0% found this document useful (0 votes)
14 views5 pages

Confidence Intervals & Hypothesis Testing Guide

Uploaded by

conlobmac
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
14 views5 pages

Confidence Intervals & Hypothesis Testing Guide

Uploaded by

conlobmac
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

STAT 224 / 231 Assignment #4 Confidence Intervals and Hypothesis Testing

1. A random sample of a distribution (42 samples drawn) yields a sample mean value of 102.4 and
a standard deviation of 8.6.
a) Calculate the 95% confidence interval
b) Calculate the 99% confidence interval
c) Calculate the 90% confidence interval
2. What Z scores correspond to the boundaries of a 99.9% confidence interval?
3. Name two conditions that must be met in order to create a valid confidence interval for the true
population mean.
4. You sample a population and get the following set of measurements: [1.4 1.2 1.1 0.6 0.9 0.5 1.4
5.6]
a) Calculate the standard error of the mean using the sample’s standard deviation
b) If the underlying population had a true mean value of 1.04, calculate the sample’s t statistic.
c) Given that you don’t know the population’s underlying standard deviation, calculate the
95% confidence interval on the mean.
d) Sampling is performed on a totally new underlying distribution. Repeated sampling results in
an estimate of the standard error of the mean of 0.5. If a given sample yielded a sample
standard deviation of 50, and this value reflects the underlying population’s true standard
deviation, what would you expect the number of samples in the sampling distribution to be?
What does this tell you about the relationship between a large standard deviation in the
underlying population and having a small standard error on the mean?
5. You sample a population which results in the following outcomes [1,0,0,0,0,0,1,0,0,0,1,0,0,0]
where 0 indicates a failure and 1 indicates a success.
a) What is the proportion of samples that result in a success?
b) What is the standard deviation of the sampling distribution (proportions equation!)?
c) What is the 95% confidence interval for the proportion of successes?
d) Name two conditions that must be satisfied in order to create a valid confidence interval for
the proportion of successes.
e) Based on the proportion of successes assessed in part A and the standard deviation
calculated in part B, is a sample with a successful proportion of the population of 0.2,
statistically significant? (i.e. p < 0.05)
6. A manufacturer of pipes is upset at the variability in how tall their pipes are. An average pipe
that they manufacture is 75 cm tall with a standard deviation of 5 cm. The manufacturer wants
to improve the fabrication process, towards that end they would like to employ a margin of
error of 3 cm with a confidence level of 95%. What should the sample size be? What should the
sample size be if the standard deviation were 10 cm? What effect does increasing the standard
deviation of the underlying distribution have on the number of samples?
7. A survey of 37 StFX students asks how often they eat a delicious savory waffle from the waffle
bus stop each week. This yielded a sample average of 1.7 waffles with a standard deviation of 1
waffle. A separate detailed analysis indicates that the population of non-student residents of
Antigonish, eat on average 0.9 waffles per week. We are interested in answering the question of
whether this data supports the notion that StFX students eat on average more waffles than a
non-student resident of Antigonish. This question will be answered with hypothesis testing.
a) What is the null hypothesis?
b) What is the alternative hypothesis?
c) What is the Z score for the mean of the sampling distribution?
d) Does this support the hypothesis that StFX students eat more waffles than the non-student
population of Antigonish?
8. As n, the number of samples increases, do the following variables increase, decrease or remain
constant?
a) Standard error
b) Z score of the distribution of the sampled mean
c) t statistic
d) p-value
9. A population has a true underlying mean value of 100 but an unknown standard deviation. A
sample of 10 is drawn from that population yielding a mean value of 98.2 and a standard
deviation of the sample of 10. Is this a statistically significant finding (p<0.05)?
10. Two separate populations are sampled, the sample from the first population had 100 samples
and yielded a sample mean of 50 and a standard deviation of 10. 82 samples were obtained
from the second population yielding a sample mean of 40 and a standard deviation of 7.
a) Calculate the 95% confidence interval for comparing these two samples
b) Is a sampled difference between the means of the two groups of 13 statistically significant?
c) Is a sampled difference between the means of the two groups of 8 statistically significant?
(have answer be diff from previous)
d) What is the pooled standard deviation of these samples?
e) What is the correct number of degrees of freedom for a t statistic that makes use of these
pooled standard deviations?
11. You are running an experiment to determine if males have worse senses of humour than
females. Out of a large population of students, you pair up students who have the same age and
hair colour but different genders (this helps make sure your experiment is not testing the effect
of age or hair colour on sense of humour). For each sampled pair of males and females, both are
told the same awful toilet humour joke and each are asked to rate how funny it was on a scale
of 0 to 10. The female’s score is subtracted from the correspondingly paired male’s score,
yielding a difference in how funny each pairing of students found the toilet joke (it wasn’t very
funny!). The average of the differences between the paired male and female subjects is 3 and
the standard deviation of the sample of differences is 1. There were 30 male and 30 female
students included in the sample.
a) Calculate the 95% confidence interval.
b) Do the data provide convincing evidence that males have worse senses of humour than
females?
Additional questions from textbook:
Chapter 7 (note the margin of error = half the ‘width’, i.e. we have used a slightly different sample size
equation in the slides than in the text): Questions: 4, 6, 14, 24, 28, 32

