Marine Directorate statistics: Code of Practice compliance

Each Marine Directorate statistical product has different quality considerations, therefore the quality management and quality assurance approach is described separately for each output:

• Scottish Sea Fisheries Statistics (provisional and final)
• Marine Economic Statistics
• Scottish salmon and sea trout fishery statistics (provisional and final)
• Aquaculture production surveys (fish and shellfish)

However, they all follow the Scottish Government Quality Management approach and quality assurance process, aligned with version 3.0 of the Code of Practice for Statistics. Information is provided on:

• Proportionality
• The European Statistics Code of Practice five dimensions of quality (Relevance, Accuracy, Timeliness, Accessibility, Comparability)
• Communicating uncertainty
• Quality assurance process

Scottish Sea Fisheries Statistics (provisional and final)

Proportionality

The Scottish Sea Fisheries Statistics are fit for their intended use, which is to provide an overall picture of the fishing industry in Scotland. This includes landings by Scottish vessels and landings into Scotland as well as information on number of fishers and number of fishing vessels. The information comes from management information apart from the number of fishers which is based on an annual survey completed by fishery offices.

Relevance

The key stakeholders for our data are summarised in the Public Involvement and Engagement section. In addition to the publication and supplementary tables, spatial information is also published based on this data. This data is used to inform policy development and implementation. In particular, spatial data is used for offshore wind impact assessments. The spatial data is currently only published at ICES rectangle level and we are aware that more detailed granularity is desired.

Accuracy

Although most of our data comes from management information, there is often a delay in receiving landings and sales information for smaller vessels. The provisional statistics published in May are less accurate and comprehensive, but timelier and enable users access to the fisheries data at an earlier point. Further quality assurance is carried out between the provisional and final data to resolve outstanding issues where possible, correct implausible data and impute missing values. Information on the size of the manual amendments, imputation of values and changes between the provisional and final publications can be found in tables 52 – 58 of the supplementary tables to the final Scottish sea fisheries statistics publication.

Timeliness

Provisional statistics are published in May, less than 5 months after the end of the year. This time allows for some quality assurance to be done and allows time for most of the paper fishing forms to be returned and entered into the database. Further quality assurance, liaison with the rest of UK for consistency and imputation of missing values using average prices per species is carried out for the final publication in September, 9 months after the end of the year.

Accessibility

The publication follows the Scottish government statistical template, accessible publication guidance and the analytical guidance on accessible tables. The text was written in plain English and charts with alt text were supplied. Key points were supplied at the start.

Comparability

In general, the data is comparable over time and with other fisheries data across the UK and Europe. The Scottish fishing data feeds into the UK fisheries data base and we also have access to data from the rest of the UK. When preparing the final publication, we work closely with the Marine Management Organisation who produces the UK fisheries statistics publication. This includes sharing manual amendments and using the same average prices to impute missing values. Where data is incomparable over time or across administrations, this is noted in the publication.

Communicating uncertainty

The key strengths of the fisheries data is its timeliness and coverage. The information is taken from a live management information database and it is a legal requirement for fishers to provide this information. This does not mean that the information is 100% complete or accurate. However, there is a high level of completeness and accuracy, particularly for larger vessels where the information is supplied electronically. In our latest publication we provided information on the changes to data between publications, the level of manual additions and manual imputation of values. The data we publish is suitable for providing an overall picture of the fishing industry in Scotland and the different sectors within it. It is not suitable for providing detailed fishing information on individual vessels as some information may be missing or imputed.

Quality assurance process

The quality of the source fisheries data is improved by running code to check for errors and inconsistencies. These are raised with fishery officers for further investigation and correction at source where required and possible. This process is clearly documented and conducted several months prior to producing the publication. Additional checks are added to this code when any new issues are discovered. The data is also checked by operational staff in compliance and expert users.

Detailed documentation, commented code and standard input files are used to produce the publication data and tables consistently. The code includes checks for anomalies, corrections to outliers, manual amendments using input files and imputation of missing values. Statisticians review outputs for consistency and check any anomalies with topic experts. More information on our data quality is available in Scottish sea fisheries statistics: technical manual - gov.scot

Where any known issues remain in this publication they are communicated briefly to users in the publication and associated tables. More detail on the known issues is available in Scottish sea fisheries statistics: technical manual - gov.scot and is signposted from the publication.

Marine Economic Statistics

Proportionality

The Marine Economic Statistics are fit for their intended use, which is to provide information on selected sectors of the marine and coastal economy in Scotland. This publication is still under development as it is not possible to include some marine and coastal sectors due to lack of data. Most of the key sectors are included, however information is generally only available at a Scotland level.

Relevance

The key stakeholders for our data are summarised in the Public Involvement and Engagement section. It is used as context in many policies to specify the size and relative importance of specific marine sectors to Scotland and the changes over time.

Accuracy

Most of the data in this publication comes from the Scottish Annual Business Statistics 2023 - gov.scot. This is taken from a sample survey of businesses, so information for sectors with a small number of businesses, such as shipbuilding are subject to high levels of annual fluctuations, so longer time trends are more reliable. The publication uses the confidence interval data for the offshore wind data to provide information on which changes are statistically significant.

