Key and supplementary comparisons - Print out

The pair-wise temperature differences and their associated uncertainties were calculated for the various laboratories by two paths. Those paths are indicated by the terms ‘direct’ and ‘inclusive’, or the corresponding abbreviations ‘dir.’ and ‘inc.’, and yield results that are indicated in the equivalence matrices and graphs by the same terms. In some cases, the paths are identical.

The ‘direct’ differences are the temperature differences between the realizations of a pair of laboratories that were obtained using the shortest available path linking those laboratories. In some cases, this means that the laboratories are compared using measurements made on the same SPRT or SPRTs. Where necessary, however, NIST or NML serves as an intermediary, linking the measurements of the two laboratories.

The ‘inclusive’ differences are those in which all of the data obtained by the pair of laboratories involved are used in the calculations of their pair-wise temperature differences and their associated uncertainties. In inclusive differences, whether or not there is a direct linkage between the laboratories, NIST (and possibly also NML) may be used as an intermediary so that the temperature differences can be computed using all relevant data. For the inclusive path, NIST was selected as the intermediate laboratory because, overall, that results in a minimum number of intermediate laboratories in the pair-wise differences.

The expanded uncertainty, U_L (95%), listed in the matrices of equivalence and used in the ‘no baseline’ graphs of approximate equivalence is derived for each laboratory by combining the Type A standard uncertainty for the smallest number of measurements (n) across SPRTs and the Type B standard uncertainty for each fixed point. The uncertainty U_L (95%) is only for the respective laboratory with no contribution from the other laboratories.

The pair-wise temperature differences and their associated uncertainties were calculated for the various laboratories by two paths. Those paths are indicated by the terms ‘direct’ and ‘inclusive’, or the corresponding abbreviations ‘dir.’ and ‘inc.’, and yield results that are indicated in the equivalence matrices and graphs by the same terms. In some cases, the paths are identical.

The ‘direct’ differences are the temperature differences between the realizations of a pair of laboratories that were obtained using the shortest available path linking those laboratories. In some cases, this means that the laboratories are compared using measurements made on the same SPRT or SPRTs. Where necessary, however, NIST or NML serves as an intermediary, linking the measurements of the two laboratories.

The ‘inclusive’ differences are those in which all of the data obtained by the pair of laboratories involved are used in the calculations of their pair-wise temperature differences and their associated uncertainties. In inclusive differences, whether or not there is a direct linkage between the laboratories, NIST (and possibly also NML) may be used as an intermediary so that the temperature differences can be computed using all relevant data. For the inclusive path, NIST was selected as the intermediate laboratory because, overall, that results in a minimum number of intermediate laboratories in the pair-wise differences.

The expanded uncertainty, U_L (95%), listed in the matrices of equivalence and used in the ‘no baseline’ graphs of approximate equivalence is derived for each laboratory by combining the Type A standard uncertainty for the smallest number of measurements (n) across SPRTs and the Type B standard uncertainty for each fixed point. The uncertainty U_L (95%) is only for the respective laboratory with no contribution from the other laboratories.

The pair-wise temperature differences and their associated uncertainties were calculated for the various laboratories by two paths. Those paths are indicated by the terms ‘direct’ and ‘inclusive’, or the corresponding abbreviations ‘dir.’ and ‘inc.’, and yield results that are indicated in the equivalence matrices and graphs by the same terms. In some cases, the paths are identical.

The ‘direct’ differences are the temperature differences between the realizations of a pair of laboratories that were obtained using the shortest available path linking those laboratories. In some cases, this means that the laboratories are compared using measurements made on the same SPRT or SPRTs. Where necessary, however, NIST or NML serves as an intermediary, linking the measurements of the two laboratories.

The ‘inclusive’ differences are those in which all of the data obtained by the pair of laboratories involved are used in the calculations of their pair-wise temperature differences and their associated uncertainties. In inclusive differences, whether or not there is a direct linkage between the laboratories, NIST (and possibly also NML) may be used as an intermediary so that the temperature differences can be computed using all relevant data. For the inclusive path, NIST was selected as the intermediate laboratory because, overall, that results in a minimum number of intermediate laboratories in the pair-wise differences.

The expanded uncertainty, U_L (95%), listed in the matrices of equivalence and used in the ‘no baseline’ graphs of approximate equivalence is derived for each laboratory by combining the Type A standard uncertainty for the smallest number of measurements (n) across SPRTs and the Type B standard uncertainty for each fixed point. The uncertainty U_L (95%) is only for the respective laboratory with no contribution from the other laboratories.

