6. Accuracy and reliability |
Top |
|
|
6.1. Accuracy - overall |
6.1.1 How good is the accuracy? |
Very good |
6.1.2 What are the main factors lowering the accuracy? |
Non-response error Processing error Other |
6.1.3 If Other, please specify |
Late or lacking sales notes on landings in foreign ports or prices from foreign byers. |
6.1.4 Additional comments |
Every day logbooks, landing declarations and sales notes are attempted correlated and matched by the computer system, according to vessel registration number, date of landing, port, etc. Lacking or late sales notes or sales notes dated before the logbook (i.e. if landed around midnight) can occur and can lead to processing errors. A number of internal reports provide information on potential data errors in both logbooks and sales notes. Our data quality staff subsequently check these, and if errors are encountered, they are corrected manually in the database. However, it cannot be excluded that a few pass unnoticed. |
|
6.2. Sampling error |
Sample survey
These questions only apply to surveys. If there is more than one survey, please describe the main survey below and the additional ones in table 6.2 of the annexed Excel file |
6.2.1 Name/Title |
|
6.2.2 Methods used to assess the sampling error |
|
6.2.3 If Other, please specify |
|
6.2.4 Methods used to derive the extrapolation factor |
|
6.2.5 If Other, please specify |
|
6.2.6 If coefficients of variation are calculated, please describe the calculation methods and formulas |
|
6.2.7 Sampling error - indicators |
Please provide the coefficients of variation in the worksheet CV of the annexed Excel file |
6.2.8 Additional comments |
|
|
6.3. Non-sampling error |
See sections below. |
6.3.1. Coverage error |
Census
These questions only apply to censuses. If there is more than one census, please describe the main census below and the additional ones in table 6.3 of the annexed Excel file |
6.3.1.1 Name/Title |
|
Over-coverage
|
6.3.1.2 Does the sample frame include wrongly classified units that are out of scope? |
|
6.3.1.3 What methods are used to detect the out-of scope units? |
|
6.3.1.4 Does the sample frame include units that do not exist in practice? |
|
6.3.1.5 Over-coverage - rate |
|
6.3.1.6 Impact on the data quality |
|
Under-coverage
|
6.3.1.7 Does the sample frame include all units falling within the scope of this survey? |
|
6.3.1.8 If Not, which units are not included? |
|
6.3.1.9 How large do you estimate the proportion of those units? (%) |
|
6.3.1.10 Impact on the data quality |
|
Misclassification
|
6.3.1.11 Impact on the data quality |
|
Common units
|
6.3.1.12 Common units - proportion |
Not applicable. |
6.3.1.13 Additional comments |
|
Sample survey
These questions only apply to surveys. If there is more than one survey, please describe the main survey below and the additional ones in table 6.3 of the annexed Excel file |
6.3.1.14 Name/Title |
|
Over-coverage
|
6.3.1.15 Does the sample frame include wrongly classified units that are out of scope? |
|
6.3.1.16 What methods are used to detect the out-of scope units? |
|
6.3.1.17 Does the sample frame include units that do not exist in practice? |
|
6.3.1.18 Over-coverage - rate |
|
6.3.1.19 Impact on the data quality |
|
Under-coverage |
6.3.1.20 Does the sample frame include all units falling within the scope of this survey? |
|
6.3.1.21 If Not, which units are not included? |
|
6.3.1.22 How large do you estimate the proportion of those units? (%) |
|
6.3.1.23 Impact on the data quality |
|
Misclassification
|
6.3.1.24 Impact on the data quality |
|
Common units |
6.3.1.25 Common units - proportion |
Not applicable. |
6.3.1.26 Additional comments |
|
Administrative data
These questions only apply to administrative sources. If there is more than one administrative source, please describe the main source below and the additional ones in table 6.3 of the annexed Excel file |
6.3.1.27 Name/Title of the administrative source |
Sales note register |
Over-coverage
|
6.3.1.28 Does the administrative source include wrongly classified units that are out of scope? |
No |
6.3.1.29 What methods are used to detect the out-of scope units? |
Control and validation of data occurs in the process of entering data into the IT-system. A number of cross validation systems highlights potential errors in data. These are subsequently checked by our data quality staff, and if errors are encountered, they are corrected manually in the database.
Every day logbooks and sales notes are attempted correlated and matched by the computer system, according to vessel registration number, date of landing, port, etc. A number of internal reports provide information on potential data errors in both logbooks and sales notes. These are subsequently checked by our data quality staff, and if errors are encountered, they are corrected manually in the database. Occasionally The Danish Agrifish Agency launches campaigns targeted at specific data quality issues. Further the rules of Control Regulation 1224/2009 article 109 onwards is implemented in a VALID system that checks for inconsistencies according to the national plan sent to DGMARE.