Timeliness

This publication is based on a financial survey, so there is a long time lag, almost 2 years. We publish within about two months of the Scottish Annual Business Statistics publication. This is the time it takes to process the bespoke extract we receive from the Scottish Annual Business Statistics and add in the additional information we collect separately on offshore wind, aquaculture and fishing.

Accessibility

The publication follows the Scottish government statistical template, accessible publication guidance and the analytical guidance on accessible tables. The text was written in plain English and charts with alt text were supplied. Key points were supplied at the start.

Comparability

There are no directly comparable publications currently. However, as this publication pulls data from a variety of existing published sources, each individual component is comparable to the sources. Data is comparable over time and the time series is updated annually as the economic values are shown in real terms.

Communicating uncertainty

The publication clearly communicates the marine sectors that are included and those that are not due to a lack of available data. The main source publication Scottish Annual Business Statistics 2023 - gov.scot contains information on standard errors. Information on confidence intervals are available for the offshore wind data.

Quality assurance process

Detailed documentation, commented code and standard input files are used to produce the publication data and tables consistently. The code also checks for anomalies in the aquaculture economic data, which is collected directly and corrects outliers. Statisticians review outputs for consistency and check any anomalies with topic experts. Where any known issues remain in this publication, they are communicated to users in the publication and associated tables.

Scottish salmon and sea trout fishery statistics (provisional and final)

Proportionality

The Scottish salmon and sea trout fishery statistics are fit for their intended use, which is to provide an overall picture of the salmon and sea trout catch in Scotland. The time-series began in 1952 and includes the number and weight of salmon and sea trout caught, and netting effort. Rod effort for salmon has been included (since 2023). The information comes from an annual postal survey sent to all known salmon and sea trout fisheries in Scotland.

The provisional publication includes the most recent two fishing seasons, as required for use in international stock assessments by the International Council for the Exploration of the Sea (ICES) Working Group on North Atlantic Salmon (WGNAS). The final publication includes the all-Scotland time-series (from 1952).

Relevance

The key stakeholders for our data are summarised in the Public Involvement and Engagement section. In addition to the final publication and supplementary tables, the following datasets are published:

• number and weight of fish caught, netting effort, and rod effort (for salmon), by year, month, method and statistical salmon fishery district
• number of fish caught by rod and rod effort (for salmon), by year, month and stock assessment area
• form and rod effort return rate, all Scotland

The data is used in national (final statistics) and international (provisional statistics) Atlantic salmon stock assessments, and by FMS and District Salmon Fishery Boards to monitor trends in salmon catch at national, statistical salmon fishery district and stock assessment level. Feedback from stock assessment scientists, researchers and policy makers informs the development of publications, datasets and statistical outputs.

Accuracy

Our data is provided to us directly from salmon fishery proprietors, their tenants, or agents. Data providers collate information provided to them by anglers using their fishery. The data is reliant on the angler’s (or ghillie’s) ability to identify migratory fish to species level, and to assess the sea-age of salmon and the maturity of sea trout. It is also reliant on catch and effort information being passed to the fishery operator, or their agent. Most caught salmon are released, and catch and release best practice guidelines state that salmon should not be handled or removed from the water. Therefore, the weights of released fish are assumed to be estimates.

The information published is the summation of the reported data. No attempt is made to correct data for non-returns or incomplete coverage of existing fisheries. Form return rates are published from 1997 onwards, and are typically around 92% for the most recent season. This rises to around 94% in the next season’s publication following the addition of late returns. No attempt is made to correct for “grilse error”, the mis-classification by fishers between one sea-winter salmon (1SW/grilse) and multi sea-winter salmon (MSW).

Timeliness

Provisional statistics are published in late February, less than 2 months after the deadline for the most recent season’s returns. This time allows for the inclusion of returns received up to one month after the return deadline, and for quality assurance to be carried out. Historically, the final statistics have been published in April. More recently, they have been published in May due, initially, to the Coronavirus pandemic, then by resources pressures and finally the inclusion of salmon rod effort for the first time in the 2024 season publication. Publication reverted to April for the 2025 season publication. This time allows for production of the statistical summary, supplementary tables, accompanying datasets and the quality assurance of these products.

Accessibility

The statistical summary follows the Scottish government statistical template and accessible publication guidance and is provided as accessible html webpages and as downloadable (pdf) documents. The text is written in plain English and charts are provided with alternative text. The supplementary tables and accompanying datasets are published in Excel format and follow accessible spreadsheet guidelines. The accompanying datasets are also published in comma-separated value (csv) format to allow machine reading.

Comparability

In general, the data is comparable over time and with other salmon and sea trout catch data from across the UK and Europe. England, Wales and Ireland sell rod licenses to anglers and the individual anglers are responsible for making returns to the relevant authorities. In Norway, angler returns are collected by land owners, and provided to County Governors who make returns to Statistics Norway. Return rates differ: for Scotland, the Marine Directorate received 94% of returns in 2023, whereas return rates for England and Wales, for Ireland and for Norway, in 2023, were 59%, 49% and 96%, respectively.

