Friday, 13 December 2019

ESTIMATION and SIGNIFICANCE of SAMPLING

ESTIMATION AND TEST OF SIGNIFICANCE.
----------+------++++++++++------------------

Approximate Confidence Limits(Large Samples any Distribution)

1) for Mean:
• 95% C.L= mean ±1.96(S.E of mean
•99% C.L= mean ±2.58(S.E of mean
• almost sure limits= mean ±3(S.E of mean.

2) for Proportion:
• 95% C.L= p ±1.96(S.E of p)
• 99% C.L= p ±2.58(S.E of p)
• almost sure =p ±3(S.E of p)

3) for difference of mean:
4) for difference of proportions:

             EXERCISE

1) A sample of 600 screws is taken from a large consignment and 75 are to be found to be defective. Estimate the percentage of defective in the consignment and assign limits within the percentage lies. 

12.5%, 8.45% to 16.55%


2) A random sample of 100 ball bearing selected from a shipment of 2000 ball bearings has an average diameter of 0.354 inch with a S. D is .048 inch. Find 95% c.l for the average diameter of these 2000 ball bearings.                       0.3448 to 0.3632 inch.


3) A random sample of 100 article from a large batch of articles contains 5 defective.
A) 96% of c.l.      0.95 to .005


B) if the batch contains 2696 articles set up 95% c.l for the proportion of defective.  

                                .008 to.092


4) 10 Life insurance policies in a sample of 200 taken out of 50000 were found to be insured for less than 5000. How many policies can be reasonably expected to be insured for less than Rs5000 in the whole lot of 95% C.L.                              1000 to 4000


5) A random sample of size 10 was drawn from a normal population with an unknown mean and a variance of 44.1 . If the observations are: 65,71,80,76,78,82,68,72,65,81 inches, obtain the 95% confidence interval for the position mean.   69.7 to 77.9


DEFINATION OF USEFUL TERMS:
-------------- ****** -------------
* STATISTICAL HYPOTHESIS:
Any statement or assertion about a statistical population or the values of its parameters is called a STATISTICAL HYPOTHESIS. There are two types of hypothesis 

a) simple

b) composite.


* SIMPLE HYPOTHESIS:
A statistical hypothesis which specifies the population completely (i e. the probability distribution and all perimeter are known) is simple hypothesis.


* COMPOSITE HYPOTHESIS:
A statistical hypothesis which does not specify the population completely (i.e, either the form of probability distribution or some parameters remains unknown) 


* TEST OF HYPOTHESIS (OR TEST OF SIGNIFICANCE) :


A test of Hypothesis is a procedure which Specifies a set of "rules for decision" whether to "accept" or 'reject' the Hypothesis under consideration (i.e., null hypothesis).


* NULL HYPOTHESIS:
A statistical hypothesis which is set up (i.e., assumed) and whose validity is tested for possible rejection on the basis of sample observations is called NULL HYPOTHESIS. It is denoted by H₀ and tested against alternatives. Tests of hypothesis deal with rejection or acceptance of null hypothesis only.
ALTERNATIVE HYPOTHESIS:
A statistical hypothesis which differs from the null hypothesis is called an Alternative Hypothesis, and is denoted by H₁ . The alternative Hypothesis is not tested, but it's acceptance (rejection) depends on the rejection (acceptance) of the null hypothesis. Alternative Hypothesis contradicts the null hypothesis. The choice of an appropriate critical region depends on the type of alternative Hypothesis, viz. Whether both-sided, one-sided (right) left) or specified alternative.


* TEST STATISTICS:
A function of sample observations (i.e., Statistics) whose computed value determines the final decision regarding acceptance or rejection of H₀ is called a TEST STATISTICS. The appropriate test statistic has to be chosen very carefully and a knowledge of its sampling distribution under H₀(i.e., when the null hypothesis is true) is essential in framing the decision rules. If the value of the test statistic falls in the critical region, the null hypothesis is rejected.


