be large or human.
In a simple linear regression model, if it is assumed that the intercept parameter is equal to zero, then:
With a = 0 the line is Y = bX, which passes through (0, 0).
A failing student is passed by an examiner is an example of:
Taking H0 "the student deserves to fail", passing him means accepting a false H0, a Type II error.
Which dispersion is calculated from all the observations?
S.D uses every observation; range and Q.D use only two values.
Standard deviation of the data 7, 7, 7, 7, 7, 7, 7 is
All values equal the mean 7, so every deviation is 0 and S.D = 0.
Which one is the poor measure of dispersion in open-end distribution?
In an open-ended distribution the extreme values are unknown, so the range cannot be found.
Men tend to marry women who are slightly younger than themselves. Suppose that every man married a woman who was exactly 5 years younger than themselves. Which of the following is correct:
Wife's age = husband's age − 5 is an exact increasing straight line, so r = +1.
Sum of absolute deviations of the values is least when deviations are taken from:
Σ|x − A| is minimum when A is the median.
Which of the following measures of central location is affected most by extreme values:
The arithmetic mean uses every value's magnitude, so extremes affect it most.
The difference between expected value of statistic and parameter is called:
Bias = E(T) − θ.
The following data shows the number of hours worked by 200 statistics students. Number of Hours: 0 - 9, 10 - 19, 20 - 29, 30 - 39 (frequencies). What is its class interval?
Class interval = difference of successive lower limits = 10 − 0 = 10.
The degrees of freedom for a t-test with sample size 6 is:
df = n − 1 = 6 − 1 = 5.
The difference between statistic and parameter is called:
Statistic − parameter is the sampling error; bias is E(statistic) − parameter.
Mean deviation is always:
For any data M.D ≤ S.D, with equality only in special cases.
Evaluate: (9-4)!
(9 − 4)! = 5! = 5×4×3×2×1 = 120.
Which formula represents the probability of the complement of event A:
P(A) + P(A′) = 1, so P(A′) = 1 − P(A).
Ideally the width of confidence interval should be:
The narrower the interval the more precise the estimate; ideally its width would be 0.
The probability distribution of a statistic is called the:
The distribution of a statistic over all possible samples is its sampling distribution.
If the following is a probability distribution, then what is the value of 'a': X: 1 2 3; P(X): 0.1 a 0.1
Probabilities sum to 1: a = 1 − 0.1 − 0.1 = 0.8.
A discrete probability function f(x) is always:
A probability function satisfies f(x) ≥ 0 with Σf(x) = 1; individual values need not be 1.
An expected value of a random variable is equal to:
E(X) = Σx f(x) is the mean of the distribution.
Which of the following is impossible in sampling:
A voters list must include every eligible person, so it needs complete enumeration, not a sample.
Which one of the following statements is true regarding a sample?
A sample is a part (subset) of the population.
The data for an ogive is found in which distribution?
An ogive plots cumulative frequencies against class boundaries.
10! = ………….
10! = 10 × 9! = 10 × 362880 = 3628800.
When E is an impossible event, then P(E) is:
An impossible event has probability 0.
When a fair die is rolled, the sample space consists of:
S = {1, 2, 3, 4, 5, 6} has 6 outcomes; 36 is for two dice.
Which graph is made by plotting the mid point and frequencies?
A frequency polygon plots class frequencies against class midpoints joined by straight lines.
In a set of 20 values all the values are 10, what is the value of median?
Every value is 10, so the middle value is 10.
E(4X + 5) = __________
E(aX + b) = aE(X) + b = 4E(X) + 5.
A set that contains all possible outcomes of a system is known as
The set containing all elements under consideration is the universal set.
A population that can be defined as the aggregate of all the conceivable ways in which a specified event can happen is known as:
A population of all conceivable ways an event can happen (e.g. all possible coin tosses) is hypothetical.
Which of the following pairs of events are mutually exclusive?
No number is both above 100 and below −200; each other pair has common members (5; 12; −1).
The collection of all outcomes for an experiment is called:
The set of all possible outcomes of an experiment is its sample space.
Symbolically, a conditional probability is:
P(A/B) denotes the probability of A given that B has occurred.
Which of the following statements is INCORRECT about the sampling distribution of the sample mean?
The sampling distribution shows how sample means vary from sample to sample around μ, not how one sample is spread around its own mean.
If one event is unaffected by the outcome of another event the two events are said to be:
Events are independent when the occurrence of one does not change the probability of the other.
The sample mean is an unbiased estimator for the population mean. This means:
Unbiased means E(X̄) = μ, i.e. the mean of all possible sample means equals μ.
Probability of an impossible event is always:
An impossible event has probability 0.
When two coins are tossed the probability of at most one head is:
At most one head = {HT, TH, TT}, so P = 3/4.
For exhaustive events, the P(AUBUC) is equal to:
Exhaustive events together make up the sample space, so P(A∪B∪C) = P(S) = 1.
One card is drawn from a standard 52 card deck. In describing the occurrence of two possible events, an Ace and a King, these two events are said to be:
One card cannot be both an Ace and a King, so the events are mutually exclusive.
The number of parameters in hypergeometric distribution is (are):
The hypergeometric distribution has three parameters N, n and k.
If f(x) is a continuous probability function, then P(X = 2) is:
For a continuous random variable P(X = any single value) = 0.
In regression line Y=a+bX, Y is called:
Y is the variable being predicted, the dependent variable.
If A and B are mutually exclusive events with P(A) = 0.25 and P(B) = 0.50, Then P(A or B) = ………
For mutually exclusive events P(A∪B) = 0.25 + 0.50 = 0.75.
In a well shuffled pack of 52 playing cards, the probability of drawing any one diamond card is
There are 13 diamonds, so P = 13/52.
If Y = 3X + 5, then S.D of Y is equal to
S.D(aX + b) = |a| S.D(X) = 3 s.d(x); the constant 5 does not affect spread.
The probability of drawing a red queen card from well-shuffled pack of 52 playing cards is
There are two red queens (hearts and diamonds), so P = 2/52.
If P(B|A) = 0.25 and P(A and B) = 0.20, then P(A) is
P(A) = P(A∩B)/P(B|A) = 0.20/0.25 = 0.80.
The classical definition of probability assumes:
Classical probability m/n requires the outcomes to be equally likely.
In scatter diagram, the variable plotted along Y-axis is:
The dependent variable is plotted on the Y-axis and the independent on the X-axis.
Which of the following measures of dispersion are based on deviations from the mean?
Variance, S.D and mean deviation (about the mean) are all computed from x − x̄.
What does it mean when a data set has a standard deviation equal to zero?
S.D = 0 only when every deviation from the mean is 0, i.e. all values are equal.
A set of possible values that a random variable can assume and their associated probabilities of occurrence are referred to as ________.
The values of a random variable together with their probabilities form its probability distribution.
Which of the following can never be probability of an event?
A probability cannot be negative.
The standard deviation of -1, -1, -1, -1 will be
All values are equal (−1), so every deviation is 0 and S.D = 0.
The Special Rule of Addition is used to combine:
The special addition rule P(A∪B) = P(A) + P(B) holds for mutually exclusive events.
The set which is the sub-set of every set is
The empty set ∅ is a subset of every set.
When two dice are rolled the number of possible sample points is:
6 × 6 = 36 ordered outcomes.
If two events A and B are not mutually exclusive then
The general addition rule subtracts the overlap: P(A∪B) = P(A) + P(B) − P(A∩B).
Evaluate (10-4)!
(10 − 4)! = 6! = 720.
When we toss a coin, we get only:
S = {H, T} has 2 outcomes.
If we roll a die then probability of getting a ‘6’ will be
One favourable face out of 6: P = 1/6.
If P(A) = 0.45, P(B) = 0.35, and P(A and B) = 0.25, then P(A | B) is:
P(A|B) = P(A∩B)/P(B) = 0.25/0.35 = 0.714.
Which of the following is not a measure of central tendency?
Standard deviation measures dispersion, not location.
Random experiment can be repeated any no. of times under the……… conditions.
A random experiment can be repeated under the same (similar) conditions.
The simultaneous occurrence of two events is called:
The probability that two events occur together, P(A∩B), is a joint probability.
In regression analysis, the variable that is being predicted is the
The predicted variable is the dependent variable.
The probability of continuous random variable x on any particular point is always zero.
For a continuous random variable P(X = x) = 0 for every single point.
P(an event) = no. of favorable outcomes/total no. of outcomes is the definition of
m/n is the classical definition, which is an objective approach to probability.
If C is a constant, then E(c) =
E(c) = c.
When we toss a fair coin 4 times, the sample space consists of….points.
2⁴ = 16 sample points.
If we roll three fair dice then the total number of outcomes is:
6 × 6 × 6 = 216.
The probability of an event is always:
0 ≤ P(A) ≤ 1 for every event.
In a multiplication theorem P(A and B) equals:
General multiplication rule: P(A∩B) = P(A)·P(B|A).
If a die is rolled, what is the probability of getting an even number greater than 2?
Favourable outcomes {4, 6}: 2/6 = 1/3.
In a Discrete probability distribution, P(x > 23) is read as:
x > 23 means strictly more than 23 successes.
A dormitory on campus houses 200 students. 120 are male, 50 are upper division students, and 40 are upper division male students. A student is selected at random. The probability of selecting a lower division student, given the student is a female, is:
Females = 80, upper-division females = 50 − 40 = 10, lower-division females = 70: P = 70/80 = 7/8.
In a simple regression line model, it is assumed that the intercept parameter is equal to zero, then
With intercept 0 the line is Y = bX, which passes through (0, 0).
If P(A∩B) = P(A/B).P(B), then A and B are
P(A∩B) = P(B)P(A|B) is the general multiplication rule used for dependent events; for independent events it reduces to P(A)P(B).
A fair coin is tossed three times, the probability that at least one head appears,
P(at least one head) = 1 − P(TTT) = 1 − 1/8 = 7/8.
In probability distribution, the sum of probabilities is equal to
Σf(x) = 1 for any probability distribution.
When a coin is tossed 3 times, the probability of getting 3 tails is
Only TTT out of 8 outcomes: P = 1/8.
In how many ways can a team of 11 players be chosen from a total of 16 players?
16C11 = 16C5 = 16!/(5!·11!) = 4368.
The standard deviation of c (constant) is
A constant does not vary, so its S.D = 0.
Let E and F be events associated with the same experiment. Suppose the E and F are independent and that P(E) = 1/4 and P(F) = 1/2. Then P(E U F) is:
P(E∪F) = 1/4 + 1/2 − (1/4)(1/2) = 6/8 − 1/8 = 5/8.
In which of the following situations binomial distributions is approximate to normal distribution?
Normal approximation needs np and nq large (about 5 or more): np = 0.5, 0.5, 5, 1, so only n = 100, p = 0.05 qualifies.
A random experiment has five outcomes in its sample space {s1, s2, s3, s4, s5}. If P(s1)=0.2, P(s2)=0.3, P(s3)=0.1 and P(s4)=0.2 then P(s5)=?
P(s5) = 1 − (0.2 + 0.3 + 0.1 + 0.2) = 1 − 0.8 = 0.2.
The joint density function f(x,y) will be a pdf if
A joint density must integrate to 1 over the whole plane.