Chapter 8: Questions: 4, 8, 24, 26

Want more problems?????

Try the uneven numbered questions in the textbook – you won’t get solutions but you will get answers
in the back of the book.

Common questions

Powered by AI

As the sample size increases, the standard error decreases because it is inversely proportional to the square root of the sample size. This decrease in standard error leads to a larger Z score or t statistic for a given difference, improving the power of the test. Consequently, the p-value decreases, making it easier to reject the null hypothesis .

Two conditions necessary for a valid confidence interval for proportions are: 1) the sample must be randomly selected, ensuring independence of observations, and 2) the sample size must be large enough that the number of successes and failures both exceed 5. This can be checked with np ≥ 5 and n(1-p) ≥ 5, where n is the sample size and p is the proportion of successes .

The degrees of freedom for a t statistic using pooled standard deviations can be calculated as n1 + n2 - 2, where n1 and n2 are the sample sizes of the two populations. In this case, from the example with sample sizes of 100 and 82, the degrees of freedom is 180 .

Calculate the t-statistic using the formula: t = (sample mean - population mean)/(sample standard deviation/√n). Here, t = (1.7 - 0.9) / (1/√37), resulting in t ≈ 4.39 . This t-statistic is then compared to critical t-values to assess significance.

For a 99.9% confidence interval, the Z score corresponds to the value that leaves 0.05% in each tail of the standard normal distribution. The Z score is approximately 3.291 . This is found using statistical tables that provide Z scores for given confidence levels.

To calculate the 95% confidence interval, use the formula: CI = mean ± Z*(standard deviation/√n). Here, mean = 102.4, standard deviation = 8.6, sample size n = 42, and Z = 1.96 for 95% confidence. So, CI = 102.4 ± 1.96 * (8.6/√42) = (100.0, 104.8).

The paired t-test is appropriate here. Calculate the t-statistic: t = mean difference / (standard deviation/√n) = 3/(1/√30). Compare this t value to critical t-value at the desired confidence level (e.g., 95%). If the calculated t is greater than the critical value, there is significant evidence to claim males have worse senses of humour .

The two conditions that must be met are: 1) the samples must be drawn randomly from the population, and 2) the sampling distribution of the sample mean is approximately normally distributed, which can often be satisfied if the sample size is large enough (Central Limit Theorem).

The standard error of the mean (SEM) is calculated by dividing the sample standard deviation by the square root of the sample size. With SEM = 0.5 and population standard deviation = 50, the sample size calculation follows: SEM = standard deviation/√n => 0.5 = 50/√n => √n = 100 => n = 10,000 . This illustrates that a large sample size reduces the SEM.

An increase in the standard deviation from 5 cm to 10 cm would require a larger sample size to maintain the same margin of error. This is because sample size is directly proportional to the square of the standard deviation when keeping other factors constant . With a greater variability, more samples are needed to precisely estimate the population mean.

You might also like