Errors and Statistics

Measurement

• Process of determining the extent, size, or

dimensions of a particular quantity in
comparison to a given standard.
• Consists of several physical operations
which renders a numerical value.
• Maybe direct or indirect.
 Measurement vs Observation

• A measurement entails the entire process of

obtaining a desired quantity, including
preparations (instrument calibration and
setup), pointing, matching, and comparing.
• An observation is a single, unadjusted
determination of a linear or angular value.
 Variability in Repeated
Measurements

• Measurements are numerical values for

random variables which are subject to
statistical fluctuations.
• An inherent quality of physical properties.
• Statistical variations due to observational
errors.
“ No exact or true measurement is ever
possible and the value of a measured
quantity is never known.”
 Error and Correction
Error - difference between the measured or
calculated value of a quantity and given or
established (“true”) value of that quantity.

   
Correction - the negative of error.

correction    
 Sources of Errors
1. Natural Errors - caused by variations in
the phenomena of nature such as
changes in magnetic declination,
temperature, etc.
2. Instrumental Errors - due to
imperfections in the instruments used.
3. Personal Errors - arise principally from
the limitations of the senses of sight,
touch, and hearing of the observer.
 Types of Errors

1. Mistake or Blunders
2. Systematic or Cumulative Errors
3. Random or Accidental Errors
 Mistakes or Blunders

• Actually not errors because they are

usually so gross in magnitude compared
to the other types of errors.
• One of the most common reasons is
simple carelessness on the part of the
observer.
• An observation with a mistake is not
useful unless the mistake is removed.
Common Mistakes or Blunders
1. Reading the wrong graduation on the
tape.
2. Omitting a whole length of tape.
3. Transposition of figures.
4. Misplacing a decimal point.
5. Incorrect recording of field notes.
6. Sighting the wrong target.
 Systematic or Cumulative Errors

• So called because they occur according to

some deterministic system which, when
known, can be expressed by some
functional relationship.
• Caused by physical and natural conditions
that vary in accordance with known
mathematical or physical laws.
 Types Systematic Error

Constant Error - if its magnitude and sign

remains the same throughout the
measuring process or field conditions are
unchanged.

Counteracting - if its sign changes while its

magnitude remains the same.
– due to personal bias of the observer.
 Common Systematic or
Cumulative Errors

1. Equipment out of calibration.

2. Personal biases of the observer.
3. Use of incorrect units (feet instead of
meters).
 Random or Accidental Errors

• Produced by irregular causes that are

beyond the control of the observer.
• This variation results from observational
errors which have no known functional
relationship based upon a deterministic
system.
• Must use probability models.
 Statistics
General Uses of Statistics:
• Statistics aids in decision making.
– Provides comparison.
– Explains action that has taken place.
– Justifies a claim or assertion.
– Predicts future outcome.
– Estimates unknown quantities.
• Statistics summarizes data for public use.
 Precision vs Accuracy
Precision - degree of refinement and measure
of uniformity of the result.

Accuracy - degree of conformity with a

standard or accepted value.
- denotes how close a given measurement is
to the absolute value of the quantity.
 GOOD ACCURACY POOR ACCURACY
& GOOD PRECISION & GOOD PRECISION

GOOD ACCURACY BAD ACCURACY

& BAD PRECISION & BAD PRECISION
 The Concept of Probability
Probability – is the likelihood associated
with a random event.

Random Variable – a variable that takes on

any of several possible values.

Random Event – is one whose relative

frequency of occurrence approaches a
stable limit as the number of observations
is increased to infinity.
 Representations of the
Probability Density
Frequency Diagrams:

Histogram – represents the probability

density of a single random variable.

Stereogram - represents the probability

density of two random variables.
Measures of Central Tendency

1. Median
2. Mean
3. Mode
4. Midrange
Sample Median - positional middle of the
arrayed data.

• Characteristics:
- Affected by the position of each item
but not by the value of each item.
- A stable measure of central tendency.
Sample Mean - sum of all the values of the
observations divided by the number of
observations. (Most Probable Value)

• Characteristics:
- Most familiar measure of central tendency
used.
- Affected by the value of every observation.
- In particular, it is strongly influenced by
extreme values.
- Since it is a calculated number, it may not be an
actual number in the data set.
Sample Mode - value that occurs most
frequently in the sample.

• Characteristics:
- Not always exist. If it does, it may not be
unique (2 or more sample modes).
- Not affected by extreme values.
- Easiest to compute.
Midrange - value of observation that is
midway along the range.
- arithmetic mean of the largest and
smallest observations.
 Sample Statistics for
Dispersion
1. Range
• The total spread of the sample.
Range = Largest value - Smallest value

2. Mean Deviation
• Arithmetic mean of the absolute values of
the deviation from any measure of
position.
1 n
MD   X i  X
n i 1
3. Variance n
• Parameter of
dispersion or spread.
(X i  X) 2

S 
2 i 1

n 1
n
4. Standard Deviation
• Positive square root (X i  X) 2

of the variance. S i 1

n 1
 Measures of Quality

Weight - the quantity that is inversely

proportional to variance.