A random variable that can assume every possible value in an interval [a, b]:
A variable taking every value in an interval is continuous.
Normal approximation to the binomial distribution is used when:
Both np and nq must exceed 5 for the normal approximation to be good.
A probability density function ‘f(x)’ has the following property:
A density must satisfy f(x) ≥ 0; it can be 0 outside the range.
For a continuous random variable X, P(X = x) is:
The probability at any single point of a continuous variable is 0.
An urn contains 4 red balls and 6 green balls. A sample of 4 balls is selected from the urn without replacement. It is the example of:
Sampling without replacement from a finite population of two kinds gives a hypergeometric distribution.
The probability distribution of the proportion of successes in all possible samples is called the:
The distribution of p̂ over all possible samples is the sampling distribution of sample proportions.
If an estimator is more efficient than the other estimator, its shape of the sampling distribution will be
A more efficient estimator has smaller variance, so its sampling distribution is more concentrated (peaked).
Match the binomial probability P(x < 23) with the correct statement.
x < 23 means fewer than 23 successes.
In statistics, the term ‘expected value’ implies the ___ value.
The expected value of a random variable is its mean.
A quantity obtained by applying certain rule or formula is known as
The rule or formula is the estimator; the numerical value it gives is the estimate.
Total no. of possible samples of size 2 (without replacement) from the population of size 6, will be:
6C2 = 6×5/2 = 15.
When a coin is tossed 3 times, the probability of getting 3 or less tails is
In 3 tosses the number of tails is always 0–3, so P(X ≤ 3) = 1.
Total no. of possible samples of size 3 (with replacement) from the population of size 6, will be:
With replacement N^n = 6³ = 216.
Using the normal approximation to the binomial distribution with n = 3 and p = 0.0571 the value of mean is:
μ = np = 3 × 0.0571 = 0.1713.
The probability of success changes from trial to trial, is the property of:
Without replacement the success probability changes after each draw (hypergeometric); in a binomial experiment it stays constant.
If a statistic used as an estimator, has its expected value equal to the true value of the population parameter being estimated then it is called........
E(T) = θ is the definition of an unbiased estimator.
Which of the following is desirable for a good point estimator?
A good estimator should be unbiased, consistent, efficient (and sufficient).
If X and Y are two discrete r.v’s with joint probability function f(x,y), then the conditional probability function Y given X, f(y|x) is given by
f(y|x) = f(x, y)/g(x), where g is the marginal of X.
Which of the probability distributions has three parameters?
Hypergeometric has N, n, k; binomial and normal have two, Poisson one.
A standard deviation obtained from sampling distribution of sample statistics is known as
The S.D of a sampling distribution is called the standard error.
A parameter is a ..... quantity.
A parameter describes a population and is a fixed (constant) value.
How can you define statistical inference?
Inference generalizes from a sample to the population.
The probabilities of the various values of the sample statistic can be computed using the......... definition of probability.
All possible samples are equally likely, so probabilities of a statistic's values are found by the classical definition.
E(10X + 3) = _______
E(aX + b) = aE(X) + b = 10E(X) + 3.
If a random variable X denotes the number of heads when we toss a fair coin 5 times, the X assumed the values:
In 5 tosses the number of heads can be 0 to 5.
The probability can never be:
A probability cannot be negative.
The conditional probability P(A\B) is:
P(A|B) = P(A∩B)/P(B), P(B) > 0.
When we draw the sample with replacement, the probability distribution to be used is:
With replacement trials are independent with constant p, so the binomial applies.
If P(A) = 0.3 and P(B) = 0.5, find P(A/B) where ‘A’ and ‘B’ are independent:
For independent events P(A|B) = P(A) = 0.3 (0.15 is P(A∩B)).
If the second moment ratio is less than 3 the distribution will be:
β2 < 3 means flatter than normal: platykurtic.
For the independent events A and B if P(A) = 0.25, P(B) = 0.40 then P(A and B) = ……
P(A∩B) = P(A)P(B) = 0.25 × 0.40 = 0.10.
A random variable X has a probability distribution as follows: X | 0 1 2 3; P(X) | 2k 3k 13k 2k. What is the possible value of k:
2k + 3k + 13k + 2k = 20k = 1, so k = 0.05.
The probability of drawing any one spade card is:
There are 13 spades in 52 cards, so P = 13/52.
The probability distribution of X is given below. x: 0 1 2 3 4; p(x): 0.73 ? 0.06 0.04 0.01. What is the value of missing probability?
Missing p = 1 − (0.73 + 0.06 + 0.04 + 0.01) = 1 − 0.84 = 0.16.
P(|X| < 1.5) =
|X| < 1.5 is the same event as −1.5 < X < 1.5.
An estimator is always a ..........?
An estimator is computed from sample values, so it is a statistic (a random variable).
Consistency is a ....... property.
Consistency describes the estimator's behaviour as n → ∞, a large-sample property.
What should be the type of both variables in joint probability distribution of X and Y?
A joint distribution is treated for two discrete (summed) or two continuous (integrated) variables.
If X and Y are two discrete r.v’s with joint probability function f(x,y), then the conditional probability function X given Y, f(x|y) is given by
f(x|y) = f(x, y)/h(y), where h is the marginal of Y.
A random sample of n = 6 has the elements 6, 10, 13, 14, 18 and 20. What is the point estimate of the population mean?
x̄ = (6 + 10 + 13 + 14 + 18 + 20)/6 = 81/6 = 13.5.
If our sampled population is normal, then sampling distribution of the sample mean will ………:
If the population is normal, x̄ is exactly normal for any sample size.
In sampling without replacement, fpc is not used when:
The finite population correction (N − n)/(N − 1) can be ignored when the sample is at most 5% of the population.
The Gallup Poll has decided to increase the size of its random sample of Canadian voters from about 1200 people to about 4000 people. The effect of this increase is to:
A larger n reduces the standard error σ/√n, i.e. the variability of the estimate; it does not remove bias.
The variance of t-distribution, for v > 2, is always:
Var(t) = ν/(ν − 2) > 1 for ν > 2.
A quantity obtained by applying certain rule or formula is known as
The rule or formula is the estimator; the number it produces is the estimate.
Criteria to check a point estimator to be good involves
A good point estimator should be consistent, unbiased and efficient.
1-α is the probability of ......
1 − α is the probability that the statistic falls in the acceptance region when H0 is true.
To find the estimate of a parameter …….methods are used.
Several methods exist (moments, least squares, maximum likelihood, etc.), so "many".
A failing student is passed by an examiner. It is an example of:
Passing a student who should fail is accepting a false H0, a Type II error.
Which statement is NOT CORRECT?
Bias comes from the sampling method; a large but badly chosen sample is still biased.
Probability of type II error is
P(Type II error) = β; 1 − β is the power.
With increase in sample size, distribution tends to be a.......
By the central limit theorem the sampling distribution tends to the normal, which is bell-shaped and mesokurtic.
How many numbers of parameter(s) are in t-distribution?
The t-distribution has one parameter, its degrees of freedom ν.
A standard deviation obtained from sampling distribution of sample statistics is known as
The S.D of a sampling distribution is the standard error.
A standardized estimate has mean and variance
Standardizing (T − E(T))/S.E gives mean 0 and variance 1.
A composite hypothesis comprises of ………
A composite hypothesis specifies a range (≠, <, >), not a single value.
For a particular hypothesis test, α = 0.05 and β = 0.05. The power of test is equal to:
Power = 1 − β = 1 − 0.05 = 0.95.
In a t-distribution
The t-distribution is symmetric and unimodal about 0, so mean = median = mode.
If we reject the null hypothesis, we might be making
Rejecting H0 risks a Type I error (rejecting a true H0); a Type II error can only occur when H0 is accepted.
Herbicide A has been used for years in order to kill a particular type of weed, but an experiment is to be conducted in order to see whether a new herbicide, Herbicide B, is more effective than Herbicide A. Herbicide A will continue to be used unless there is sufficient evidence that Herbicide B is more effective. The alternative hypothesis in this problem is that
H1 is the claim needing evidence: Herbicide B is more effective than A.
How many steps are involved in general procedure for testing hypothesis:
The general procedure has six steps: H0/H1, α, test statistic, critical region, computation, conclusion.
The critical region is in:
The rejection region lies in the tail(s) of the sampling distribution.
t-distribution is used to test the hypothesis about.....
The t-test is used for means (small samples, σ unknown); proportions use Z.
The term 1-β is called
1 − β is the power of the test.
Which one provides the basis for hypothesis testing?
The test is built on H0, which is assumed true and then accepted or rejected.
Alpha is the probability of ......
α = P(rejecting H0 when it is true).
An expected value of a random variable is equal to:
E(X) is the mean of the distribution.
The best unbiased estimator for population variance is:
S² = Σ(x − x̄)²/(n − 1) is the unbiased estimator of σ².
The mean of a population is 6. A sample of size 20 is taken with replacement from this population. Which one of the following is the mean of the sampling distribution of mean?
The mean of the sampling distribution of x̄ equals the population mean, μ(x̄) = μ = 6.
The value of χ² can never be:
χ² is a sum of squared terms divided by positive expected values, so it cannot be negative.
The probability of an event always lies between:
By the axioms of probability, 0 ≤ P(A) ≤ 1.
The number of parameters in a Poisson distribution is (are):
The Poisson distribution has one parameter, μ = λ (its mean).
A good way to get a small standard error is to use a ________.
Standard error σ/√n decreases as the sample size n increases.
Which of the following is a measure of absolute dispersion?
Mean deviation is expressed in the units of the data, so it is an absolute measure; CV is relative.
An automobile is running, during the first 60 Km, at the rate of 10 Km/hr. During the second 60 Km at the rate of 30 Km/hr, while during the third 60 Km its speed was 40 Km/hr. What method is more appropriate to calculate the average speed?
Speeds over equal distances are averaged by the harmonic mean: 3/(1/10 + 1/30 + 1/40) ≈ 18.95 km/hr.
If S.D(X) = 5 then S.D((2X + 5)/2) = ______
(2X + 5)/2 = X + 2.5; adding a constant does not change SD, so SD = 5.
For a particular data set the Pearson's coefficient of skewness is greater than zero. What will be the shape of distribution?
A positive Pearson coefficient (mean > mode) indicates a positively skewed distribution.
The total number of samples when sampling is done without replacement is equal to:
Without replacement, the number of possible samples of size n from N is the combination ᴺCₙ.
Which of the following is correct option to find P(X + Y ≤ 1)?
Pairs with x + y ≤ 1 (non-negative integer values) are (0,0), (0,1) and (1,0).
The conditional probability function f(x | 1) = __________
f(x | y) = f(x, y)/h(y), where h is the marginal of Y; with y = 1 it is f(x, 1)/h(1).
If f(x, y) is bivariate probability density function of continuous random variables X and Y then marginal density function of y i.e. h(y) is:
The marginal of Y is obtained by integrating the joint density over x: h(y) = ∫ f(x, y) dx.
________ is a range of numbers inferred from the sample that has a certain probability of including the population parameter over the long run.
A confidence interval is a sample-based range that covers the parameter with a stated long-run probability.
