You are on page 1of 10

Maths T

Assignment Viva
Title : Sampling
Distributions

Prepared by :
zMarcus Ng Zi Jian

Examinate by :
Mr Suhaimi Aziz
Introduction
z
Confidence Interval
• Range of values we are fairly sure our true value lies in
• set of numbers bounded above and below the mean of a statistic

Sample Size
• Number of participants or observations included in a study.
• A group of participants chosen from the general public who are regarded
to representative of the real population for that specific study
Normal Distribution
z
▪ Continuous probability distribution that is symmetrical around its mean

▪ Graphically represented as a "bell curve"

Chi-Squared Distribution
• Degree-of-freedom continuous distribution
• It describes the distribution of a sum of squared random variables
• To evaluate the goodness of fit of a data distribution, determine if data
series are independent
Methodology
z
Task 1
1. Generate 500 observations by using Microsoft Excel spreadsheets for
the normal distribution with mean 10, variance 2.
2. Generate the x-values with inverse normal distribution formula and copy
the data values to the next sheet.
3. Find the range and calculate the probability, cumulative frequency and
frequency of the data values.
4. Plot the histogram by using the range and probability.
5. Repeat step 1 to 4 for mean 10, variance 4.
6. Next, generate 500 observations for Chi-Squared distribution with
degree of freedom=1.
7. Repeat step 2 to 4 for this Chi-Squared distribution.
8. Repeat step 6 to 7 by changing the degree of freedom to 10 and 30.
Task 2
1. Generate
z n=10 observations for every distribution in Task 1 by using
Microsoft Excel spreadsheets.
2. Calculate the sample size, alpha, standard deviation, mean, confidence
level and confidence interval for step 1.
3. Generate 200 sets of data and confidence intervals for step 1 and 2.
4. Repeat step 1 to 3 for n=30 and n=50.

Task 3
1. Compare every distribution and data of their sample size, standard
error, estimation error and coverage probability that are discussed for
the findings on the coverage probability of the confidence intervals for
the population means.
2. Tabulate all these findings.
z
Results
SET SAMPLE SIZE CONFIDENCE INTERVAL OF ESTIMATION STANDARD ERROR COVERAGE
PROPORTION ERROR, r PROBABILITY
NORMAL 10 (0.89706,0.90294) 0.003 0.0015 0.9
DISTRIBUTION WITH 30 z (0.912267,0.917733) 0.0025 0.00132 0.915
MEAN 10 AND 8
VARIANCE 2
50 (0.942766,0.947234) 0.945
0.00214 0.00109
NORMAL 10 (0.942766,0.947234) 0.00266 0.00136 0.945
DISTRIBUTION WITH 30 (0.947864,0.952136) 0.0025 0.001276 0.95
MEAN 10 AND 50 (0.947863,0.952136) 0.00214 0.00109 0.95
VARIANCE 4

CHI-SQUARED 10 (0.755815,0.764185) 0.00399 0.00204 0.76


DISTRIBUTION WITH
1 DEGREE of
FREEDOM 30 (0.907195,0.912805) 0.00307 0.00156 0.91
50 (0.942766,0.947234) 0.00242 0.00123 0.945

CHI-SQUARED 10 (0.907195,0.912805) 0.0028 0.00143 0.91


DISTRIBUTION WITH
10 DEGREE OF
FREEDOM 30 (0.932584,0.937416) 0.00242 0.00123 0.935

50 (0.912267,0.917733) 0.00203 0.00104 0.915

CHI-SQUARED 10 (0.891996,0.898004) 0.0028 0.00143 0.895


DISTRIBUTION WITH
30 DEGREE OF
FREEDOM 30 (0.922419,0.927581) 0.0025 0.000128 0.925
50 (0.963199,0.966801) 0.00223 0.00114 0.965
z

▪ The result in the previous slide


shows that when sample size is
increasing from n=10 to n=50, the
confidence interval is getting closer
and the estimation error becomes
smaller

▪ This is because when sample size


increase, the estimation error will
decrease.
z
Conclusion

Sample sizes of any population


distribution will affect the coverage
probabiltiy. As the sample size
increase, the coverage probability will
approach the confidence interval with
smaller standard error.
z Thank you

You might also like