* CRITICAL REGION:
The set of values of the test statistic which lead to rejection of the null hypothesis is called CRITICAL REGION of the test. The probability with which a true Hypothesis is rejected by the test is often referred to as "size" of the CRITICAL REGION. Geometrically, a sample x₁ , x₂ , .... xₙ of size n is looked upon as just a point x, called SAMPLE POINT, within the region of all possible Samples, called the SAMPLE SPACE (W). The critical region is then defined as a subset (w) of those sample points which lead to the rejection of the null hypothesis.


*  LEVEL OF SIGNIFICANCE:
The maximum probability with which a true null hypothesis is rejected is known as LEVEL OF SIGNIFICANCE of the test, and is denoted by α. In framing decision rules, the level of significance is arbitrary chosen in advance depending on the consequences of a statistical decision. Customarily, 5% or 1% level of significance is taken, although other levels such as 2% or 1/% is also used. The level of significance α is used to indicate the upper limit of the probability of committing Type I error, i.e., the size of the critical region.


* TYPE I ERROR (or ERROR OF FIRST KIND):
This is the error committed in rejecting a null hypothesis by the test when it is really true. The critical region is so determined that the probability of Type I error does not exceed the level of significance of the test.


* TYPE II ERROR (or ERROR OF SECOND KIND). 

This is the error committed in accepting a null hypothesis by the test when it is really false. The probability of Type II error depends on the specified value of the alternative hypothesis, and is used in evaluating the efficiency of a test.


* POWER:
The probability of rejecting a false Hypothesis is called POWER OF THE TEST. Therefore Power is the hypothesis is false. For a specified value of the parameter consistent with the alternative hypothesis.
Power = 1 - probability of Type II error.
If the power is graphically plotted against all specified alternatives, the curve obtained is known as POWER CURVE. The power of a test Determines between true and false null hypothesis.


STEPS IN TEST OF SIGNIFICANCE:


1) SET UP the " null hypothesis" H₀ and the "Alternative Hypothesis" H₁ on the basis of the given Problem. The null hypothesis usually specifies the value of some parameters involved in the population: H₀(θ = θ₀). The alternative hypothesis may be any one of the following types:
H₁(θ≠θ₀), H¹(θ>θ₀), H₁(θ≺θ₀). The type of alternative hypothesis determine whether to use a two-tailed or one-tailed test(right or left).


2) STATE the appropriate "test statistic" T and also its sampling distribution, when the null hypothesis is true. In large sample tests the statistics z= (T - θ₀)/ S.E.(T), which approximately follows Standard Normal distribution, is often used. In small sample tests, the population is assumed to be Normal and various test Statistics are used which follow Standard Normal, Chi-square, t of F distribution exactly.


3) Select the " level of significance"  α of the test, it is not specified in the given Problem. This represents the maximum probability of committing a Type I error, i.e., of making a wrong decision by the test procedure when in fact the null hypothesis is true. Usually, a 5% or 1% level of significance is used (If nothing is mentioned, use 5% level).


4) Find the 'critical region' of the test at the chosen level of significance, This represents the set of values of the test statistic which lead to rejection of the null hypothesis. The critical region always appears in one or both tails of the distribution, depending on whether the alternative hypothesis is one-sided or both-sided. The area in the tail (called 'size of the critical region') must be equal to the level of significance α. For a one-tailed test, α appears in one tail and for a two-tailed test α/2 appears in each tail of the distribution. The critical region is
T≥Tα/₂. or T≤T₁₋α/₂ when H₁(θ≠θ₀)
                  T≥Tα       when H₁(θ≻θ₀)
                  T≤T₁₋α    when H₁(θ≺θ₀)
Where Tα is the value of T such that the area to its right is α.


5) Compute the value of the test statistic T on the basis of sample data and the null hypothesis. In large sample tests, if some parameters remains unknown they should be estimated from the sample.