By definition f(y | x) = __________
Conditional of Y given X is the joint divided by the marginal of X: f(y | x) = f(x, y)/h(x).
Which of the following is an unbiased estimator of population σ²?
E[Σ(x − x̄)²/(n − 1)] = σ², so dividing by n − 1 gives the unbiased estimator.
For α = 0.01, the critical values of z for two tailed test are equal to:
Two-tailed α = 0.01 puts 0.005 in each tail, giving z = ±2.58.
A judge can acquit a guilty person is the example of:
With H₀: the accused is innocent, acquitting a guilty person means accepting a false H₀, a Type II error.
The proportion of males in Pakistan is at least 0.48, the alternative hypothesis H₁ is
The claim P ≥ 0.48 (contains equality) is H₀, so H₁: P < 0.48.
What factor determines the shape of the t-distribution?
The t-distribution's shape depends only on its degrees of freedom.
Hypergeometric probability distribution has:
The hypergeometric distribution is specified by N (population size), n (sample size) and k (number of successes in the population).
For any two estimators T1 and T2, if VAR(T1) < VAR(T2), then T1 is:
The estimator with the smaller variance is the more efficient one.
Atmosphere pressure is the example of:
Atmospheric pressure is measured numerically and varies, so it is a quantitative (continuous) variable.
Which type of the curve is represented by the following shape? [figure: a curve with a long tail to the left and its peak toward the right]
A curve whose long tail extends to the left is negatively skewed.
Which one of the following is not a type of frequency curve?
A frequency polygon is a graph of a frequency distribution, not a type (shape) of frequency curve.
Which one is the measure of central tendency:
Measures of central tendency are averages; the others describe spread.
In a set of 10 values all the values are 5, what will be the P₅₀?
When all values equal 5, every percentile, including P₅₀ (the median), is 5.
P(A or B) = P(A) + P(B), then A and B are:
The addition rule reduces to P(A) + P(B) only when P(A and B) = 0, i.e. the events are mutually exclusive.
If σ² is unknown, then we use Z-test if the sample size is:
With σ² unknown, a large sample (n ≥ 30) lets s replace σ and the z-test is used.
When a coin is tossed 3 times, the probability of 3 tails is:
P(TTT) = (1/2)³ = 1/8.
Rumour has reached the Trading Standards Officer that the manufacturer ABC is deliberately underfilling his cartons of orange juice. It is decided that a sample should be taken to check this claim. The stated contents on the carton are 100 ml on the average, then the alternative hypothesis is:
Underfilling means the mean content is below 100 ml, so H₁: μ < 100.
Rumour has reached the Trading Standards Officer that the manufacturer ABC is deliberately underfilling his cartons of orange juice. It is decided that a sample should be taken to check this claim. The stated contents on the carton are 100 ml on the average, then the null hypothesis is:
The null hypothesis states the claimed value, H₀: μ = 100 ml.
By definition f(xᵢ | yⱼ) = __________
The conditional of X given Y = yⱼ is the joint probability divided by the marginal of Y: f(xᵢ, yⱼ)/h(yⱼ).
Which one of the following is the most common example of a situation for which the main parameter of interest is a population proportion?
A binomial experiment counts successes, so its parameter of interest is the proportion p.
An estimator which has the smallest standard error among all unbiased estimators fulfills the property of ____________.
The unbiased estimator with the smallest variance (standard error) is the most efficient.
Which of the following can never be taken as the probability of an event?
A probability must lie between 0 and 1, so −0.5 is impossible.
A set is any well-defined collection of
A set is a well-defined collection of distinct objects.
Measure of dispersion is used to calculate the:
Dispersion measures how scattered the values are around the centre.
If X and Y are independent variables then Var(X−Y) = ______
For independent X and Y, Var(X−Y) = Var(X) + (−1)²Var(Y) = Var(X) + Var(Y).
How the standard error is decreased:
Standard error σ/√n falls as n grows.
The total number of samples when sampling is done with replacement is equal to:
With replacement each of the n draws can be any of N units, giving Nⁿ samples.
If you draw all possible samples from some population, calculate the mean for each of the sample and construct the probability distribution of the sample means, what would you have?
The probability distribution of a statistic over all possible samples is its sampling distribution.
A deserving player is not selected in the team is an example of:
Taking H₀ as 'the player is deserving', rejecting a true H₀ is a Type I error.
“A point estimate plus/minus a few times the standard error of that estimate”. This statement represents:
A confidence interval has the form estimate ± (critical value) × standard error.
If X̄ is the mean of the n observations, then which test statistic will be used to calculate the confidence limits of the population variance σ²?
Confidence limits for σ² use (n−1)s²/σ², which follows a χ² distribution with n−1 df.
Which of the following value could not represent a coefficient of correlation?
The correlation coefficient must satisfy −1 ≤ r ≤ 1, so 1.09 is impossible.
When f(x) is continuous probability function, then P(X = 2) is:
For a continuous random variable the probability of any single exact value is 0.
Which of the following scale has true zero point?
Only the ratio scale has a true (absolute) zero point.
The difference between the largest and the smallest data values is called the
Range = largest value − smallest value (xₘ − x₀).
Which is appropriate average for finding the average speed of a car:
Speeds (rates) over equal distances are averaged with the harmonic mean.
Which one is the formula of mid range:
Mid-range is the average of the smallest and largest values, (x₀ + xₘ)/2.
Which one of the following is a meso-kurtic curve?
The normal curve is the standard mesokurtic curve (β₂ = 3).
The critical region for H₁: μ > μ₀ when α = 0.01 is:
A right-tailed test at α = 0.01 rejects H₀ when z > z₀.₀₁ (= 2.33).
For degree of freedom ν > 2 the variance of t-distribution is always:
Var(t) = ν/(ν − 2), which exceeds 1 for every ν > 2.
If the sampling distribution of X̄ is normal, we would expect 99% of the sample means to be within the interval:
For a normal distribution 99% of values lie within ±2.58 standard errors of the mean.
Two continuous r.v.'s X and Y are said to be independent if and only if :
X and Y are independent exactly when the joint density equals the product of the marginals.
If an estimator gets closer to the population parameter by increasing sample size then it is known as:
An estimator that converges to the parameter as n increases is consistent.
Quantitative variable is further divided into:
Quantitative variables are either discrete (countable) or continuous (measurable).
Color of the dress is the example of:
Colour is a non-numeric attribute, so it is qualitative data.
Which one is commonly called a bell shaped distribution?
A bell-shaped curve is symmetrical about its centre.
A fair coin is tossed three times, the probability that at least one head appear is:
P(at least one head) = 1 − P(no head) = 1 − (1/2)³ = 7/8.
The probability of simultaneous occurrence of two events is called:
P(A∩B), the probability that both events occur together, is the joint probability.
What is the stem part of 243:
With the last digit as the leaf, 243 splits into stem 24 and leaf 3.
A numerical value used as a summary measure for a sample, such as sample mean, is known as a :
A summary measure computed from sample data is a statistic.
The critical region for H1: μ < μ0 when α = 0.05 is:
A left-tailed alternative puts the whole rejection region in the lower tail: z < −z0.05 = −1.645.
Assuming that following is a probability distribution, then what is the value of 'a': X: 1, 2, 3; P(X): 0.1, a, 0.1
Probabilities sum to 1, so a = 1 − 0.1 − 0.1 = 0.8.
A randomly selected sample of 400 students at university was asked whether or not they will participate in politics. Forty-six percent of the 400 student surveyed answered “yes”. Which one of the following statement about number 46% is correct?
46% is computed from the sample of 400, so it is a sample statistic (p̂).
In interval estimation, we always get:
Interval estimation gives a range of values (an interval) expected to contain the parameter.
How many parameter(s) are in t-distribution?
The t-distribution has one parameter, its degrees of freedom ν.
After an interval is constructed, then what is the probability of occurrence of the parameter in it?
Once computed, a fixed interval either contains the fixed parameter or not, so the probability is 1 or 0.
In the regression line Y = a + bX, the non-random variable is:
In regression of Y on X, X is the fixed (non-random) independent variable and Y is random.
The variance of the hypergeometric probability distribution is:
Hypergeometric variance is n(k/N)((N − k)/N)((N − n)/(N − 1)), the binomial npq times the finite correction.
Tabulation is the process of arranging data into:
A table arranges data systematically in rows and columns.
A simple bar chart consists of:
A simple bar chart may be drawn with bars either vertically or horizontally.
Which one is the formula to calculate the approximate value of class interval?
Class interval h ≈ Range / number of classes.
Which one of the following is equal to the 2nd quartile:
Q2 divides the data into two equal halves, so it equals the median (also D5 and P50).
For the independent events A and B if P(A) = 0.15, P(B) = 0.50 then P(A∩B) =……..
For independent events P(A∩B) = 0.15 × 0.50 = 0.075.
Which one of the following measure is not based on all the observations?
The mode depends only on the most frequent value, while the three means use every observation.
The sum of deviations is zero, when deviations are taken from:
Σ(x − x̄) = 0 holds only for the arithmetic mean.
The degrees of freedom for a t-test with sample size 'n' is:
A one-sample t-test uses ν = n − 1 degrees of freedom.
In a symmetrical distribution, the coefficient of skewness is equal to :
In a symmetric distribution mean = median = mode, so skewness is 0.
Which measure of dispersion is used to compare the variation of two data sets?
C.V = (s/x̄)×100 is unit-free, so it compares variation between data sets.
What are the number of ways in which four books can be arranged on a shelf?
Four books can be arranged in 4! = 24 ways.
When f(x) is continuous probability function for 1 < X < 5, then P (X < 1) is:
f(x) is zero outside 1 < x < 5, so P(X < 1) = 0.
What curve shape would you expect for the distribution of death rates of population of all age groups?
Death rates are high for infants and the very old and low in between, giving a U-shaped curve.
Test Statistics χ² is equal to:
With S² = Σ(x − x̄)²/n, the test statistic for H₀: σ² = σ₀² is χ² = nS²/σ₀² with n − 1 d.f.
What is the probability of drawing a red-queen card from a well shuffled pack of 52 playing cards?
There are 2 red queens (hearts and diamonds) among 52 cards, so P = 2/52.
The probability of drawing a spade card is:
A pack of 52 cards has 13 spades, so P(spade) = 13/52.
In construction of a histogram, what would be taken along X-axis?
Histogram rectangles are drawn between class boundaries on the X-axis so adjacent bars touch.
If A = {1, 2, 3, 4, 5, 10} and B = {1, 3, 5} then B ⊂ A means:
B ⊂ A means every element of B (1, 3, 5) is in A, i.e. B is a subset (contained in) A.
What is the mode in the word STATISTICS:
Letter counts: S = 3, T = 3, I = 2, A = 1, C = 1, so S and T are both modes (bimodal).
In a symmetrical distribution, if Q₁ = 20, Q₂ = 30 then Q₃ is:
In a symmetric distribution Q₃ − Q₂ = Q₂ − Q₁, so Q₃ = 30 + (30 − 20) = 40.
The sum of deviations from mean is:
Σ(x − x̄) = Σx − n·x̄ = n·x̄ − n·x̄ = 0, a basic property of the arithmetic mean.
Mean deviation is always:
Mean deviation ≤ standard deviation for any data set, with equality only when all |x − x̄| are equal.