Communicating uncertainty

The key strengths of the data are its timeliness, longevity and coverage. It is a legal requirement for proprietors, tenants, or their agents to provide this information when asked. This does not mean that the information is 100% complete or accurate. To help users assess quality and uncertainty, annual form returns rates are published in the statistical summary, supplementary tables, and supporting datasets. Weights reported for released fish are assumed to be estimates. Species identification, sea-age of salmon and maturity of sea trout are as reported, but where average weights do not meet expectations the data provider is asked to review the information reported and correct if necessary. There is no correction for non-returns or for grilse error. These uncertainties are communicated in factsheets linked to from the publication.

Quality assurance process

Reported data is verified against a series of checks which are clearly documented and run at least weekly during the data collection period. Where necessary, the data is referred back to the person completing the questionnaire for clarification. The information published, however, is the summation of the reported data. No attempt is made to correct the data for non-returns or incomplete coverage of existing fisheries. Where weights cannot be provided, the district average for that season is used. If no district average is available, regional averages are used.

Standard reports are used to produce the supplementary tables that inform the statistical summary. Standard reports are also used to extract catch and net effort data used for the accompanying datasets. However, rod effort and released net catch must be added separately to these reports. Due to this not being automated, checks are carried out comparing different extractions and previous publications to ensure accuracy.

Post-production, the statistical summary and supplementary tables are quality assured by subject matter experts.

As part of quality assurance, a range of quality indicators are reviewed, including return rates, follow-up queries and the use of estimated weights. These indicators help identify emerging quality issues and guide future improvements to the collection process.

Aquaculture production surveys (fish and shellfish)

This section applies to the Scottish Shellfish Farm Production Survey and the Scottish Fish Farm Production Survey.

Proportionality

The Scottish Aquaculture Production Surveys are fit for their intended purpose, which is to provide information on aquaculture production in Scotland. The surveys collect information on production volumes, species and other industry measures and provide an important evidence base for understanding the performance and development of the Scottish aquaculture sector.

Relevance

The key stakeholders for these statistics are summarised in the Public Involvement and Engagement section. The statistics are used to monitor trends in aquaculture production, inform policy development and implementation, and support economic and environmental analysis.

Accuracy

The surveys are designed as censuses of active aquaculture businesses and authorised farm sites operating in Scotland and therefore do not rely on a sample of the population. Consequently, the statistics are not subject to sampling error. The quality of the statistics depends on the completeness and accuracy of information supplied by respondents.

Survey returns are checked for completeness and consistency, with follow-up undertaken where clarification is required. Comparisons with previous years' data and other available information sources are used to identify unexpected changes, anomalies or potential reporting errors. The surveys have consistently achieved a 100% response rate. Quality assurance activities therefore focus on ensuring the completeness, consistency and accuracy of the information provided by respondents.

Timeliness

The surveys are conducted annually following the end of the production year. Time is required to collect responses, validate returns, resolve data queries and undertake quality assurance checks. Publications are released once these processes have been completed and the outputs have been reviewed by relevant scientific and technical experts.

Accessibility

Publications follow the Scottish Government statistical publication template, accessible publication guidance and analytical guidance on accessible tables. The text is written in plain English and charts are accompanied by alternative text. Key findings are presented at the start of publications and supplementary tables are provided to enable further analysis by users.

Comparability

The surveys have been conducted over a number of years using broadly consistent methodologies, allowing trends in aquaculture production to be monitored over time. Where changes to definitions, methodology or coverage affect comparability, these are explained within the relevant publication.

The statistics are broadly comparable with equivalent aquaculture statistics produced elsewhere in the UK, although differences in data collection arrangements, methodology and industry structure should be considered when making comparisons.

Communicating Uncertainty

The statistics are based on survey returns and are therefore subject to potential non-response, reporting errors and revisions identified through validation processes. Extensive follow-up activity is undertaken to maximise response rates and improve data quality.

Any limitations, methodological changes or issues affecting interpretation are explained within the publications and accompanying tables. Users should exercise caution when interpreting small year-on-year changes, as these may reflect normal reporting variation rather than underlying changes in production.

Quality Assurance Process

Survey returns undergo a range of validation and consistency checks. These include checks for missing values, inconsistencies between related variables, unexpected year-on-year changes and values outside expected ranges. Identified issues are investigated and respondents are contacted where clarification is required.

Detailed process documentation, documented analytical code and standard input files support the consistent production of publication outputs and reduce the risk of processing errors. Statistical outputs are reviewed by analysts and discussed with relevant scientific and technical experts in aquaculture to ensure they are consistent with sector knowledge and expectations prior to publication.

Outputs are also reviewed prior to publication to ensure that confidential business information is protected and that published statistics comply with relevant disclosure control requirements. Any known limitations or remaining issues are communicated to users through the publication and accompanying tables.