Loading...
 
Read Edit History
Tags: bias

Bias Factors

Bias Factors

Definition



Full text

As stated in the introduction to this methodological wiki, archaeological field-walking survey is inherently biased by manifold natural, anthropogenic and methodological factors. The following table (from Given 2003, p. 19, table 2.2) sums up aspects which might have affected the archaeological data as we perceive it, from the artefact production in antiquity up to the presentation of archaeological research today. In the following text some of the methodological and post-depositional bias factors will be explained in more detail.

"Factors that might create or affect surface artefact density figures:

Cultural

1. Rate of artefact production

2. Rate of artefact disposal

3. Duration of artefact production

4. Duration of artefact deposition (e.g. continuous manuring for 300 years)

5. Manner of deposition: abandoned settlement, burial, dumping, manuring

Post-depositional

6. Build-up of incoming sediment over artefacts, covering and hiding them

7. Build-up of incoming sediment containing new artefacts from elsewhere

8. Erosion of existing sediment, removing artefacts with it

9. Stripping of existing sediment, exposing buried artefacts

10. Deflation of sediment leaving artefacts in place, causing greater surface densities

11. Movement because of recent anthropogenic disturbance (ploughing, dumping, bulldozing, etc.)

12. Differential sherd survival

13. Sebakh: the spreading of soil and sherds from ancient sites as agricultural fertiliser

Methodological

14. Ground visibility, and varying ways of estimating and compensating for it

15. Background confusion, and varying ways of estimating and compensating for it

16. Effects of sunlight, weather and soil moisture

17. Intensity of recording (how much of their strip do the fieldwalkers actually examine?)

18. Varying abilities of fieldwalkers to identify artefacts; fieldwalkers’ state of mind and health

19. Definition of artefact (e.g. minimum sherd size)

20. Collection policy: relationship between sherds counted and sherds collected

21. Identification and grouping of fixed chronological periods

22. Data entry, processing and analytical methods (typos, formulae for calculating densities, etc.)

23. Presentation: choice of symbols, data ranges, and scales"

(Given 2003, p. 19, table 2.2)

An archaeological survey examines spatial phenomena (sites or survey units) at specific moments in time (visits). Because of the dynamics of the archaeological landscape through time it is important to record the visibility conditions of each visit. Visibility conditions include all attributes within the spatial-temporal unity of a visit that can influence (either facilitating or obstructing) the recognition of archaeological remains. The simplest example is ground visibility; the amount of clear surface visible to the team of field-walkers. Vegetation cover and stoniness or rockiness can strongly decrease this, as well as the presence of modern structures (i.e. sheds, (dirt) roads, etc.). Of course, agricultural activities (most importantly ploughing) can heavily affect visibility conditions. It is therefore advisable to plan visits of the same survey unit in various (agricultural) seasons.

Furthermore, the weather can manipulate visibility conditions in various ways. Sunlight can increase contrasting colours, which makes it easier to detect potsherds, but it can also cast an irregular pattern of shadows on the field, making it almost impossible to discern anything at all. Shadiness rather than sunlight is therefore often recorded as an important visibility attribute. Rain, on the other hand (frustrating as it may be for a survey team), washes the dirt and dust from the exposed materials on the surface, leaving the clean sherds brightly coloured and easy to spot. Hence, soil humidity is another crucial factor. Most systematic intensive survey projects meticulously record visibility conditions, either all individual attributes or a more judgmental mean. However, it is important to note that none of these factors is practically measurable on a ratio scale. Recorded conditions are always subjective, arbitrary decisions made by a survey team (or its administrator).

This brings us to another crucial bias factor in survey archaeology: the field-walker. Whether they consist of students or professional archaeologists, survey teams (and their results!) are bounded by individual abilities and a wide range of climatic, social and health conditions. Experienced field-walkers might be more able to find hardly recognizable pottery classes, as it is far easier to recognize objects once you have seen them before. In the end, every field-walker has his or her own set of capacities and imperfections and consequently, all survey data will be subject to biases. Trying to control these biases, some survey projects have experimented with the monitoring and comparing of the number of finds collected by individual walkers.

Post-depositional biases cover all natural and anthropogenic processes that have occurred after the deposition of archaeological materials and which might have transformed them into their present state. Our understanding of post-depositional processes is vital to assessing the possible biases they might have caused in the archaeological record. Some of the most evident examples of post-depositional processes encountered in survey archaeology are erosion and agricultural activities, such as ploughing.

Bibliography and further reading

Given, M., 2003b, ‘Mapping and Manuring: Can we compare sherd density figures?’, in: Alcock, S. & J. Cherry, Side by Side Survey: Comparative Regional Studies in the Mediterranean World, Oxford, pp. 13-21.

Taylor, J., 2000, ‘Cultural depositional processes and post-depositional problems’, in: Francovich, R., Patterson, H. & G. Barker (eds.), The Archaeology of Mediterranean Landscapes 5: Extracting Meaning from Ploughsoil Assemblages, Oxford, pp. 16-26.

Terrenato, N., 2000, ‘The visibility of sites and the interpretation of survey results: towards an analysis of incomplete distributions’, in: Francovich, R., Patterson, H. & G. Barker (eds.), The Archaeology of Mediterranean Landscapes 5: Extracting Meaning from Ploughsoil Assemblages, Oxford, pp. 60-71.

Yorston, R., Gaffney, V. & P. Reynolds, 1990, ‘Simulation of artifact movement due to cultivation’, in: JASc 17, pp. 67-83.

Research topics: Survey Methodology