{"id":51,"date":"2024-04-04T00:39:48","date_gmt":"2024-04-03T21:39:48","guid":{"rendered":"https:\/\/sisu.ut.ee\/lcms_method_validation\/7-accuracy\/"},"modified":"2025-10-17T16:15:37","modified_gmt":"2025-10-17T13:15:37","slug":"7-accuracy","status":"publish","type":"page","link":"https:\/\/sisu.ut.ee\/lcms_method_validation\/7-accuracy\/","title":{"rendered":"7. Accuracy"},"content":{"rendered":"<h5 style=\"text-align: center;\"><div class=\"ratio ratio-16x9 mb-3\"><div class=\"video-placeholder-wrapper video-placeholder-wrapper--16x9\">\n\t\t\t    <div class=\"video-placeholder d-flex justify-content-center align-items-center\">\n\t\t\t        <div class=\"overlay text-white p-2 w-100 text-center d-block justify-content-center align-items-center\">\n\t\t\t            <div>To view third-party content, please accept cookies.<\/div>\n\t\t\t            <button class=\"btn btn-secondary btn-sm mt-1 consent-change\">Change consent<\/button>\n\t\t\t        <\/div>\n\t\t\t    <\/div>\n\t\t\t<\/div>\n<\/div>Accuracy\u00a0\u2013 relations to different concepts<\/h5>\n<h5 style=\"text-align: center;\"><a href=\"http:\/\/www.uttv.ee\/naita?id=23668\" target=\"_blank\" rel=\"noopener\">http:\/\/www.uttv.ee\/naita?id=23668<\/a><\/h5>\n<h5 style=\"text-align: center;\"><a href=\"https:\/\/www.youtube.com\/watch?v=iunClGAivzo\" target=\"_blank\" rel=\"noopener\">https:\/\/www.youtube.com\/watch?v=iunClGAivzo<\/a><br>\u00a0<\/h5>\n<p>Measurement result <strong>accuracy<\/strong> indicates its closeness to the true value [ref 6].\u00a0Accuracy differs from trueness: accuracy can be used to characterize an individual result, but trueness always refers to the mean value of a large number of results. Because of that and because accuracy can characterize an individual result, accuracy involves also precision. So, the accuracy of a method (i.e. accuracy of the results delivered by the method)\u00a0is affected by a systematic (bias) as well as random (precision) error components and is, therefore, studied as two components: trueness and precision [ref 2].<\/p>\n<p style=\"text-align: center;\"><img loading=\"lazy\" decoding=\"async\" width=\"1434\" height=\"797\" class=\"alignnone wp-image-426\" style=\"width: 700px; height: 389px;\" title=\"accuracy_16_9_gray_blue_01.png\" src=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/accuracy_16_9_gray_blue_01.png\" alt=\"accuracy\" srcset=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/accuracy_16_9_gray_blue_01.png 1434w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/accuracy_16_9_gray_blue_01-300x167.png 300w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/accuracy_16_9_gray_blue_01-1024x569.png 1024w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/accuracy_16_9_gray_blue_01-768x427.png 768w\" sizes=\"auto, (max-width: 1434px) 100vw, 1434px\"><\/p>\n<h4 style=\"text-align: center;\">Figure 1. Interrelations between the different error types, the performance characteristics used to estimate them and the ways of expressing the estimates quantitatively. This type of scheme was originally published in\u00a0ref 56.\u00a0<a href=\"#\" data-bs-toggle=\"modal\" data-bs-target=\"#popup-modal\" data-title=\"(See Note 1)\" data-content=\"Literature reference in the video has incorrect number of year. The correct reference is as stated here:\u00a0A. Menditto, et al &lt;em&gt;Accred. Qual. Assur&lt;\/em&gt;. 2007, 12, 45&lt;br \/&gt;\">(See Note 1)<\/a><\/h4>\n<h5 style=\"text-align: center;\"><div class=\"ratio ratio-16x9 mb-3\"><div class=\"video-placeholder-wrapper video-placeholder-wrapper--16x9\">\n\t\t\t    <div class=\"video-placeholder d-flex justify-content-center align-items-center\">\n\t\t\t        <div class=\"overlay text-white p-2 w-100 text-center d-block justify-content-center align-items-center\">\n\t\t\t            <div>To view third-party content, please accept cookies.