Relative Error or Precision – ratio of the error

to the measured or estimated quantity.

Ratio of Misclosure - ratio between the total

error and the total length of the survey.
Probable Errors
 Residual

• Sometimes called the deviation.

• Defined as the difference between any
measured quantity and its most probable
value (MPV).

v  residual

vX X X  an individual measurement

X  MPV of the quantity
 Probable Error

• A quantity which, when added to and

subtracted from the MPV, defines a range
within which there is 50% chance that the
true value lies inside (or outside) the
limits thus set.

X  PE
 Probable Error

PE s  0.6745
v 2 Probable error of
any single
n1 measurement of
a series.

PE m  0.6745
v 2

Probable error of
n(n  1) the mean.
 Relative Error or Precision
• Expressed by a fraction having the
magnitude of the error in the numerator
and the magnitude of a measured quantity
in the denominator.

EXAMPLE:
If PEm= 0.10 m and MPV = 235.50 m,
Relative Precision = 0.10/235.50 = 1/2355 or 1:2355.
EXAMPLE:
The following values were determined in a series of tape
measurements of a line in meters:
1000.58, 1000.40, 1000.38
1000.48, 1000.40, 1000.46
Determine:
a) MPV of the measured length.
b) PEs and PEm.
c) Final expression for most probable length.
d) Relative Precision of the measurement.
SOLUTION:
a) MPV:

n = 6 (number of observations)

X  X 1  X2  X3  X4  X5  X6
 1000.58  1000.40  1000.38  1000.48  1000.40  1000.46
 6002.70 m
MPV or X   X / n
 6002.70 / 6
 1000.45m
b) PEs and PEm:
Measured Length Residual Square of
(X) m (v) Residual (v2)
1000.58 -0.13 0.0169
1000.40 +0.05 0.0025
1000.38 +0.07 0.0049
1000.48 -0.03 0.0009
1000.40 +0.05 0.0025
1000.46 -0.01 0.0001
X=6002.70 v=0.00 v2=0.0278
 PE s  0.6745
v 2

 0.6745
0.0278
n 1 6 1
 0.05m

PE m  0.6745
v 2

 0.6745
0.0278
n(n  1) 6(6  1)
 0.02m
c) Final Expression for the MPV of the length:
1000.45 ± 0.02 m

d) Relative Precisions:
RPs = PEs/MPV = 0.05/1000.45
RPs = 1/20000
RPm = PEm/MPV = 0.02/1000.45
RPm = 1/50000
 Weighted Observations
• Degree of reliability.
• Usually based upon:
1. Judgment of the surveyor.
2. Number of measurements taken for a
particular quantity.
3. The assumption that it is inversely
proportional to the square of the probable
error.
EXAMPLE:

X (m) W P=X*W
Measured length Assigned weight Weighted length
110.26 6 661.56
110.13 2 220.26
110.20 3 330.60
110.31 1 110.31

Sums: W = 12 P= 1322.73

Weighted Mean = ∑P/∑W = 110.23 m.

 Interrelationship of Errors

Summation of Errors:

PESUM   PE12  PE22  PE32    PEn2

Where:
PESUM = Probable error of the sum
PE1, PE2, etc. = PE of each measurement
EXAMPLE:
A triangular-shaped tract of land has sides a =
35.02 m, b = 22.15 m and c = 12.65 m, each having
probable errors of 0.05, 0.07 and 0.03 m,
respectively. Compute for the probable error of
the perimeter.
PE SUM   PE12  PE 22  PE32    PE n2
  (0.05) 2  (0.07) 2  (0.03) 2
 0.09m
Note: The MPV of the perimeter is 69.82 ± 0.09 m.
 Interrelationship of Errors

Product of Errors:

PE P   (Q1  PE 2 ) 2  (Q2  PE1 ) 2

Where:
PEp= Probable error of the product
Q1 & Q2 = Measured quantities
PE1, PE2 = PE corresponding to each quantity
EXAMPLE:

A rectangular lot has sides a = 51.17 m and b =

27.30 m each having probable errors 0.02, 0.05
meters respectively. Compute for the probable
error of the calculated area.
PE P   (Q1  PE 2 )  (Q2  PE1 )
2 2

  (51.17  0.05) 2  (27.30  0.02) 2

 2.62m 2
Note: The MPV of the area is 1396.94 ± 2.62 m2.
REFERENCE

La Putt, Juny Pilapil. “Elementary

Surveying”, 3rd Ed.National Bookstore: 2007.
pp. 44-50