Scribd
Upload
8 views

4. Estimation

The document discusses the concept of estimating with uncertainty, particularly focusing on the standard error of the mean and its implications for confidence intervals. It highlights the importance of sample size in reducing variation in sample means and the common issue of pseudoreplication in data collection. Additionally, it emphasizes the correct interpretation of confidence intervals in statistical analysis.

Uploaded by

hikkihikaruww
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
8 views

4. Estimation

The document discusses the concept of estimating with uncertainty, particularly focusing on the standard error of the mean and its implications for confidence intervals. It highlights the importance of sample size in reducing variation in sample means and the common issue of pseudoreplication in data collection. Additionally, it emphasizes the correct interpretation of confidence intervals in statistical analysis.

Uploaded by

hikkihikaruww
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Estimating with uncertainty

https://www.zoology.ubc.ca/~whitlock/
Kingfisher/SamplingNormal.htm
Chapter 4

Sample size 10 from Normal distribution with µ=13 and s2=16 Another sample of 10 from same distribution

Frequency
Frequency
Distribution of the means of many samples,
A third sample of 10 from the same distribution
each of sample size 10
Frequency

A sample of 100 from the same population distribution A sample of 1000 from the same population distribution
Frequency

Frequency
Variation in sample means decreases with sample size
Distribution of the means of many samples,
each of sample size 100
n = 10

n = 100

1000 samples each

The standard error of an


estimate is the standard
deviation of its sampling
distribution.

The standard error predicts Mean = 67.4


the sampling error of the
estimate.
Standard error of the N

mean

Mean = 67.4
The math works!

The problem is,


we rarely know s.

Estimate of the standard N In most cases, we


don’t know the real

error of the mean population


distribution.

We only have a
sample.

This gives us some knowledge of the likely


difference between our sample mean and the
We use this as an
true population mean. estimate of
Confidence interval The 2SE rule-of-thumb
The 95% confidence interval provides a
The interval from - 2 to + 2
plausible range for a parameter. All values
provides a rough estimate of the 95%
for the parameter lying within the interval
confidence interval for the mean.
are plausible, given the data, whereas
those outside are unlikely.
(Assuming normally distributed population and/or sufficiently
large sample size.)

Use correct language when talking


about confidence intervals

Not correct:
https://www.zoology.ubc.ca/~whitlock/ “There is a 95% probability that the population
Kingfisher/CIMean.htm mean is within a particular 95% confidence interval”

Correct:
“We are 95% confident that the population mean
lies within the 95% confidence interval.”
Sample means of gene sizes
Confidence interval

US counties with high kidney cancer US counties with low kidney cancer
death death
Variation in cancer rates decreases
with population size of counties Pseudoreplication

The error that occurs when samples are not


independent, but they are treated as though they are.

Wainer (2007) The most dangerous equation. American Scientist 95: 249-256.

Example: Pseudoreplication Avoiding pseudoreplication

You are interested in average pulse rate of mountain You are interested in average pulse rate of mountain
climbers. Since they are hard to find, you decide to take climbers. Since they are hard to find, you decide to take
10 measurements from each climber. You study 6 10 measurements from each climber. You study 6
climbers, so you have 60 measurements. climbers, so you have 60 measurements.

What is your sample size (n)? Take the mean blood pressure for each climber, so that
you have 6 pulse rates, one for each climber (n = 6).

You might also like