<\/div>\n\t\t\t            <button class=\"btn btn-secondary btn-sm mt-1 consent-change\">Change consent<\/button>\n\t\t\t        <\/div>\n\t\t\t    <\/div>\n\t\t\t<\/div>\n<\/div>Accuracy, trueness, precision and measurement uncertainty<\/h5>\n<h5 style=\"text-align: center;\"><a style=\"text-align: center; font-family: inherit; font-weight: 600;\" href=\"http:\/\/www.uttv.ee\/naita?id=23345\" target=\"_blank\" rel=\"noopener\">http:\/\/www.uttv.ee\/naita?id=23345<\/a><\/h5>\n<h5 style=\"text-align: center;\"><a href=\"https:\/\/www.youtube.com\/watch?v=NfEsN1Gaq5k\" target=\"_blank\" rel=\"noopener\">https:\/\/www.youtube.com\/watch?v=NfEsN1Gaq5k<\/a><br>\u00a0<\/h5>\n<p>A number of regulatory bodies (ICH, FDA, EMA) define accuracy as the degree of agreement between the experimental value, obtained by replicate measurements, and the accepted reference value. This definition is identical to the currently accepted\u00a0definition of trueness. For the evaluation of acceptability of measurement\u00a0accuracy, different evaluation criteria can be used: <em>E<\/em><sub>n<\/sub> -numbers, <em>z<\/em>-scores or zeta-scores.<\/p>\n<p>Accuracy is often considered as a qualitative term [ref 6]. However, in practice it is useful to consider that accuracy is quantitatively expressed as a measurement uncertainty. There are different approaches for the measurement uncertainty estimation, but in practice the approach based on the\u00a0validation data is often the most convenient. The following video explains the basics of this approach. In-depth treatment of the topic of measurement uncertainty, together with numerous examples and self-tests can be found in the on-line course <a href=\"https:\/\/sisu.ut.ee\/measurement\/\" target=\"_blank\" rel=\"noopener\">Estimation of Measurement Uncertainty in Chemical Analysis<\/a>. Uncertainty estimation on the basis of validation and quality control data is covered in section <a href=\"https:\/\/sisu.ut.ee\/measurement\/node\/613\" target=\"_blank\" rel=\"noopener\">10. The Single-lab validation approach<\/a>.<\/p>\n<p><\/p><div class=\"ratio ratio-16x9 mb-3\"><div class=\"video-placeholder-wrapper video-placeholder-wrapper--16x9\">\n\t\t\t    <div class=\"video-placeholder d-flex justify-content-center align-items-center\">\n\t\t\t        <div class=\"overlay text-white p-2 w-100 text-center d-block justify-content-center align-items-center\">\n\t\t\t            <div>To view third-party content, please accept cookies.<\/div>\n\t\t\t            <button class=\"btn btn-secondary btn-sm mt-1 consent-change\">Change consent<\/button>\n\t\t\t        <\/div>\n\t\t\t    <\/div>\n\t\t\t<\/div>\n<\/div>\n<h5 style=\"text-align: center;\">Measurement uncertainty estimation approaches<\/h5>\n<h5 style=\"text-align: center;\"><a href=\"http:\/\/www.uttv.ee\/naita?id=23667\" target=\"_blank\" rel=\"noopener\">http:\/\/www.uttv.ee\/naita?id=23667<\/a><\/h5>\n<h5 style=\"text-align: center;\"><a href=\"https:\/\/www.youtube.com\/watch?v=syB2RKAEeMs&amp;t=46s\" target=\"_blank\" rel=\"noopener\">https:\/\/www.youtube.com\/watch?v=syB2RKAEeMs&amp;t=46s<\/a><br>\u00a0<\/h5>\n<p style=\"text-align: center;\"><img loading=\"lazy\" decoding=\"async\" width=\"550\" height=\"251\" class=\"alignnone wp-image-424\" title=\"7-fig2_uus2.png\" src=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/7-fig2_uus2.png\" alt=\"7-fig2_uus2.png\" srcset=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/7-fig2_uus2.png 550w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/7-fig2_uus2-300x137.png 300w\" sizes=\"auto, (max-width: 550px) 100vw, 550px\"><\/p>\n<h4 style=\"text-align: center;\">Figure 2. The main idea of estimating measurement uncertainty using validation and quality control data.<\/h4>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<h1><strong>Zeta-scores, z-scores and <em>E<\/em><sub>n<\/sub> numbers<\/strong><\/h1>\n<p>Zeta-scores (\u03b6-scores) and <em>E<\/em><sub>n<\/sub> numbers [ref 69] are used for comparing a measurement result with a reference value (e.g. certified value of a certified reference material or independent reference value of an interlaboratory comparison) or two measurement results against each other, to determine whether they are in agreement. An important condition for using them is that uncertainty estimates are needed for both compared values.