Scribd
Upload
147 views

Stats Part2 Week9 020507

sampling assemblages in SPSS

Uploaded by

Dr Richard Haddlesey
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
147 views

Stats Part2 Week9 020507

sampling assemblages in SPSS

Uploaded by

Dr Richard Haddlesey
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPT, PDF, TXT or read online on Scribd
You are on page 1/ 33

Computing in

Archaeology
Session 9. Sampling
Assemblages
© Richard Haddlesey www.medievalarchitecture.net
Aims
 To become familiar with sampling
practices in an archaeological context
Introduction to Sampling
 An area of excavation is a sample of
the complete site which in itself is a
sample of all sites of that type. The
same goes for artefact assemblages.

 The essence of all sampling is to gain


the maximum amount of information
by measuring or testing just a part of
the available material

Fletcher & Lock 2005, 66


Archaeological sample

Sampled population

Target population
Formal definitions
 Population: the whole group or set of
objects about which inference is to be
made

 Sampling fame: a list of the items, units


or objects that could be sampled

 Variable: a characteristic which is to be


measured for the units, such as weight of
spearheads

Fletcher & Lock 2005, 66


Formal definitions
 Sample: the subset or part of the
population that is selected

 Sample size: the number in the sample.


A sample size of 5 is considered small,
while, formally, a sample size of 50 is
large. The sample size maybe stated as a
percentage of the sampling frame, e.g. a
10% sample

Fletcher & Lock 2005, 67


Sampling strategies

• a simple random sample (probability sample USA)

• a systematic sample

• a stratified sample

• a cluster sample
population – 100 units

. . . etc

100 obsidian spearheads


population – 100 units
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
A simple random number sample
Random sampling
 If we have a sample of 100 spearheads,
we simply pick 10 random numbers (i.e.
10%)

 Computers can help generate random


sequences, but are not necessary

 You must avoid bias in your selection as


this can result in scrutiny from others
a simple random number sample
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
A systematic sample
Systematic sampling
 To take a systematic approach, we
could choose every number ending in
4. Once again this would give us our
10%

 This method has the advantage of


being easy to design unless the units
have inherent patterning in their
order
a systematic sample
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
A stratified sample
Stratified sampling
 Here we take a random sample 5
from the top and five from the
bottom

 Or 5 from the left, 5 right etc


a stratified sample
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
a stratified sample
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
A cluster sample
Cluster sampling
 Rather than select individual items,
select clusters or groups of items
that are close together

 This may result in bias values


a cluster sample
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
a cluster sample
1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90

91 92 93 94 95 96 97 98 99 100
Downside to systematic

Totally miss
this context
Common sample statistics:
 x – the sample mean

 s – the sample standard deviation

 p – the sample proportion (i.e. the


proportion of the sample having a
particular characteristic)
Stats
 The true population values for these
statistics are usually unknown, and
formally denoted by Greek letters
Common sample statistics:

known value estimate for

x – the sample mean μ – the population mean

s – the sample standard


deviation

p – the sample
proportion
Common sample statistics:

known value estimate for

x – the sample mean μ – the population mean

s – the sample standard σ – the population


deviation standard deviation

p – the sample
proportion
Common sample statistics:

known value estimate for

x – the sample mean μ – the population mean

s – the sample standard σ – the population


deviation standard deviation

p – the sample π – the population


proportion proportion
The central-limit theorem
(the law of averages)

 In order to comment on how good an


estimate the sample statistics are,
the nature of their distribution needs
to be known

 See
• Fletcher & Lock (2nd ED) 2005, Digging
Numbers Oxbow 70-9

You might also like