Statistics criteria, classification, means, dispersion, concentration, frequency curves, data collection, sampling errors

1
Statistics criteria, classification, means,
dispersion, concentration, frequency curves, data
collection, sampling errors

• Krisztian Orban, Ministry of Justice, Hungary
• Regional Strategic Workshop on Impact Assessment
• Riga, Latvia 2-4 February, 2005
• Organized by Open Society Institute and UNDP

2
population, criteria, classification

• population entities analyzed together due to a
  specific reason
• criteria analysis aspects, by which the
  population can be separated into groups not
  folding each other
• two criteria of classification
• to be definite
• neither gaps nor foldings

3
examples of criteria
4
means

• arithmetic mean
• sum of units / number of units
• weighted arithmetic mean
• sum of units multiplied with their weights / sum
  of weights

5
indicators of dispersion (example)

• row of numbers 1 7,8 7,9 8,1 8,2
• row of numbers 2 -294 1 33 292
• The mean is 8 in both cases.

6
dispersion indicators

• deviation from the mean
• average deviation
• (in weighted case)

7
dispersion indicators II.

• dispersion quadratic mean of the deviations from
  the mean (di)
• row 1 s1 0,1732,
• row 2 s2 201,9442

8
dispersion (example)
Average monthly income of the gipsy and non-gipsy housings, 1996 Average monthly income of the gipsy and non-gipsy housings, 1996 Average monthly income of the gipsy and non-gipsy housings, 1996
  Average monthly income of the housing dispersion
Gipsy 33.288 HUF 19.258 HUF
Non-gipsy 58.340 HUF 45.719 HUF
source Sági Matild Társadalmi folyamatok a
rendszerváltás után, OKI, 1997
9
concentration

• A significant part or the whole of the total sum
  of values belonging to a population is
  concentrated to a few units
• Lorenz-curve
• Herfindahl-index (HI)

10
Lorenz-curve (example)
50,7 38,5 10,8
11
frequency curves

• a curve relating the volume (magnitude) of a
  variable characteristic to its frequency of
  occurrence

12
frequency curve (example)
Distribution of the employed population by
age (in percent of the total population),
Hungary, 1994
13
characterization of the curves

• the curves can be characterized with the
  following parameters
• situation
• dispersion
• shape

14
situation
situation how the curve situated on the X-axis
15
dispersion

• dispersion influences the peak- and flatness of
  the curve

16
shape

• shape influences the symmetry and the monotonity
  of the curve

17
data collection

• sampling
• using administrative sources and
• experiments

18
sampling

• the adequate data can be collected
• no impact on the analyzed area
• full (all items)
• partial (a part of the whole)
• representative (items chosen properly)

19
administrative resources

• advantage cost- and time-effective
• disadv. more difficult to use the results
• e.g.
• economic and social research, analysis
  institutes,
• research groups

20
experiments

• the expert plays an active role in the analysis,
  he/she can modify or filter some factors
  determining the parameters of the mass.

21
sampling error

• it depends basically on the volume of the sample