The pair-wise temperature differences and their associated uncertainties were calculated for the various laboratories by two paths. Those paths are indicated by the terms ‘direct’ and ‘inclusive’, or the corresponding abbreviations ‘dir.’ and ‘inc.’, and yield results that are indicated in the equivalence matrices and graphs by the same terms. In some cases, the paths are identical.

The ‘direct’ differences are the temperature differences between the realizations of a pair of laboratories that were obtained using the shortest available path linking those laboratories. In some cases, this means that the laboratories are compared using measurements made on the same SPRT or SPRTs. Where necessary, however, NIST or NML serves as an intermediary, linking the measurements of the two laboratories.

The ‘inclusive’ differences are those in which all of the data obtained by the pair of laboratories involved are used in the calculations of their pair-wise temperature differences and their associated uncertainties. In inclusive differences, whether or not there is a direct linkage between the laboratories, NIST (and possibly also NML) may be used as an intermediary so that the temperature differences can be computed using all relevant data. For the inclusive path, NIST was selected as the intermediate laboratory because, overall, that results in a minimum number of intermediate laboratories in the pair-wise differences.

The expanded uncertainty, U_L (95%), listed in the matrices of equivalence and used in the ‘no baseline’ graphs of approximate equivalence is derived for each laboratory by combining the Type A standard uncertainty for the smallest number of measurements (n) across SPRTs and the Type B standard uncertainty for each fixed point. The uncertainty U_L (95%) is only for the respective laboratory with no contribution from the other laboratories.

The pair-wise temperature differences and their associated uncertainties were calculated for the various laboratories by two paths. Those paths are indicated by the terms ‘direct’ and ‘inclusive’, or the corresponding abbreviations ‘dir.’ and ‘inc.’, and yield results that are indicated in the equivalence matrices and graphs by the same terms. In some cases, the paths are identical.

The ‘direct’ differences are the temperature differences between the realizations of a pair of laboratories that were obtained using the shortest available path linking those laboratories. In some cases, this means that the laboratories are compared using measurements made on the same SPRT or SPRTs. Where necessary, however, NIST or NML serves as an intermediary, linking the measurements of the two laboratories.

The ‘inclusive’ differences are those in which all of the data obtained by the pair of laboratories involved are used in the calculations of their pair-wise temperature differences and their associated uncertainties. In inclusive differences, whether or not there is a direct linkage between the laboratories, NIST (and possibly also NML) may be used as an intermediary so that the temperature differences can be computed using all relevant data. For the inclusive path, NIST was selected as the intermediate laboratory because, overall, that results in a minimum number of intermediate laboratories in the pair-wise differences.

The expanded uncertainty, U_L (95%), listed in the matrices of equivalence and used in the ‘no baseline’ graphs of approximate equivalence is derived for each laboratory by combining the Type A standard uncertainty for the smallest number of measurements (n) across SPRTs and the Type B standard uncertainty for each fixed point. The uncertainty U_L (95%) is only for the respective laboratory with no contribution from the other laboratories.

The pair-wise temperature differences and their associated uncertainties were calculated for the various laboratories by two paths. Those paths are indicated by the terms ‘direct’ and ‘inclusive’, or the corresponding abbreviations ‘dir.’ and ‘inc.’, and yield results that are indicated in the equivalence matrices and graphs by the same terms. In some cases, the paths are identical.

The ‘direct’ differences are the temperature differences between the realizations of a pair of laboratories that were obtained using the shortest available path linking those laboratories. In some cases, this means that the laboratories are compared using measurements made on the same SPRT or SPRTs. Where necessary, however, NIST or NML serves as an intermediary, linking the measurements of the two laboratories.

The ‘inclusive’ differences are those in which all of the data obtained by the pair of laboratories involved are used in the calculations of their pair-wise temperature differences and their associated uncertainties. In inclusive differences, whether or not there is a direct linkage between the laboratories, NIST (and possibly also NML) may be used as an intermediary so that the temperature differences can be computed using all relevant data. For the inclusive path, NIST was selected as the intermediate laboratory because, overall, that results in a minimum number of intermediate laboratories in the pair-wise differences.

The expanded uncertainty, U_L (95%), listed in the matrices of equivalence and used in the ‘no baseline’ graphs of approximate equivalence is derived for each laboratory by combining the Type A standard uncertainty for the smallest number of measurements (n) across SPRTs and the Type B standard uncertainty for each fixed point. The uncertainty U_L (95%) is only for the respective laboratory with no contribution from the other laboratories.