Which one is the formula of mid range:
Mid-range is the average of the smallest (x0) and largest (xm) values: (x0 + xm)/2.
The critical region for H1: μ > μ0 when α = 0.01 is:
A right-tailed test at α = 0.01 rejects H0 when z > z0.01 (= 2.33).
No MCQs match this filter.Is filter mein koi MCQ nahi.
How many parameters are involved in hypergeometric distribution?
Three parameters: N (population size), n (sample size) and k (number of successes in the population).
If an automobile is driven on the average no more than 16000 Km per year, then formulate the null and alternative hypothesis.
The claim "no more than 16000 km" is taken as the null hypothesis: H0: μ ≤ 16000 km and H1: μ > 16000 km (a right-tailed test).
Write down the test statistic when chi-square goodness of fit test is performed.
χ² = Σ (oi − ei)²/ei, where oi are observed and ei expected frequencies; it has approximately a χ² distribution with k − 1 − r degrees of freedom (k classes, r parameters estimated from the data).
For the given grouped data, compute the MEDIAN value. Group (Frequency): 51 – 55 (1), 56 – 60 (7), 61 – 65 (8), 66 - 70 (4).
n = 20, n/2 = 10. Cumulative frequencies: 1, 8, 16, 20, so the median class is 61–65 (class boundaries 60.5–65.5), h = 5, f = 8, C = 8. Median = l + (h/f)(n/2 − C) = 60.5 + (5/8)(10 − 8) = 60.5 + 1.25 = 61.75.
Calculate first two moments about mean from the following data. X = 11, 12, 14, 15, 16, 18, 19
x̄ = 105/7 = 15. Deviations x − x̄: −4, −3, −1, 0, 1, 3, 4. First moment about mean m1 = Σ(x − x̄)/n = 0/7 = 0. Second moment m2 = Σ(x − x̄)²/n = (16 + 9 + 1 + 0 + 1 + 9 + 16)/7 = 52/7 = 7.43.
In a binomial distribution, the mean and the variance were found to be np = 12.38 and npq = 8.64 respectively. Find the two parameters of binomial distribution.
q = npq/np = 8.64/12.38 = 0.698, so p = 1 − 0.698 = 0.302. n = np/p = 12.38/0.302 ≈ 41. The parameters are n ≈ 41 and p ≈ 0.302.
Use the data 71, 64, 54, 51, 65, 74, 52 to complete the following stem-and-leaf table (Stem, Leaf).
Taking the tens digit as stem and the units digit as leaf: Stem 5 | 1 2 4; Stem 6 | 4 5; Stem 7 | 1 4.
Define Cartesian product. Construct a Cartesian product for the following two sets A = {a,b} and B = {1,3,5}
The Cartesian product A × B of two sets is the set of all ordered pairs (x, y) with x ∈ A and y ∈ B. Here A × B = {(a,1), (a,3), (a,5), (b,1), (b,3), (b,5)}, which has 2 × 3 = 6 elements.
A fair coin is tossed 25 times. Applying normal approximation to the binomial distribution, find the mean and variance.
n = 25, p = q = 0.5. Mean μ = np = 25 × 0.5 = 12.5; variance σ² = npq = 25 × 0.5 × 0.5 = 6.25 (σ = 2.5).
Formulate 90% confidence interval for population variance.
For a sample of size n from a normal population with sample variance s² (divisor n − 1), the 90% confidence interval for σ² is (n − 1)s²/χ²(0.05, n−1) < σ² < (n − 1)s²/χ²(0.95, n−1), where χ²(0.05, n−1) and χ²(0.95, n−1) are the chi-square values with upper-tail areas 0.05 and 0.95 for n − 1 degrees of freedom.
Write down the formula of 78th Percentile.
For ungrouped data P78 = value of the 78(n + 1)/100 th ordered observation. For grouped data P78 = l + (h/f)(78n/100 − C), where l is the lower class boundary of the class containing the 78n/100 th value, h the class width, f its frequency and C the cumulative frequency before it.
Definition of quantitative variable.
A quantitative variable is one whose values are numbers obtained by counting or measuring (e.g. number of children, height); it can be discrete or continuous.
Var(X) with E(X²) = 43/6 and E(X) = 5/2: Var(X) = 43/6 − (5/2)². Is the result 11/12, or (172 − 150)/12 = 11/6?
Var(X) = E(X²) − [E(X)]² = 43/6 − 25/4 = 86/12 − 75/12 = 11/12 ≈ 0.917. So 11/12 is correct; 11/6 is an arithmetic error (172/24 − 150/24 = 22/24 = 11/12).
A sample of size n = 3 is drawn without replacement from a population of N = 5 items whose values are 0, 2, 3, 6, 7. Write all possible samples.
Number of samples = 5C3 = 10: (0,2,3), (0,2,6), (0,2,7), (0,3,6), (0,3,7), (0,6,7), (2,3,6), (2,3,7), (2,6,7), (3,6,7).
Calculate the sampling error if the sample mean is 102 and the population mean is 100.
Sampling error = x̄ − μ = 102 − 100 = 2.
Define unbiased estimator.
An estimator θ̂ is unbiased if its expected value equals the parameter it estimates, i.e. E(θ̂) = θ (for example E(x̄) = μ).
Define quartile deviation.
Quartile deviation is half the difference between the third and first quartiles: Q.D = (Q3 − Q1)/2. It is also called the semi-interquartile range.
Describe the significance level.
The level of significance α is the probability of rejecting the null hypothesis when it is true (Type I error). It is fixed before the test and sets the critical region; common values are 0.05 and 0.01.
Discuss the central limit theorem.
If a population has mean μ and finite variance σ², then the sampling distribution of the sample mean x̄ approaches a normal distribution with mean μ and variance σ²/n as the sample size n becomes large, whatever the shape of the population.
Briefly explain an experimental design.
An experimental design is a plan for collecting data on the problem under study (assigning treatments to experimental units) so that valid and objective inferences can be drawn about the problem.
Define the independent and dependent variable in regression.
In the regression Y = a + bX, Y is the dependent (response) variable whose value is predicted, and X is the independent (explanatory) variable used to predict Y.
What is the difference between an outcome and an event?
An outcome is a single result of one trial of a random experiment (one sample point). An event is a subset of the sample space and may contain one outcome or several outcomes.
The mean of a Poisson distribution is 5 while its standard deviation is 4. Comment on it.
In a Poisson distribution mean = variance. Here variance = 4² = 16, which is not equal to the mean 5, so the statement is not possible for a Poisson distribution.
Discuss three properties of normal distribution.
(1) The normal curve is bell-shaped and symmetric about μ, so mean = median = mode and the third moment about the mean is zero. (2) The curve is asymptotic to the x-axis as x → ±∞ and the total area under it is 1. (3) Its kurtosis β2 = μ4/μ2² = 3 (mesokurtic); also about 68%, 95% and 99.7% of the area lie within μ ± σ, μ ± 2σ and μ ± 3σ.
The 90% confidence interval for the population mean is 11 to 20. Interpret this result.
If many samples were drawn and an interval built from each in the same way, about 90% of these intervals would contain the true population mean. We are 90% confident that μ lies between 11 and 20.
Define LSD test.
The least significant difference (LSD) test is used after ANOVA rejects H0 of equal means. It computes the smallest difference between two treatment means that is significant, LSD = t(α/2, v)√(2MSE/r), where v is the error degrees of freedom and r the number of replications; any pair of means whose absolute difference exceeds LSD is declared significantly different.
How many parameters are associated with the F-distribution and what is the range of the distribution?
The F-distribution has two parameters, the degrees of freedom v1 (numerator) and v2 (denominator). F ranges from 0 to +∞.
If the population proportions are given as P1 = 0.4, P2 = 0.20, find σ²(p̂1 − p̂2), where n = 12.
σ²(p̂1 − p̂2) = P1Q1/n1 + P2Q2/n2 = (0.4)(0.6)/12 + (0.2)(0.8)/12 = 0.02 + 0.0133 = 0.0333.
What are the steps involved in statistical research?
Steps: choose the topic and state its significance; set the objectives; decide the methodology for data collection (source of data, sampling method, instrument); collect the data; organise and present it in tables and graphs; analyse it with suitable statistical methods; interpret the results and draw conclusions and recommendations.
Define the sampling error and also write its formula.
Sampling error is the difference between a sample statistic and the population parameter it estimates, e.g. sampling error = x̄ − μ.
Properties of hypergeometric distribution.
Each trial gives a success or a failure; sampling is without replacement from a finite population, so the trials are not independent and the probability of success changes from trial to trial; the number of trials n is fixed. Its parameters are N, n and k, with mean nk/N.
Five number summary.
A five-number summary consists of the smallest value X0, the first quartile Q1, the median, the third quartile Q3 and the largest value Xm. It gives a quick idea of the centre, spread and shape (symmetry or skewness) of a distribution and is used to draw a box plot.
What is bias?
Bias of an estimator θ̂ is E(θ̂) − θ. If E(θ̂) ≠ θ the estimator is biased; if the difference is zero it is unbiased.
Advantages and disadvantages of median.
Advantages: easy to understand and calculate; not affected by extreme values; can be found for open-ended classes and located graphically from an ogive; usable for ordinal data. Disadvantages: data must be arranged in order, which is tedious for large data; it does not use all observations; it is less suitable for further algebraic treatment and fluctuates more from sample to sample than the mean.
Mathematical expectation of discrete random variable.
The expected value of a discrete random variable X is the sum of each possible value multiplied by its probability: E(X) = Σ x f(x). It is the mean of the probability distribution.
Any two properties of mathematical expectation.
(1) If c is a constant, E(c) = c. (2) If a and b are constants, E(aX + b) = aE(X) + b. Also E(X + Y) = E(X) + E(Y).
What is a statistical test?
A statistical test is a rule, based on a test statistic computed from sample data, for deciding whether to reject the null hypothesis. The test statistic has a known sampling distribution when H0 is true.
Decide a small sample and large sample.
A sample with n ≤ 30 is usually called a small sample and one with n > 30 a large sample.
In which condition is the Poisson distribution used to approximate the hypergeometric distribution?
When the population is large relative to the sample (n < 0.05N) so that the hypergeometric is close to binomial, and n is large with a small success probability p = k/N; the Poisson mean is then μ = nk/N.
Elaborate the Least Significant Difference (LSD) test.
After ANOVA rejects equality of means, LSD = t(α/2, error df)√(2MSE/r) is computed; each pair of treatment means whose absolute difference exceeds LSD is significantly different.
Write down the formula of combined or pooled proportion of two samples.
p̂c = (X1 + X2)/(n1 + n2) = (n1p̂1 + n2p̂2)/(n1 + n2).
If the approximate value of class interval is 2.96 and range = 14.8, then find the number of classes.
Number of classes = range/class interval = 14.8/2.96 = 5.
Find the coefficient of variation (C.V) for the following price of a commodity. Price (X): 8, 13, 18, 23, 30
x̄ = 92/5 = 18.4; Σx² = 1986; S² = 1986/5 − 18.4² = 397.2 − 338.56 = 58.64; S = 7.66. C.V = (S/x̄) × 100 = (7.66/18.4) × 100 = 41.6%.