6) If the computed value of tests Statistics T lies in the critical region, " reject H₀"; otherwise " do not reject H₀. The decision regarding rejection or otherwise of H₀ is made after a comparison of the completed value of T with the critical value (i.e., boundary value of the appropriate critical region).


7) Write the conclusion in plain non-technical language. If H₀ is rejected, the interpretation is: "the data are not consistent with the assumption that the null hypothesis is true and hence H₀ is not tenable" . If H₀ is not rejected, " the data cannot provide any evidence against the null hypothesis and hence  H₀ may be accepted to be true". The conclusion should preferably be given in the words stated in the Problem.

* LARGE SAMPLE TESTS*
The following tests of significance are valid only for large sample sizes. These tests are also called " approximate test", because the sampling distribution used are only approximately true, when the number of observations in the sample is large (say, more than 30). The accuracy however increases with larger sample sizes.


*** FOR USING NORMAL DISTRIBUTION****


1) TEST FOR A SPECIFIED PROPORTION:
A random sample of size n shows that the Proportion of members possessing a certain attribute (e.g., defectives) is p. It is required to test the hypothesis that the proportion P in the population has a specified value P₀      
             H₀ (P =P₀)
If the sample size n is large, we use the test statistic
    z = (p - P₀/(S.E of p)
Which approximately follows Standard Normal distribution.
For testing significance at 5% level, the rules are as follows:
(i) If the alternative hypothesis is that the population proportion P is  'different' from P₀. reject H₀  when the value of z lies outside the range - 1.96 to 1.96.

H₁(P≠P),  Critical Region |z| ≥ 1.96
(ii) If the alternative hypothesis is that the population proportion P is ′greater′ than P₀, reject H₀ when the value of z is greater than 1.645.
H₁(P > P₀),Critical Region z≥1.645
(iii) If the alternative hypothesis is that the population proportion P is ′′ less′′ than P₀ , reject H₀ when the value of z is less than - 1.645,
H₁ (P Otherwise, do not reject the null hypothesis H₀.

For  testing at 1% level, the value 1.96, 1.645 and -1.645, given above, should be replaced by 2.58, 2.33 and -2.33 respectively.

   Rejection Rules for H₀(P = P₀)

Alternative  Critical region

hypothesis  5%level  1%level
   (H₁)

P≠ P₀        |z|≥ 1.96  |z|≥2.58
P>P₀          z≥ 1.645     z≥2.33
P

Confidence limits for the population proportion P are
95% C. L = p± 1.96 (S .E of p)
99% C. L = p± 2.58 (S .E of p)
Where S. E of p = √(p q/n) if H₀ is rejected.

                  

               EXERCISE
               ----------------

1) A die was thrown 400 times & 'six' resulted 80 times. Do the justify the hypothesis of an unbiased die ?
     *die may be unbiased


2) In a big city 325 men out of 600 were found to be smokers. Does this information support the conclusion that the majority of men in this city are smokers?  

*Yes


3) A manufacturer claimed that atleast 90% of the components which he supplied, confirmed to speculations. A random sample of 200 components showed that only 164 were upto the standard. Test his claim at 1% level of significance.
*Manufacturer's claim is not justified.


4) In a large city A, 20% of a random sample of 900 school children had defective eye sight. In another large city B, 15 % of a random sample of 1600 children had the same defect. Is this difference between the two Proportion significant? Obtain 95% confidence limits for the difference in the population proportions.      

* Significant.


5) A machine produced 20 defective articles in a batch of 400. After overhauling it produced 10 defective in a batch of 300. Has the machine improved.
* Machine has not improved after overhauling.


6) In a certain city 100 men in a sample of 400 were found to be smokers. In another city, the number of smokers was 300 in a random sample of 800. Does this indicate that there is a greater proportion of smokers in the second city than in the first ?
* Proportion of smokers is greater in the second city than in the first.






No comments:

Post a Comment