PW 01 - 25.02-1.03.2019 - Physical Quantities and Error Calculations PDF
PW 01 - 25.02-1.03.2019 - Physical Quantities and Error Calculations PDF
PW 01 - 25.02-1.03.2019 - Physical Quantities and Error Calculations PDF
Practical works
01 Introduction
25.02 – 1.03.2019
Physical quantities
1) Measuring physical quantities
QUANTITIES AND UNITS IN PHYSICS
equation of dimensions.
[ ]
measurement units
Median (Me) is the value of the set of measured values for which half
of the values are strictly lower and half of the values are higher than
the median.
301 K < 305 K < 307 K < 308 K < 310 K
Average or mean value
or taking into account the next notation for the sum of n values:
∑
̅ (∑ )
If the set of measurements contains several equal values (we call the
absolute frequency and we note by fi the occurrences for a
repetitive value), then the mean value is calculated as the weighted
average of the sample using the following formula:
with or
(∑ )
Mode
The representative parameter indicating the value occurring at the
highest frequency and describing the central trend, i.e. the the
distribution symmetry of the measurement values around a central
value, is the selection mode defined as:
̅
̅ the values set has a perfect symmetry.
̅ the values set has a negative symmetry
̅ the values set has a positive symmetry
13) Statistical errors. Numerical
application for the temperature
measurement set of values (K): 305
K, 307 K, 301 K, 310 K şi 308 K
***The numerical application is done in the classroom.
TYPES OF STATISTICAL ERRORS
Partial error (for each measurement )
̅
Absolute error
(| | | | | |) ∑| |
The downside to the absolute error is that it does not provide much information
about the accuracy of the measurements.
Relative error :
̅
This type of error has the advantage of evaluating the precision of
the measurement chain.
The relative error can also be presented as a percentage error:
( ) ∑( )
√ √ ∑( )
Coefficient of variation
̅
evaluates the homogeneity of the studied population.
o if cv < 10 % the population is homogeneous and the sample is
representative;
o if cv > 50 % the population is heterogeneous and the sample is
unrepresentative.
Mean standard deviation
√
It can be proved that with the increase in the number of
measurements, the value of the mean standard deviation is constant,
in other words the absolute error of measurement is constant.
the mean standard deviation is not calculated for less than five
measurements (n < 5)
over 20 measurements (n> 20) its value is independent of the
number of measurements, in other words the mean standard
deviation is independent of the precision of the measuring
device.
The confidence interval (CI)
Dispersion around the central value can also be represented in the
form of the confidence interval [ ], cu:
̅
̅
where t is the value of a statistical parameter related to the statistical
test (the Student test, test, the Fisher test etc.) applied to the set
of measurements.
Numerical applications
No. Value Partial error Parameter Calculated
mes. value
1. 301 amplitude, A
2. 305 dispersion,
3. 307 Standard deviation, s
4. 308 The coefficient of
variation %
5. 310 Mean standard
deviation
Confidence interval
(t = 0,05)
Remarks