Flaws in plywood occur at random with an average of one flaw per 50 square feet. What is the probability that 32 square feet will have no flaws?
Poisson with μ = 32/50 = 0.64. P(X = 0) = e^−0.64 = 0.527.
If X is binomial with n = 5, p = 0.5, q = 0.5, then find SD(X).
SD(X) = √(npq) = √(5 × 0.5 × 0.5) = √1.25 = 1.118.
n = 24, Mean = 33, s = 15, x̄ = 40.4. Compute the t statistic.
Taking μ0 = 33: t = (x̄ − μ0)/(s/√n) = (40.4 − 33)/(15/√24) = 7.4/3.062 = 2.42, with 23 degrees of freedom.
90% confidence interval of population mean is 11 to 20. Interpret the result.
We are 90% confident that the true population mean lies between 11 and 20; about 90% of intervals built this way from repeated samples would contain μ.
E(XY) = 421, E(X) = 42, E(Y) = 15. Check independence.
Cov(X, Y) = E(XY) − E(X)E(Y) = 421 − 630 = −209 ≠ 0, so E(XY) ≠ E(X)E(Y) and X and Y are not independent.
Difference between statistics and statistic.
Statistics is the science of collecting, organising, presenting, analysing and interpreting numerical data (or, in plural sense, the numerical facts themselves). A statistic is a single numerical value computed from a sample, such as the sample mean, used to estimate a population parameter.
Why do we call the standard deviation of a sampling distribution the standard error?
The standard deviation of the sampling distribution of a statistic measures the error (variation) of that statistic as an estimate of the parameter, so it is called the standard error, e.g. S.E.(x̄) = σ/√n.
If E(Y) = 0.5 then find E(10.5 + 2Y).
E(10.5 + 2Y) = 10.5 + 2E(Y) = 10.5 + 1 = 11.5.
What is the meaning of 'b' in Y = a + bX?
b is the slope (regression coefficient): the change in Y for a unit increase in X.
Explain level of significance.
The level of significance α is the probability of rejecting H0 when it is true (Type I error); it is fixed before the test, commonly 0.05 or 0.01.
Can all deciles be expressed as percentiles? Explain.
Yes. Each decile equals a percentile: D1 = P10, D2 = P20, ..., D9 = P90, because Dk divides off 10k percent of the data.
What is meant by sampling distribution?
The sampling distribution of a statistic is the probability distribution of that statistic (such as x̄ or p̂) over all possible samples of the same size from the population.
How do we decide that the drawn sample is a "small sample" or a "large sample"?
If n ≤ 30 the sample is treated as small; if n > 30 it is large.
If E(X) = 0.7 then find E(2X).
E(2X) = 2E(X) = 2 × 0.7 = 1.4.
State Bayes' theorem.
If A1, A2, ..., Ak form a partition of the sample space S (mutually exclusive and exhaustive) and B is any event with P(B) > 0, then P(Ai|B) = P(Ai)P(B|Ai) / Σj P(Aj)P(B|Aj), for i = 1, ..., k.
"The 95% confidence interval for population mean is 1.3 to 4.7". Interpret this result.
We are 95% confident that the population mean lies between 1.3 and 4.7; about 95% of such intervals from repeated samples would contain μ.
If E(XY) = 0.45, E(X) = 0.50 and E(Y) = 0.90, are X and Y independent?
E(X)E(Y) = 0.50 × 0.90 = 0.45 = E(XY), so Cov(X, Y) = 0 and the condition E(XY) = E(X)E(Y) for independence holds. (Strictly, zero covariance alone does not prove independence; it shows X and Y are uncorrelated.)
If we draw a card from an ordinary deck of 52 cards, can king and diamond be mutually exclusive events? Give reason.
No. The king of diamonds is both a king and a diamond, so the two events can occur together and are not mutually exclusive.
Suppose that we toss a fair coin three times. What is its sample space?
S = {HHH, HHT, HTH, THH, HTT, THT, TTH, TTT}, 2³ = 8 sample points.
Why do we take β2 = 3 as the criterion for measuring the kurtosis of any distribution?
For the normal distribution β2 = μ4/μ2² = 3, so the normal (mesokurtic) curve is the standard; β2 > 3 is leptokurtic and β2 < 3 platykurtic.
Define level of significance.
The level of significance α is the probability of rejecting a true null hypothesis (Type I error), fixed before the test.
If the mean of a distribution is 0.925 and the standard deviation is 0.132, find the limits by applying Chebychev's inequality for k = 2. How much fraction of data will lie between the two limits?
Limits: x̄ ± kS = 0.925 ± 2(0.132) = 0.925 ± 0.264, i.e. 0.661 to 1.189. By Chebyshev at least 1 − 1/k² = 1 − 1/4 = 3/4 (75%) of the data lie between these limits.
If E(XY) = 7.5, E(X) = 2.4 and E(Y) = 4.3, calculate the covariance of X and Y.
Cov(X, Y) = E(XY) − E(X)E(Y) = 7.5 − (2.4)(4.3) = 7.5 − 10.32 = −2.82.
Write down the two properties of the sampling distribution of proportion p̂ when the sampling is performed with replacement.
(1) The mean of the sampling distribution equals the population proportion: μ(p̂) = p. (2) Its variance is σ²(p̂) = pq/n, so the standard error is √(pq/n).
What is natural pairing in observations? Give example.
Natural pairing occurs when two measurements are taken on the same unit or individual, e.g. at two different times. Example: weights of ten recruits recorded before and after a training programme; each before-after pair is naturally paired and analysed with the paired t-test.
Write down the significance of experimental design.
Experimental design plans how treatments are assigned to units so that data give valid, objective and efficient conclusions; it controls extraneous variation through randomisation, replication and local control (blocking), so treatment effects can be estimated and tested with minimum error and cost.
Explain the formula of hypergeometric probability distribution.
P(X = x) = [kCx × (N−k)C(n−x)] / NCn, x = 0, 1, ..., min(n, k), where N = population size, n = sample size and k = number of successes in the population. Its three parameters are N, n and k.
Suppose that we toss a fair coin three times. What is the sample space?
S = {HHH, HHT, HTH, THH, HTT, THT, TTH, TTT}.
If X is a Poisson random variable with parameter value 4.5, find P(X = 1).
P(X = 1) = e^−4.5 (4.5)¹/1! = 0.01111 × 4.5 = 0.0500.
When is the value of chi-square equal to zero?
χ² = Σ(o − e)²/e is zero only when every observed frequency equals its expected frequency.
Describe the formula for hypergeometric distribution.
P(X = x) = kCx × (N−k)C(n−x) / NCn, where N is the population size, n the sample size and k the number of successes in the population.
What is graphical representation?
Graphical representation is showing data visually with graphs or charts (such as histograms, frequency polygons, ogives and bar charts) so that patterns, comparisons and relationships are easy to see.
What is the t formula for paired observations?
t = d̄/(sd/√n), where d̄ is the mean of the paired differences, sd = √[Σ(d − d̄)²/(n − 1)] their standard deviation, and n the number of pairs; df = n − 1.
What is the mean and variance of Poisson distribution?
If X has a Poisson distribution with parameter μ, then E(X) = μ and Var(X) = μ.
Explain the Chi-square test of goodness of fit.
It tests whether observed frequencies agree with expected frequencies from a hypothesised distribution, using χ² = Σ(o − e)²/e with k − 1 − m degrees of freedom (k classes, m estimated parameters); a large χ² leads to rejecting the fit.
Define Null and Alternative hypothesis.
The null hypothesis H0 is the statement being tested, usually of no difference or no effect, assumed true until evidence rejects it. The alternative hypothesis H1 is the statement accepted when H0 is rejected; it may be one-sided (< or >) or two-sided (≠).
A random sample of 100 is taken from a population with mean 30 and standard deviation 5. The probability distribution of the parent population is unknown. Find the mean and standard error of the sampling distribution of x̄.
μ(x̄) = μ = 30; S.E.(x̄) = σ/√n = 5/√100 = 0.5. Since n is large, x̄ is approximately normal by the central limit theorem.
Define normal distribution.
A continuous random variable X is normally distributed with mean μ and standard deviation σ if its density is f(x) = [1/(σ√(2π))] e^(−(x − μ)²/(2σ²)), −∞ < x < ∞.
What is natural pairing in observations? Also give an example.
Two measurements on the same unit, e.g. weights of recruits before and after training.
Explain the method of maximum likelihood in point estimation.
Introduced by R. A. Fisher, it takes as the estimate the parameter value that makes the observed sample most probable. Form the likelihood function L(θ) = ∏ f(xi; θ), usually take ln L, differentiate with respect to θ, set it to zero and solve; the solution is the maximum likelihood estimator.
Write down the formula for binomial distribution.
P(X = x) = nCx p^x q^(n−x), x = 0, 1, ..., n, where q = 1 − p.
Write down the formula for testing the equality of two population proportions.
z = (p̂1 − p̂2)/√[p̂c q̂c (1/n1 + 1/n2)], where p̂c = (X1 + X2)/(n1 + n2) and q̂c = 1 − p̂c.
Define moment ratios. In which unit are they expressed?
Moment ratios are b1 = m3²/m2³ (measure of skewness) and b2 = m4/m2² (measure of kurtosis). They are independent of origin and units of measurement, i.e. they are pure numbers with no unit.
Write down the formula of mathematical expectation.
For a discrete random variable E(X) = Σ x f(x); for a continuous random variable E(X) = ∫ x f(x) dx over −∞ to ∞.
Discuss the statistical independence of two discrete random variables.
Two discrete random variables X and Y are statistically independent if their joint probability function equals the product of the marginal ones for every pair: f(x, y) = g(x)h(y), i.e. P(X = x, Y = y) = P(X = x)P(Y = y). Then also E(XY) = E(X)E(Y) and Cov(X, Y) = 0.
Write down the name of methods/techniques that are used to represent the quantitative discrete data.
Quantitative discrete data are presented by an ungrouped (discrete) frequency distribution table and graphically by a line chart / vertical line diagram (stick diagram), in which a vertical line of height equal to the frequency is drawn at each value of X; simple bar charts and pie charts may also be used, and a cumulative frequency step graph shows cumulative frequencies.
A coin is tossed 900 times and heads appear 490 times. State the null and alternative hypotheses to show that the coin is unbiased.
Let p be the probability of a head. H₀: p = 0.5 (the coin is unbiased); H₁: p ≠ 0.5 (the coin is biased). This is a two-tailed test (sample p̂ = 490/900 ≈ 0.544).
Write down the names of two types of experimental designs.
(1) Completely Randomized Design (CRD); (2) Randomized Complete Block Design (RCBD). (A third common design is the Latin Square Design.)
What are the mean and variance of binomial distribution?
For a binomial distribution with parameters n and p (q = 1 − p): Mean = np; Variance = npq (so standard deviation = √(npq)).
Find the probability of drawing a white ball from a bag containing 4 red, 8 black and 3 white balls.
Total balls = 4 + 8 + 3 = 15; favourable (white) = 3. P(white) = 3/15 = 1/5 = 0.2.
Write down the properties of sampling distribution of proportion p̂, when sampling is performed without replacement.