|
6.3.1.30 Does the administrative source include units that do not exist in practice? |
No |
6.3.1.31 Over-coverage - rate |
None |
6.3.1.32 Impact on the data quality |
None |
Under-coverage
|
6.3.1.33 Does the administrative source include all units falling within the scope of this survey? |
Yes |
6.3.1.34 If Not, which units are not included? |
Illegal trade does occur, but at a very low scale and our control officers takes a strong line against it. |
6.3.1.35 How large do you estimate the proportion of those units? (%) |
Very low |
6.3.1.36 Impact on the data quality |
None |
Misclassification |
6.3.1.37 Impact on the data quality |
Low |
6.3.1.38 Additional comments |
|
|
6.3.2. Measurement error |
Census
These questions only apply to censuses. If there is more than one census, please describe the main census below and the additional ones in table 6.3 of the annexed Excel file |
6.3.2.1 Name/Title |
|
6.3.2.2 Is the questionnaire based on usual concepts for respondents? |
|
6.3.2.3 Number of censuses already performed with the current questionnaire? |
|
6.3.2.4 Preparatory testing of the questionnaire? |
|
6.3.2.5 Number of units participating in the tests? |
|
6.3.2.6 Explanatory notes/handbook for surveyors/respondents? |
|
6.3.2.7 On-line FAQ or Hot-line support for surveyors/respondents? |
|
6.3.2.8 Are there pre-filled questions? |
|
6.3.2.9 Percentage of pre-filled questions out of total number of questions |
|
6.3.2.10 Other actions taken for reducing the measurement error? |
|
6.3.2.11 Additional comments |
|
Sample survey
These questions only apply to surveys. If there is more than one survey, please describe the main survey below and the additional ones in table 6.3 of the annexed Excel file |
6.3.2.12 Name/Title |
|
6.3.2.13 Is the questionnaire based on usual concepts for respondents? |
|
6.3.2.14 Number of surveys already performed with the current questionnaire? |
|
6.3.2.15 Preparatory testing of the questionnaire? |
|
6.3.2.16 Number of units participating in the tests? |
|
6.3.2.17 Explanatory notes/handbook for surveyors/respondents? |
|
6.3.2.18 On-line FAQ or Hot-line support for surveyors/respondents? |
|
6.3.2.19 Are there pre-filled questions? |
|
6.3.2.20 Percentage of pre-filled questions out of total number of questions |
|
6.3.2.21 Other actions taken for reducing the measurement error? |
|
6.3.2.22 Additional comments |
|
|
6.3.3. Non response error |
Census
These questions only apply to censuses. If there is more than one census, please describe the main census below and the additional ones in table 6.3 of the annexed Excel file |
6.3.3.1 Name/Title of the survey |
|
6.3.3.2 Unit non-response - rate |
|
6.3.3.3 How do you evaluate the recorded unit non-response rate in the overall context? |
|
6.3.3.4 Measures taken for minimising the unit non-response |
|
6.3.3.5 If Other, please specify |
|
6.3.3.6 Item non-response rate |
|
6.3.3.7 Item non-response rate - Minimum |
|
6.3.3.8 Item non-response rate - Maximum |
|
6.3.3.9 Which items had a high item non-response rate? |
|
6.3.3.10 Additional comments |
|
Sample survey
These questions only apply to surveys. If there is more than one survey, please describe the main survey below and the additional ones in table 6.3 of the annexed Excel file |
6.3.3.11 Name/Title of the survey |
|
6.3.3.12 Unit non-response - rate |
|
6.3.3.13 How do you evaluate the recorded unit non-response rate in the overall context? |
|
6.3.3.14 Measures taken for minimising the unit non-response |
|
6.3.3.15 If Other, please specify |
|
6.3.3.16 Item non-response rate |
|
6.3.3.17 Item non-response rate - Minimum |
|
6.3.3.18 Item non-response rate - Maximum |
|
6.3.3.19 Which items had a high item non-response rate? |
|
6.3.3.20 Additional comments |
|
|
6.3.4. Processing error |
Census
These questions only apply to censuses. If there is more than one census, please describe the main census below and the additional ones in table 6.3 of the annexed Excel file |
6.3.4.1 Name/Title |
|
6.3.4.2 Imputation - rate |
|
6.3.4.3 Imputation - basis |
|
6.3.4.4 If Other, please specify |
|
6.3.4.5 Additional comments |
|
Sample survey
These questions only apply to surveys. If there is more than one survey, please describe the main survey below and the additional ones in table 6.3 of the annexed Excel file |
6.3.4.6 Name/Title |
|
6.3.4.7 Imputation - rate |
|
6.3.4.8 Imputation - basis |
|
6.3.4.9 If Other, please specify |
|
6.3.4.10 How do you evaluate the impact of imputation on Coefficients of Variation? |
|
6.3.4.11 Additionnal comments |
|
|
6.3.5. Model assumption error |
Not applicable. |
6.4. Seasonal adjustment |
Not applicable. |
6.5. Data revision - policy |
Logbooks and sales notes are processed daily. Every night a computerized system ensures correlation between logbooks and sales notes on the basis of a complex algorithm. Furthermore overnight new statistical data sets are compiled for management and control purposes. For use in the information system for statistics, new data sets with preliminary figures are released about two weeks after the end of each month. Thus, preliminary figures for a year are released by the end of January the following year. It is difficult to estimate when final figures are available since corrections to information can be necessary due to more recent and correct information about data. It is the objective to obtain as updated figures as possible in the databases. However, after the month of June there are few corrections with regard to figures for the previous year. For statistical purposes, a version of selected data from the previous year is ‘frozen’ each in the beginning of May. This is done in agreement with Statistics Denmark. |
6.6. Data revision - practice |
6.6.1 Data revision - average size |
Not available. |
6.6.2 Were data revisions due to conceptual changes (e.g. new definitions) carried out since the last quality report? |
No |
6.6.3 What was the main reason for the revisions? |
|
6.6.4 How do you evaluate the impact of the revisions? |
|
6.6.5 Additional comments |
|
|