On the first exam, your score reflects two things:

- Your ability in applying the material learned in this course
- Errors associated with how the exam was graded. These
errors come in two forms; one is much worse than the other:
1. Random error (can be corrected for - see below)

2. Systematic Error (extremely serious if you don't know it exists)These Errors works like the following:

What we measure is X but what we are interested in is the distribution of the true variable, T. To measure T, however, we have to know what the random error, e

_{r}and systematic error e_{s}is.Without knowledge of e

_{r}and e_{s}, T can never be accurately measured. This potentially is a huge problem.What is e

_{r}Random errors increase the dispersion. These errors are associated with apparatus or method used in obtaining the data. All data sampling is subject to random error, period. There is no way to avoid it. However, if you know what the value of e

_{r}is, you can "subtract" it out of the data.Here is how you due to subtraction. In a bell-shaped distribution, the various components are added or subtracted in quadrature. What does that mean?:

(measured dispersion) ^{2}= (true dispersion)^{2}+ e_{r}^{2}so

(true dispersion)

^{2}= (measured dispersion)^{2}- e_{r}^{2}Example: Suppose a measured distribution has a dispersion of 8 units and we know that the measuring error e

_{r}is 6 units. What is the true dispersion?(true dispersion) ^{2}= 8^{2}- 6^{2}

(true dispersion)^{2}= 64 -36

(true dispersion)^{2}= 28

(true dispersion) = 5.3 (square root of 28)What is e

_{s}?Systematic errors mean that different methods of measurement are being applied to the same variable. This means that the position of the mean is strongly effect. For example, suppose there are two patrolment on the freeway both with identical radar guns. Except that one of them systematically reads 5 mph higher than the other due to a "calibration" error back at the station. Which policeman do you want to speed by?

Your exam was graded by 3 different people although all have the
same exam key. The exams were * randomly * sorted as they
were handed in and the 3 graders randomly picked a pile. Each pile
has 35 exams.

Since the intrinsic distribution of exam scores will be normal (bell- shaped) then by the sampling principles established earlier we know that 30 samples is enough to accurately reflect the distribution.

Therefore, each of the 3 piles should have the same mean and dispersion. Differences in grading style, however, will cause each grader to have a slightly different mean and dispsersion.

e_{s} will then manifest itself
by different mean scors for different graders. This can be corrected
for easily.

e_{r} manifests itself by different
dispersions for different graders. This is due to random errors
associated with grading the questions and assigning a point value
in a lightly subjective manner.

For this exam, your dispersion score was calculated by the instructor
after e_{r} and e_{s} had been determined. In principle,
this needs to be done for every exam you take in large classes.
Most instructors don't do this, period (its too much work).

You should always demand that your instructors prove to you that in the case of multiple graders on the exam, no bias exists

For this exam:

- e
_{r}was much smaller than the intrinsic dispersion - e
_{s}was not significant (all graders, graded in the same manner to within the errors).

Introduction to Measurement Errors:

In class exercise.

On a piece of paper, write down your estimates for the following quantities and give them to one of the TAs.

- The age of the professor
- This guys salary (he's the UO prez)
- The number of oil wells in the United States

Understanding the role of measurement errors is crucial to proper data interpretation. For instance, the measured dispersion in some distribution represents the convolution of

- the intrinsic dispersion
- measurement error
- the precision of the measurements

In general, you only care about the intrinsic dispersion in some distribution. That is, you don't want to have the dispersion dominated by measurement error or poor precision because then you can't draw any valid conclusion.

Example: Column 1 contains the data that was measured with good precision. That is, the measuring error of the instrument was less than 0.1. Column 2 represents the same data that was measured with and instrument that had a measuring error of +/- 1 unit:

- 20.1 18
- 20.2 19
- 20.3 19
- 20.3 21
- 20.4 20
- 20.4 19
- 20.5 22
- 20.6 21
- 20.7 23
- 20.8 21
- 20.9 20

Clearly the first column is a better measure of the intrinsic distribution of the sample than the second column. Essentially the numbers in the second column are meaningless.

In principle, every measurement has an error associated with it and hence a measurement is only as good as its error. Knowing the size of measuring or sampling errors is often difficult.

For some kind of sampling, error estimation is straight forward. For instance, opinion poll sampling has an error that depends only on the Number of people in the sample. This error has to do with counting statistics and is expressed as

- Percentage yes = 75 +/- 25%
- Percentage no = 25 +/- 25%

- Percentage yes = 75 +/- 3%
- Percentage no = 25 +/- 3%

Conclusion: Always ask what the measuring errors are!!!

Data reliability involves the determination of the extent to which the data sample can address the question. For data samples widely distributed over time, one can raise the issue of the reliability of data taken with older instruments.

Data reliability is an issue when you can not measure a phenomena directly but instead you have to use some tracer.

For example:

What is the best way to measure the extent of urban air pollution?

- Measure daily pollutants combined with some circulation model?
- Set up many measuring stations around the city? costly
- Use Lichen ?

- Direct Measure of ph in Rain at various sites? this depends on many factors, some natural
- Use the New Brunswick Data ?
- Measure SO_2 output directly;
Global patterns ;
More recent annual data
- Examine effect on certain kinds of trees -e.g. sugar maples in NE USA
- Measure soil acidity?
- Measure the rate of degradation of historic monuments?

Issues to raise:

- Is the determination of the world's average annual temperature
as a function of time the correct diagnostic for showing global
warming?
- Is the Northern Hemisphere Data more reliable than the Southern
Hemisphere Data? Is it a better diagnostic.
- Are measurements made 100 years ago as reliable as measurements
made now?
- Are the thermometers used located in the correct areas?
- Are their enough individual measurements to reliably establish what the global mean temperature is?

Let's see the data:

Here is some Global Data whose reliability is high. There is no doubt that CO_2 in the atmosphere is increasing and is now at unprecedented levels

Methane concentration from ice core data!

Previous Lecture Next Lecture Course Page