When samples of size n are drawn without replacement from a finite population of size N with proportion p (q = 1 − p): (1) the mean of the sampling distribution of p̂ equals the population proportion, μ_p̂ = p (p̂ is unbiased); (2) its standard error is σ_p̂ = √(pq/n) · √((N − n)/(N − 1)), the factor (N − n)/(N − 1) being the finite population correction; (3) for large n (np and nq both at least 5) the sampling distribution of p̂ is approximately normal.
If X = 255, n = 500, p₀ = 0.60 then find the z-test statistic for proportion.
p̂ = X/n = 255/500 = 0.51. z = (p̂ − p₀)/√(p₀q₀/n) = (0.51 − 0.60)/√(0.60 × 0.40/500) = −0.09/√0.00048 = −0.09/0.02191 ≈ −4.11.
What is the impact of degrees of freedom on chi-square distribution?
The χ² distribution has one parameter, its degrees of freedom ν, and its shape depends entirely on ν. Mean = ν and variance = 2ν, so as ν increases the curve shifts to the right and spreads out. For small ν (1 or 2) it is J-shaped/highly positively skewed; as ν increases the skewness decreases and the curve becomes more symmetric, approaching a normal distribution for large ν.
Describe the Poisson distribution as the limiting form of the binomial distribution.
The binomial probability P(X = x) = ⁿCₓ pˣ qⁿ⁻ˣ tends to the Poisson distribution when the number of trials n → ∞ and the success probability p → 0 in such a way that the mean np = λ stays constant (finite). In that limit P(X = x) = e^(−λ) λˣ / x!, x = 0, 1, 2, …. So for large n and small p (in practice n ≥ 20 and p ≤ 0.05) binomial probabilities can be approximated by Poisson probabilities with λ = np; for example, rare events such as defects or accidents.
Define Disjoint Sets.
Two sets A and B are disjoint (mutually exclusive) if they have no element in common, i.e. A ∩ B = ∅. Example: A = {1, 2, 3} and B = {4, 5} are disjoint. In probability, disjoint events cannot occur together, so P(A ∩ B) = 0.
What is acceptance region?
In hypothesis testing the sample space of the test statistic is divided into two parts. The acceptance region is the set of values of the test statistic that lead to accepting (not rejecting) the null hypothesis H₀; its probability under H₀ is 1 − α. Its complement is the critical (rejection) region, and the two are separated by the critical value(s). For example, for a two-tailed z-test at α = 0.05 the acceptance region is −1.96 < z < 1.96.
If there are K treatments and R rows in a Randomized Complete Block Design then calculate the total number of experimental units used.
In an RCBD every treatment appears exactly once in every block (row). With K treatments and R blocks the total number of experimental units is n = K × R = KR (for example, 4 treatments in 5 blocks need 20 units).
In how many ways a three-person committee can be formed from a group of ten persons? (Use the formula)
Order does not matter in a committee, so use combinations: ⁿCᵣ = n!/(r!(n−r)!). Here ¹⁰C₃ = 10!/(3! × 7!) = (10 × 9 × 8)/(3 × 2 × 1) = 720/6 = 120. So the committee can be formed in 120 ways.
From the given data, calculate mean and standard deviation of sampling distribution of mean if the sampling is done with replacement. N = 120, n = 64, μ = 50, σ = 2
With replacement no finite population correction is needed. Mean of the sampling distribution: μx̄ = μ = 50. Standard deviation (standard error): σx̄ = σ/√n = 2/√64 = 2/8 = 0.25. (N = 120 is not used because sampling is with replacement.)
Construct 90% confidence interval for the difference in means μ₁ − μ₂ in case of paired observations, where d̄ = 1.8, s_d = 1.32, t₀.₀₅(₉) = 1.833
Paired data: t has n − 1 = 9 df, so n = 10 pairs. 90% CI: d̄ ± t₀.₀₅(₉) × s_d/√n = 1.8 ± 1.833 × 1.32/√10 = 1.8 ± 1.833 × 0.4174 = 1.8 ± 0.765. So 1.035 < μ₁ − μ₂ < 2.565 (about 1.04 to 2.57). Assumes s_d is the sample standard deviation of the differences (divisor n − 1).
Name the measures of dispersion that are not based on all the values.
The range (it uses only the largest and smallest values) and the quartile deviation (it uses only Q1 and Q3) are not based on all the observations. The mean deviation, variance and standard deviation use every value.
When we use two-tailed test?
A two-tailed test is used when the alternative hypothesis states only that the parameter differs from the hypothesised value, without a direction (e.g. H0: μ = μ0 vs H1: μ ≠ μ0). The rejection region is split equally between both tails, α/2 in each, so H0 is rejected for values of the test statistic that are too large or too small.
For a sample data n = 15, calculate t_{α/2(ν)} for α = 0.10.
Degrees of freedom ν = n − 1 = 15 − 1 = 14 and α/2 = 0.05. From the t-table, t0.05(14) = 1.761.
Suppose we want to determine the proportions of smokers and non smoker in a city? In this situation what type of distribution we can use?
Each person is either a smoker (success) or a non-smoker (failure), so the situation is binomial: the number of smokers X in a sample of n people follows the binomial distribution b(x; n, p), and the sample proportion p̂ = X/n has mean p and standard error √(pq/n) (approximately normal for large n).
If mean of a distribution is 20 and standard deviation is 2. Find out μ ± 2σ limits by applying empirical rule. What percent of data will lies between these two limits?
μ − 2σ = 20 − 2(2) = 16 and μ + 2σ = 20 + 2(2) = 24. By the empirical rule, approximately 95% of the data lie between 16 and 24.
Suppose X is a random variable having Poisson distribution with its parameter value 3, find value of P(X=1).
For a Poisson distribution P(X = x) = e^(−μ) μ^x / x!. With μ = 3: P(X = 1) = e^(−3) × 3¹ / 1! = 0.049787 × 3 = 0.1494.
If X = 341, n = 634, p0 = 0.50 then find the z-test statistic for proportion.
np0 = 634(0.50) = 317 and √(np0(1 − p0)) = √(634 × 0.5 × 0.5) = √158.5 = 12.59. Since X > np0, use the continuity correction X − ½: Z = (341 − 0.5 − 317)/12.59 = 23.5/12.59 = 1.87. (Without the continuity correction, Z = 24/12.59 = 1.91.)
A random sample of size n is drawn from normal population with mean 5 and variance σ². If s = 2.5, x̄ = 7 and t = 3, then what is the values of n?
t = (x̄ − μ)/(s/√n) ⇒ 3 = (7 − 5)/(2.5/√n) = 2√n/2.5 ⇒ √n = 3 × 2.5/2 = 3.75 ⇒ n = 14.06 ≈ 14.
Is there any difference between distribution function and cumulative distribution function of discrete r.v. X?
No. They are the same thing. The distribution function F(x) = P(X ≤ x) gives the probability that X takes a value less than or equal to x; it accumulates the probabilities from the smallest value up to x, so it is also called the cumulative distribution function (cdf). For a discrete r.v., F(x) = Σ f(xi) over all xi ≤ x, and its graph is a step function.
What is meant by sampling distribution of sample proportion?
If all possible samples of size n are drawn from a population and the proportion of successes p̂ = X/n is computed for each, the probability distribution of these p̂ values is the sampling distribution of the sample proportion. Its mean is μp̂ = p and its standard error is σp̂ = √(pq/n) (with the finite population correction √((N − n)/(N − 1)) when sampling without replacement); for large n it is approximately normal.
While applying the chi-square goodness of fit test, if an expected frequency in any category is less than 5, then what we do to solve this problem.
Combine (pool) that category with an adjacent category (and add their observed frequencies too) until every expected frequency is at least 5. The degrees of freedom are then reduced, since they are computed from the new number of categories: ν = k − 1 − (number of estimated parameters).
Write a short note on Quota Sampling.
Quota sampling is a non-random (non-probability) sampling method. The population is divided into groups (e.g. by sex, age or income), a quota (fixed number) is set for each group, usually in proportion to its size in the population, and the interviewer then selects any units of his choice until each quota is filled. It is quick and cheap and is often used in opinion and market surveys, but because the selection within groups is left to the interviewer it can be biased and sampling error cannot be measured objectively.
Write down the properties of hypergeometric distribution.
A hypergeometric experiment has these properties: (1) a sample of n items is drawn without replacement from a finite population of N items; (2) k of the N items are successes and N − k are failures; (3) the outcome of each draw is success or failure, the trials are not independent and the probability of success changes from draw to draw. The probability is h(x; N, n, k) = C(k, x)·C(N − k, n − x)/C(N, n), x = 0, 1, …, min(n, k). Its mean is nk/N and its variance is n(k/N)((N − k)/N)((N − n)/(N − 1)); when N is large relative to n it approaches the binomial distribution with p = k/N.
Write down critical region for the following hypothesis at 5% level of significance. H0: μ = 75, H1: μ < 75
This is a left-tailed test at α = 0.05. Using the Z statistic Z = (x̄ − 75)/(σ/√n), the critical region is Z ≤ −z0.05 = −1.645: reject H0 if the computed Z is less than or equal to −1.645. (If σ is unknown and n is small, use t ≤ −t0.05(n − 1).)
For the following table we want to test the independence of smoking pattern and marital status. What will be the degrees of freedom for the chi-square test of independence? Smoking pattern (Total self-restraint, Only at times, Regular smoker, Total): Single 67, 213, 74, 354; Married 411, 633, 129, 1173; Widowed 85, 51, 7, 143; Divorced 27, 60, 15, 102; Total 590, 957, 225, 1772.
The table has r = 4 rows (marital status) and c = 3 columns (smoking pattern). Degrees of freedom = (r − 1)(c − 1) = (4 − 1)(3 − 1) = 3 × 2 = 6.
What does quartile deviation measure in a distribution?
Quartile deviation (semi-interquartile range) Q.D = (Q3 − Q1)/2 measures the dispersion (spread) of the middle 50% of the observations around the median. Because it ignores the lowest and highest 25% of values, it is not affected by extreme values and can be used with open-ended classes.
In a random sample of 1000 women in a locality, 224 women said that they use Aerial washing powder. What is the point estimator and point estimate of the proportion of the women who use Aerial washing powder?
Point estimator: the sample proportion p̂ = X/n. Point estimate: p̂ = 224/1000 = 0.224, i.e. about 22.4% of the women use Aerial washing powder.
What is the impact of level of confidence (1−α) on the value of Zα/2?
As the confidence level (1 − α) increases, α decreases and Zα/2 increases (90%: 1.645; 95%: 1.96; 99%: 2.58). A larger Zα/2 gives a wider confidence interval for the same sample size.
Formulate the hypothesis in case of goodness of fit test.
H0: the data follow the specified (hypothesised) distribution, i.e. observed frequencies agree with the expected frequencies. H1: the data do not follow the specified distribution. Test with χ² = Σ(o − e)²/e on (k − 1 − number of estimated parameters) degrees of freedom.
Why the median is suitable average for the below data set? Explain. Monthly income (in rupees) / No. of workers: Less than 2000/-: 100; 2000/- to 2999/-: 300; 3000/- to 3999/-: 500; 4000/- to 4999/-: 250; 5000/- and above: 50.
