5. SURVEY STANDARDS

5.1 STRATIFICATION
5.2 CLASSIFICATIONS
5.3 VALIDITY OF SURVEY STANDARDS OVER TIME

The definition and application of survey standards are key features of the methodological and operational framework of a sample-based fishery survey. Setting up survey standards includes the processes of stratification and classification.

Stratification of the domain to be covered by the statistical programme will include decisions on:

• Administrative strata
• Logical strata (estimation contexts)
• Sampling locations (homeports and landing sites)

Classification of the units that will be measured, including:

• Boat and gear categories
• Species and species groups
• System units (i.e. weight, currency and effort units)

Well-defined survey standards help in streamlining field operations, facilitating computerization, producing consistent reports and integrating survey outputs with the results of other application domains. Poorly defined standards have a negative effect on field operations, computer operations and on the meaning of produced estimates.

5.1 STRATIFICATION

5.1.1 Objectives of stratification

Stratification methods are applied in cases of:

• the need for more homogeneous target populations, which will provide lower variances in the estimates.
• categorization of the data population in order to respond to specific user needs.
• “forced” stratification dictated by administrative, reporting, functional and other non-methodological criteria.

5.1.2 Major Strata

The first step is to divide the entire statistical area into administrative or reporting strata, also called major strata. It should be noted that:

• Major strata do not constitute estimation contexts.
• Totals at major stratum level are computed by adding up estimates produced at lower (minor) level.
• Definition of major strata is usually dictated by external factors and not by real statistical needs.

5.1.3 Minor strata

Within each major stratum there exist “logical strata” that constitute the estimation contexts of the survey. These sub-divisions are called minor strata. It should be noted that:

• A minor stratum in a major stratum cannot be associated with another major stratum.
• Minor strata are not limited to geographical areas. They can refer to any logical estimation context including sub-periods within a month, to fishing grounds or fishing vessels. An important fishing location can itself constitute a minor stratum, if estimates are required at that level.
• Minor strata are controlled by the survey designers, and their purpose is to improve the quality and utility of estimates.
• Excessive division of potential data sources/types into minor strata can compromise the cost-effectiveness of a sampling programme.

5.1.4 Homeports and landing sites

Homeports are locations from which fishing boats operate, i.e. where they are based. Individual boats may use one or more landings sites, including the homeport. The homeport is always the basis for reporting the numbers of fishing boats and gears in frame surveys, and for the estimation of fishing effort.

The figure above illustrates a theoretical stratification approach. It should be noted that:

• Frame surveys are conducted at all homeports (shaded locations).
• Effort surveys are conducted at homeports selected for sampling.
• Locations can be both homeports and landing sites (half-shaded).
• Landing surveys are conducted at landing sites selected for sampling (white or half-shaded).
• Estimates are produced at minor stratum level.
• Totals for major strata are computed by combining the results produced at minor stratum level.

5.2 CLASSIFICATIONS

5.2.1 Boat/gear categories

In defining boat/gear categories the following points are usually considered:

• Level of detail required: This depends on the feasibility of data collection operations in frame surveys and effort surveys.
• Uniformity: Landing surveys, frame surveys and effort surveys must all use the same boat/gear classification.

The criteria for defining boat/gear types are usually based on known (or assumed) significant differences in:

• Species composition or size
• Catch rates
• Fishing trip patterns
• Fishing methods

5.2.2 Species classifications

The criteria for defining species classifications are usually based on the need to prioritize statistical monitoring of catches for:

• Commercially important species or species groups.
• Species that are important in certain areas to the local people.
• Species of particular biological interest.

5.2.3 Measurement units

Standard measurement units must be consistent throughout the statistical programme. Weight should normally be recorded in metric units, usually the kilogram.

In surveys for basic fishery data there is often a need to easily integrate catch and effort estimates deriving from different boats and gears. For small-scale fisheries the boat-day is a reasonably good way of expressing fishing effort uniformly.

5.3 VALIDITY OF SURVEY STANDARDS OVER TIME

By definition, survey standards are defined on an a priori basis. Their purpose is to provide a methodological and operational survey framework that will be valid for a certain length of time. Survey standards should be valid for a complete operational cycle, usually a year, after which period they may be reviewed.

Validity problems may occur after the first few months of survey implementation if changes are required to stratification schemes or to boat/gear or species classifications. However, as shown in table 5.3.3 below, modifications and changes to survey standards in the middle of an operational cycle are permissible when they do not affect the consistency of the survey framework.

Table 5.3.3 Implications of changes in survey standards to ongoing operations

Stratification
Type of change	Permissible	Consistency implications
Addition of new major or minor strata	Yes	None
Addition of new homeports or landing sites	Yes	Frame survey adjusted
Changes in associations between minor strata and major strata	Yes	Reports at major stratum level to be re-produced
Changes in associations between sites and minor strata	No	Else primary data must be re-organized and estimates re-produced
Classifications
Type of change	Permissible	Consistency implications
New species or name changes	Yes	None
New boat/gear types or name changes	Yes	Frame survey adjusted
More detailed or more grouped species level	No	Else data inputting and estimates re-done
More detailed or more grouped boat/gear level	No	Else frame survey re-structured Data inputting and estimates re-done

SUMMARY In this section the concept of survey standards was discussed. (a) Survey standards are the methodological and operational framework of a sample-based fishery survey through: Stratification of the domain to be covered by the statistical programme into major strata, minor strata and sites. Classifications of boat/gear categories and species. Standard System units (i.e. weight and effort units). (b) Well-defined survey standards help in streamlining field operations, facilitating computerization, producing consistent reports and integrating the survey outputs with those from other application domains. (c) Poorly defined standards have a negative effect on field operations and on the meaning of produced estimates. They also create problems in the computer-related tasks relating to data organization, processing and analysis. (d) Survey standards are defined on an a priori basis and are usually assumed to be valid for one year, after which period they are reviewed. (e) Validity and consistency problems were summarized in Table 5.3.3.