<\/p>\n<p>Zeta-scores are calculated as follows:<\/p>\n<h6 style=\"text-align: center;\">\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <img loading=\"lazy\" decoding=\"async\" class=\"alignnone wp-image-1543\" src=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/zeta-eq.png\" alt=\"\" width=\"304\" height=\"60\" srcset=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/zeta-eq.png 587w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/zeta-eq-300x59.png 300w\" sizes=\"auto, (max-width: 304px) 100vw, 304px\">\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0(Eq 1)<\/h6>\n<p><em>X<\/em><sub>lab<\/sub> is the result obtained by the lab<br><em>X<\/em><sub>ref<\/sub> is the <strong>reference value<\/strong> (or the result obtained by another lab)<br><em>u<\/em><sub>c_lab<\/sub> is the <strong>combined standard uncertainty<\/strong> of lab\u2019s result<br><em>u<\/em><sub>c_ref<\/sub> is the <strong>combined standard uncertainty<\/strong> of the reference value (or the result obtained by another lab)<\/p>\n<p>Agreement between the two results is determined as follows:<em><sup>a<\/sup><\/em><\/p>\n<table class=\"table table-hover\">\n<tbody>\n<tr>\n<td>\n<p><strong><em>zeta-score<\/em><\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Result<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Action to be taken by the lab<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>|<em>zeta<\/em>| \u2264 2<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Results agree<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>None<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>2 &lt; |<em>zeta<\/em>| &lt; 3<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Unclear situation<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Preventative action<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>|<em>zeta<\/em>| <\/strong><strong>\u00b3<\/strong><strong> 3<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Results disagree<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Corrective action<\/strong><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><em><sup>a<\/sup><\/em> |<em>zeta<\/em>| denotes absolute value of <em>zeta<\/em>.<\/p>\n<p><em>E<\/em><sub>n<\/sub> numbers are calculated as follows:<\/p>\n<h6 style=\"text-align: center;\">\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 <img loading=\"lazy\" decoding=\"async\" class=\"alignnone wp-image-1545\" src=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/En-eq.png\" alt=\"\" width=\"236\" height=\"90\" srcset=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/En-eq.png 458w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/En-eq-300x115.png 300w\" sizes=\"auto, (max-width: 236px) 100vw, 236px\"> \u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 (Eq 2)<\/h6>\n<p><em>X<\/em><sub>lab<\/sub> is the result obtained by the lab<br><em>X<\/em><sub>ref<\/sub> is the <strong>reference value<\/strong> (or the result obtained by another lab)<br><em>U<\/em><sub>lab<\/sub> is the <strong>expanded uncertainty<\/strong> of lab\u2019s result, typically at <em>k<\/em> = 2 level<br><em>U<\/em><sub>ref<\/sub> is the <strong>combined standard uncertainty<\/strong> of the reference value (or the result obtained by another lab), typically at <em>k<\/em> = 2 level<\/p>\n<p>Agreement between is determined using the following table:<em><sup>a<\/sup><\/em><\/p>\n<table class=\"table table-hover\">\n<tbody>\n<tr>\n<td>\n<p><strong><em>E<\/em><\/strong><strong><sub>n<\/sub><\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Result<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Action to be taken by the lab<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>|<em>E<\/em><sub>n<\/sub>| \u2264 1<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Results agree<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>None<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>|<em>E<\/em><sub>n<\/sub>| &gt; 1<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Results disagree<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Corrective action<\/strong><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><em><sup>a<\/sup><\/em> |<em>z<\/em>| denotes absolute value of <em>z<\/em>.