The first class (less than 2000) and last class (5000 and above) are open-ended, so their mid-points are unknown and the mean cannot be calculated exactly. The median needs only the position of the middle value, so it can be found. N = 1200, N/2 = 600; cumulative frequencies 100, 400, 900, so the median class is 3000–3999 (boundaries 2999.5–3999.5). Median = 2999.5 + (600 − 400)/500 × 1000 = Rs. 3399.5. Income data are also usually skewed, and the median is not affected by extreme values.
If Z is a standard normal random variable with mean 0 and variance 1, then find the Lower quartile.
The lower quartile Q1 satisfies P(Z < Q1) = 0.25, so P(Q1 < Z < 0) = 0.25. From the normal table the z with area 0.25 between 0 and z is 0.6745 (≈ 0.67). Since Q1 is below the mean, Q1 = −0.6745 (≈ −0.67).
The manager of a bottling plant is anxious to reduce the variability in net weight of fruit bottled. Over a long period, the standard deviation has been 15.2 gm. A new machine is introduced and the net weights (in grams) in 10 randomly selected bottles (all of the same nominal weight) are 987, 966, 955, 977, 981, 967, 975, 980, 953, and 972. State null and alternative hypothesis that machine has a better performance? Also write down test-statistics about the hypothesis.
H0: σ² = 231.04 (σ = 15.2 gm; the new machine is no better). H1: σ² < 231.04 (σ < 15.2 gm; the new machine reduces variability). Test statistic: χ² = (n − 1)s²/σ0² = Σ(x − x̄)²/σ0² with n − 1 = 9 df. Here x̄ = 971.3, Σ(x − x̄)² = 1110.1, so χ² = 1110.1/231.04 = 4.80. Left-tail critical value χ²0.95,9 = 3.325; since 4.80 > 3.325, H0 is not rejected at 5%.
In how many branches, estimation can be divided? Also write down the names of the branches.
Estimation is divided into two branches: (1) Point estimation, where a single value computed from the sample (e.g. x̄) is used as the estimate of the parameter; (2) Interval estimation, where a range of values (a confidence interval) is given that is expected to contain the parameter with a stated level of confidence.
A Sample has following values. 101, 99, 111, 199, 150, 140 Find out the maximum likelihood estimate of population mean.
For a normal population the maximum likelihood estimate of μ is the sample mean: μ̂ = x̄ = Σx/n = (101 + 99 + 111 + 199 + 150 + 140)/6 = 800/6 = 133.33.
Write down critical region for the following hypothesis. H0: σ² = 20, H1: σ² < 20. Where, α = 0.01 and n=10
Test statistic χ² = (n − 1)s²/σ0² with ν = n − 1 = 9 degrees of freedom. Since H1 is left-tailed, the critical region is χ² ≤ χ²(1−α)(9) = χ²0.99(9) = 2.088; reject H0 if the computed χ² is less than or equal to 2.088.
What is the concept of extrapolation in regression analysis?
Extrapolation means using the fitted regression line to predict Y for X values lying outside the range of the X values used to fit the line. It is risky because the linear relationship observed within the data range may not hold outside it, so such predictions can be unreliable; prediction within the range is called interpolation.
Find the proportion for the X1 and X2, Where X1 =300 with sample size= n1 = 400 and X2 = 200 with sample size = n2 =300.
Sample proportions: p̂1 = X1/n1 = 300/400 = 0.75 and p̂2 = X2/n2 = 200/300 = 0.667. The pooled (combined) proportion used for testing p1 = p2 is p̂c = (X1 + X2)/(n1 + n2) = 500/700 = 0.714.
Find the mean and variance for the sampling distribution given below: p̂ = 0, 1/3, 2/3, 1 with probabilities 1/20, 9/20, 9/20, 1/20.
Mean = Σp̂ f(p̂) = 0(1/20) + (1/3)(9/20) + (2/3)(9/20) + 1(1/20) = (0 + 3 + 6 + 1)/20 = 0.5. E(p̂²) = (1/9)(9/20) + (4/9)(9/20) + 1(1/20) = 6/20 = 0.3. Variance = 0.3 − 0.5² = 0.05.
Which of the following statements represents continuous data and discrete data? (i) Number of shops in a plaza. (ii) Hourly temperature recorded by weather bureau. (iii) Inches of rainfall in a city. (iv) Number of passengers carried by rail every year. (v) Height measurements of boys studying in a college.
(i) Discrete (count). (ii) Continuous. (iii) Continuous. (iv) Discrete (count). (v) Continuous, because height is measured and can take any value in a range.
Two different dice are thrown. Make the sample space and also find the number of sample points in an event A that the sum is 7.
S = {(i, j): i, j = 1, 2, ..., 6}, i.e. (1,1), (1,2), ..., (6,6), giving 6 × 6 = 36 sample points. A = {(1,6), (2,5), (3,4), (4,3), (5,2), (6,1)}, so A has 6 sample points (P(A) = 6/36 = 1/6).
The probability of one electric device failure is 0.01. If a sample of 400 is chosen, what is the probability that exactly 2 are defective?
Binomial: P(X = 2) = 400C2 (0.01)² (0.99)^398 = 79800 × 0.0001 × 0.0183 = 0.146. (Poisson approximation with μ = np = 4: e^−4 4²/2! = 0.147.)
Two dice are rolled. Find the probability that the outcome is equal to or more than 11.
Sum ≥ 11: (5,6), (6,5), (6,6), i.e. 3 of 36 outcomes. P = 3/36 = 1/12 = 0.083.
Interpret the concept of five number summary and also explain the purpose of five number summaries.
The five-number summary is X0 (minimum), Q1, median, Q3 and Xm (maximum). Purpose: it shows centre, spread and shape and is the basis of the box plot. For symmetric data Q1 and Q3 are equidistant from the median and X0 and Xm are equidistant from it as well; for right-skewed data the distance Q3 to Xm exceeds X0 to Q1 (and median < mid-quartile range < midrange); for left-skewed data the reverse holds.
What is the probability that a poker hand of 5 cards contains exactly 2 aces (hypergeometric distribution)?
P(X = 2) = 4C2 × 48C3 / 52C5 = (6 × 17296)/2598960 = 103776/2598960 = 0.0399.
What is the probability that at least one head occurs if a coin is tossed 6 times successively?
P(at least one head) = 1 − P(no head) = 1 − (1/2)^6 = 1 − 1/64 = 63/64 = 0.984.
From the table given below find the value of chi square. Observed frequencies: 120, 130, 80, 170; expected frequencies: 100, 150, 100, 150.
χ² = Σ(o − e)²/e = 400/100 + 400/150 + 400/100 + 400/150 = 4 + 2.667 + 4 + 2.667 = 13.33.
Describe the main steps of the general procedure for testing hypothesis.
(1) Formulate H0 and H1. (2) Choose the level of significance α. (3) Choose the test statistic and its distribution under H0. (4) Compute the test statistic from the sample. (5) Find the critical region from α and the alternative. (6) Conclude: reject H0 if the statistic falls in the critical region, otherwise do not reject it.
A candidate for mayor in a large city believes that he appeals to at least 10 per cent more of the educated voters than the uneducated voters. He hires the services of a poll-taking organization, and they find that 62 of 100 educated voters interviewed support the candidate, and 69 of 150 uneducated voters support him. Test his belief at the 0.05 significance level.
H0: p1 − p2 ≥ 0.10; H1: p1 − p2 < 0.10; α = 0.05. p̂1 = 0.62, p̂2 = 0.46. z = (0.62 − 0.46 − 0.10)/√(0.62×0.38/100 + 0.46×0.54/150) = 0.06/√(0.002356 + 0.001656) = 0.06/0.0633 = 0.95. Critical region z < −1.645. Since 0.95 is not in it, H0 is not rejected; the data support the candidate's belief.
If we have RCBD with MSE = 3.19, number of treatments = 4, number of blocks = 5, then find the value of LSD (least significant difference) for treatments by using α = 0.05 and error degrees of freedom is 12.
LSD = t(0.025, 12) √(2MSE/r) = 2.179 × √(2 × 3.19/5) = 2.179 × √1.276 = 2.179 × 1.1296 = 2.46. Any two treatment means differing by more than 2.46 are significantly different.
The following data was obtained for a randomized block design involving five treatments and three blocks: SST = 430, SSTR = 310, SSB = 85. Set up the ANOVA table.
SSE = 430 − 310 − 85 = 35. Treatments: df 4, SS 310, MS 77.5, F = 77.5/4.375 = 17.71. Blocks: df 2, SS 85, MS 42.5, F = 42.5/4.375 = 9.71. Error: df (5−1)(3−1) = 8, SS 35, MS 4.375. Total: df 14, SS 430.
In a standard normal distribution, find: mean and standard deviation, lower quartile, upper quartile, inter-quartile range, mean deviation.
Mean = 0, S.D = 1. Q1 = μ − 0.6745σ = −0.6745. Q3 = μ + 0.6745σ = 0.6745. IQR = Q3 − Q1 = 1.349 (quartile deviation = 0.6745). Mean deviation = 0.7979σ = 0.7979.
Find quartile deviation for the data given below: 18, 53, 45, 28, 39, 29, 23, 40 and 21
Arrange in order: 18, 21, 23, 28, 29, 39, 40, 45, 53 (n = 9). Q₁ = value of the (n + 1)/4 = 2.5th observation = 21 + 0.5(23 − 21) = 22. Q₃ = value of the 3(n + 1)/4 = 7.5th observation = 40 + 0.5(45 − 40) = 42.5. Quartile deviation = (Q₃ − Q₁)/2 = (42.5 − 22)/2 = 10.25.
A population consists of N=5 values 1, 2, 3, 5, 6. A sample size of n=3 is selected from the population without replacement, calculate sampling distribution of sample proportions for even numbers.
Even numbers in the population: 2 and 6, so p = 2/5 = 0.4. Number of samples = ⁵C₃ = 10: {1,2,3} 1/3, {1,2,5} 1/3, {1,2,6} 2/3, {1,3,5} 0, {1,3,6} 1/3, {1,5,6} 1/3, {2,3,5} 1/3, {2,3,6} 2/3, {2,5,6} 2/3, {3,5,6} 1/3. Sampling distribution of p̂: p̂ = 0 with probability 1/10, p̂ = 1/3 with probability 6/10, p̂ = 2/3 with probability 3/10. Check: E(p̂) = 0(0.1) + (1/3)(0.6) + (2/3)(0.3) = 0.4 = p; Var(p̂) = E(p̂²) − 0.4² = 0.2 − 0.16 = 0.04 = (pq/n)(N − n)/(N − 1) = (0.24/3)(2/4).
Write down the testing procedure in case of goodness of fit test.
(1) State H₀: the data follow the specified distribution (the observed frequencies agree with the expected ones); H₁: the data do not follow it. (2) Choose the level of significance α. (3) Compute expected frequencies eᵢ = n·pᵢ from the hypothesised distribution (combine classes so each eᵢ is at least 5). (4) Test statistic χ² = Σ (oᵢ − eᵢ)²/eᵢ, which follows approximately a chi-square distribution with ν = k − 1 − r degrees of freedom (k classes, r parameters estimated from the data). (5) Critical region: reject H₀ if computed χ² ≥ χ²α(ν) (right-tailed test). (6) Compute χ² and draw the conclusion: if it falls in the critical region reject H₀ (poor fit), otherwise do not reject (good fit).
If we have n=634 and p̂ =0.459, where Z_(0.01/2) = 2.58, then find the 99% confidence interval for population proportion.
99% CI: p̂ ± Z_(α/2) √(p̂q̂/n). q̂ = 1 − 0.459 = 0.541. SE = √(0.459 × 0.541/634) = √(0.248319/634) = √0.00039167 ≈ 0.01979. Margin = 2.58 × 0.01979 ≈ 0.0511. CI = 0.459 ± 0.0511 = (0.408, 0.510), i.e. 0.408 ≤ p ≤ 0.510.
A man travels on car from Lahore to karachi on Motor way in 8 stages of equal intervals. The speed of the car in the various stages was observed to be 110, 116, 120, 114, 115, 112, 120, 117 kilometers per hour. Find the average speed at which the car travels.
The stages cover equal distances, so the correct average of speeds is the harmonic mean H = n/Σ(1/x). Σ(1/x) = 1/110 + 1/116 + 1/120 + 1/114 + 1/115 + 1/112 + 1/120 + 1/117 = 0.009091 + 0.008621 + 0.008333 + 0.008772 + 0.008696 + 0.008929 + 0.008333 + 0.008547 = 0.069321. H = 8/0.069321 ≈ 115.40 km per hour. So the average speed of the car is about 115.4 km/h.
If n = 1150, x = 450, p = 0.39, H₀: p₀ = 0.5, and α = 0.05. Test the stated hypothesis. (Use table value of z = ±1.96)
1) H₀: p = 0.5 against H₁: p ≠ 0.5 (two-tailed, as the table value ±1.96 shows). 2) α = 0.05. 3) Test statistic z = (p̂ − p₀)/√(p₀q₀/n). 4) Computation: p̂ = 450/1150 = 0.39; √(0.5 × 0.5/1150) = √0.0002174 = 0.01474; z = (0.39 − 0.5)/0.01474 = −0.11/0.01474 ≈ −7.46. 5) Critical region: |z| > 1.96. 6) Conclusion: |−7.46| > 1.96, so reject H₀; the population proportion is significantly different from 0.5 (it is lower).
A personal manager is interested in trying to determine whether absenteeism is greater on some specific day of the week or not? His records for the past year show the following sample distributions. Day of the week: Monday, Tuesday, Wednesday, Thursday, Friday; No. of absentees: 66, 57, 54, 48, 75. Test goodness of fit. Where test statistic chi-square = Σ(oᵢ − eᵢ)²/eᵢ = 7.50
1) H₀: absenteeism is equally distributed over the five days (uniform); H₁: it is not. 2) α = 0.05 (not given; usual level assumed). 3) Total = 66 + 57 + 54 + 48 + 75 = 300, so eᵢ = 300/5 = 60 for each day. 4) χ² = Σ(oᵢ − eᵢ)²/eᵢ = (36 + 9 + 36 + 144 + 225)/60 = 450/60 = 7.50 (matches the given value). 5) df = k − 1 = 5 − 1 = 4; critical value χ²₀.₀₅(₄) = 9.488. 6) Since 7.50 < 9.488, do not reject H₀: there is no significant evidence that absenteeism is greater on any particular day.
A random variable X is normally distributed with μ = 50 and σ² = 25. Find the probability of X larger than 54.
σ = √25 = 5. Z = (54 − 50)/5 = 0.8. P(X > 54) = P(Z > 0.8) = 0.5 − P(0 < Z < 0.8) = 0.5 − 0.2881 = 0.2119.
The given data is n = 1150, x = 450, p = 0.39, H0: p0 = 0.3, α = 0.01. Test the stated hypothesis. (Use table value of z = ±2.58)
H0: p = 0.30 vs H1: p ≠ 0.30 (two-tailed, since ±2.58 is given), α = 0.01. Test statistic Z = (p̂ − p0)/√(p0q0/n). p̂ = 450/1150 = 0.391, √(0.3 × 0.7/1150) = √0.0001826 = 0.01351. Z = (0.391 − 0.30)/0.01351 = 6.76 (6.66 if p̂ = 0.39 is used). Critical region: |Z| ≥ 2.58. Since 6.76 > 2.58, reject H0: the population proportion differs significantly from 0.30 (it is greater).
Given, n1 = n2 = 16, s1² = 50, s2² = 16. Construct a 90% confidence interval for the variance ratio σ1²/σ2².
ν1 = ν2 = 15, α = 0.10, F0.05(15,15) = 2.40. s1²/s2² = 50/16 = 3.125. 90% CI: (s1²/s2²)·1/F0.05(15,15) < σ1²/σ2² < (s1²/s2²)·F0.05(15,15) ⇒ 3.125/2.40 < σ1²/σ2² < 3.125 × 2.40 ⇒ 1.30 < σ1²/σ2² < 7.50.
If X̄ is normally distributed with μx̄ = μ = 24000 and σx̄ = 412.20, then find out the P(X̄ > 24500).
Z = (X̄ − μx̄)/σx̄ = (24500 − 24000)/412.20 = 1.21. P(X̄ > 24500) = P(Z > 1.21) = 0.5 − 0.3869 = 0.1131 (about 11.3%).
If we have X̄ = 15 and Σ(Xi − X̄)² = 56 (sum over i = 1 to 10), then find the 90% confidence interval for population variance (σ²).
n = 10, df = 9. The 90% CI for σ² is Σ(Xi − X̄)²/χ²0.05,9 < σ² < Σ(Xi − X̄)²/χ²0.95,9, with χ²0.05,9 = 16.919 and χ²0.95,9 = 3.325. So 56/16.919 < σ² < 56/3.325, i.e. 3.31 < σ² < 16.84. (X̄ = 15 is not needed.)
A research worker wishes to estimate the mean of a population using a sample sufficiently large that the probability will be 0.95 that the sample mean will not differ from the true mean by more than 25 percent of the standard deviation. How large a sample should be taken?
We need P(|X̄ − μ| ≤ 0.25σ) = 0.95, so z0.025·σ/√n = 0.25σ, i.e. 1.96/√n = 0.25. √n = 1.96/0.25 = 7.84, n = 61.47. Rounding up, a sample of n = 62 should be taken.
Find the coefficient of standard deviation from the following data: Life in Hours (X): 130, 150, 180, 250, 345
n = 5, ΣX = 1055, X̄ = 211. Σ(X − X̄)² = 6561 + 3721 + 961 + 1521 + 17956 = 30720. S = √(30720/5) = √6144 = 78.38 hours. Coefficient of standard deviation = S/X̄ = 78.38/211 = 0.371. (Using the divisor n − 1: s = 87.64 and coefficient = 0.415.)
What is the probability that a poker hand of 5 cards contain (i) exactly 2 aces (ii) exactly 1 ace? (Use hypergeometric distribution)
N = 52, k = 4 aces, n = 5. P(X = x) = C(4,x)·C(48,5−x)/C(52,5), C(52,5) = 2598960. (i) P(X = 2) = C(4,2)·C(48,3)/C(52,5) = 6 × 17296/2598960 = 103776/2598960 = 0.0399. (ii) P(X = 1) = C(4,1)·C(48,4)/C(52,5) = 4 × 194580/2598960 = 778320/2598960 = 0.2995.
If n = 22, Σ(d − d̄)² = 270, d̄ = 4. Then, using a 0.05 level of significance level, test the hypothesis H0 : μd = 0 against H1 : μd ≠ 0
Paired t-test. s_d² = Σ(d − d̄)²/(n − 1) = 270/21 = 12.857, s_d = 3.586. t = d̄/(s_d/√n) = 4/(3.586/√22) = 4/0.7645 = 5.23 with 21 df. Two-tailed critical value t0.025,21 = 2.080. Since 5.23 > 2.080, reject H0: the mean difference is significantly different from zero.
A company launched new layout of its website. After a survey, 62 of 115 visitors liked the new layout while 59 of 135 visitors liked the old layout of the website. Company claims that new layout did not improve the visitor's liking about the website. By using the critical value Z < −Z0.05 = −1.645, verify the company's claim: H0 : P1 − P2 ≥ 0, H1 : P1 − P2 < 0
p̂1 = 62/115 = 0.539, p̂2 = 59/135 = 0.437, pooled p̂ = (62 + 59)/(115 + 135) = 121/250 = 0.484, q̂ = 0.516. SE = √[p̂q̂(1/115 + 1/135)] = √[0.484 × 0.516 × 0.016104] = 0.0634. Z = (0.539 − 0.437)/0.0634 = 1.61. Critical region: Z < −1.645. Since 1.61 > −1.645, H0 is not rejected at 5%: there is no evidence that the new layout is liked less (P1 < P2). Note that 1.61 is also below +1.645, so the increase in liking is not significant at 5% either; the data do not show a significant improvement.
A certain type of storage battery lasts on the average 3.0 years, with a standard deviation of 0.5 year. Assuming that the battery lives are normally distributed, find the probability that a given battery will last less than 2.3 years.
μ = 3.0, σ = 0.5. z = (x − μ)/σ = (2.3 − 3.0)/0.5 = −1.4. P(X < 2.3) = P(Z < −1.4) = 0.5 − 0.4192 = 0.0808. So about 8.08% of batteries last less than 2.3 years.
For the Given information, what is your conclusion in testing the indicated null hypothesis? n1 = 13, n2 = 41, s1² = 6.3, s2² = 15.6, H0: σ1² = σ2² and H1: σ2² > σ1²
H0: σ1² = σ2² vs H1: σ2² > σ1² (right-tailed). Level of significance not stated; take α = 0.05. Test statistic F = s2²/s1² = 15.6/6.3 = 2.476 with ν1 = n2 − 1 = 40 and ν2 = n1 − 1 = 12 df. Critical value F0.05(40,12) = 2.43. Since 2.476 > 2.43, reject H0 and conclude that σ2² is greater than σ1² at the 5% level. (At α = 0.01, F0.01(40,12) ≈ 3.43, so H0 would not be rejected.)
In a random sample of 200 persons having their lunch at the University cafeteria on meatless day it was observed that 30 percent preferred vegetable dishes. Find 95% confidence interval for p.
p̂ = 0.30, q̂ = 0.70, n = 200. SE = √(p̂q̂/n) = √(0.21/200) = √0.00105 = 0.0324. 95% CI: p̂ ± 1.96·SE = 0.30 ± 1.96(0.0324) = 0.30 ± 0.0635, i.e. 0.2365 < p < 0.3635 (about 23.6% to 36.4%).
Answer the following questions: (i) Which method has been used for colleting the data in the following case? "An investigator collects the information personally from the individuals concerned". (ii) Find the sampling error when the population mean is 15.785 and sample mean is 15.6. (iii) Write down any two types of nonrandom sampling. (iv) What is the singular of strata?
(i) Direct personal investigation (a method of collecting primary data). (ii) Sampling error = x̄ − μ = 15.6 − 15.785 = −0.185 (magnitude 0.185). (iii) Non-random sampling types: quota sampling, purposive (judgment) sampling, convenience sampling (any two). (iv) The singular of strata is stratum.