<\/p>\n<p><strong>Z-Scores<\/strong><\/p>\n<p>Z-Sores [ref 69] are the most widely used measure for evaluating the results of proficiency tests (PT, interlaboratory comparison measurements aiming at evaluating the proficiency of participating laboratories). The main advantages of z-sores against other typical measures \u2013 <em>E<\/em><sub>n<\/sub> numbers and zeta-scores \u2013 are that there is no need for an independent reference value \u2013 a <strong>consensus value of the participating laboratories<\/strong> is used instead \u2013 and uncertainty estimates of the values are not needed. Calculating the z-score of a PT participant:<\/p>\n<h6 style=\"text-align: center;\">\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 <img loading=\"lazy\" decoding=\"async\" class=\"alignnone wp-image-1544\" src=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/z-score-eq.png\" alt=\"\" width=\"197\" height=\"70\" srcset=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/z-score-eq.png 330w, https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/z-score-eq-300x106.png 300w\" sizes=\"auto, (max-width: 197px) 100vw, 197px\"> \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0(Eq 3)<\/h6>\n<p><em>X<\/em><sub>lab<\/sub> is the participant result<br><em>X<\/em><sub>cons<\/sub> is the <strong>consensus value<\/strong> of the participating laboratories<br><em>s<\/em><sub>target<\/sub> is the <strong>target standard deviation<\/strong><\/p>\n<p>Participant performance is determined using the following table:<em><sup>a<\/sup><\/em><\/p>\n<table class=\"table table-hover\">\n<tbody>\n<tr>\n<td>\n<p><strong><em>z-score<\/em><\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Result<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Action to be taken by the lab<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>|<em>z<\/em>| \u2264 2<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Good<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>None<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>2 &lt; |<em>z<\/em>| &lt; 3<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Warning<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Preventative action<\/strong><\/p>\n<\/td>\n<\/tr>\n<tr>\n<td>\n<p><strong>|<em>z<\/em>| <\/strong><strong>\u00b3<\/strong><strong> 3<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Unacceptable<\/strong><\/p>\n<\/td>\n<td>\n<p><strong>Corrective action<\/strong><\/p>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><em><sup>a<\/sup><\/em> |<em>z<\/em>| denotes absolute value of <em>z<\/em>.<\/p>\n<p>To use z-scores the <strong>consensus value<\/strong> and the <strong>target standard deviation <\/strong>has to be evaluated.<\/p>\n<p>If the intercomparison has no independent reference value then the consensus value has to be <strong>evaluated from the participant data<\/strong>. The consensus value is re-evaluated in every round of an intercomparison. <strong>Outlying results are eliminated<\/strong> before calculating the consensus value. If this is not done then some strongly outlying results can strongly affect the whole analysis. <a href=\"#\" data-bs-toggle=\"modal\" data-bs-target=\"#popup-modal\" data-title=\"(See Note 2)\" data-content=\"For example, it is not unusual to have participant values differing from other submitted values by 1000 times. This can happen if mistake is made with units, e.g. a result that should have unit \u03bcg kg&lt;sup&gt;-1&lt;\/sup&gt; is presented with the unit mg kg&lt;sup&gt;-1&lt;\/sup&gt;.&lt;br \/&gt;\">(See Note 2)<\/a> Eliminating outliers is not a trivial task. Outlier elimination of an outlier must be based on <strong>rigorous statistical tests<\/strong>, not on intuition. Currently there is no one single universally acknowledged way of determining outliers. Grubbs and Cochran tests (see the Exploratory Data Analysis section of the NIST\/SEMATECH e-Handbook of Statistical Methods [ref 70]) are often used. After elimination of outliers, the consensus value is found as the <strong>mean <\/strong>or <strong>median <\/strong>of the participant results. In the case of median outliers are less influential. For this reason, median is often preferred over mean.<\/p>\n<p>The target standard deviation <em>s<\/em><sub>target<\/sub> is a quantity that describes the <strong>expected<\/strong> variance of the results of the participants. The target standard deviation can be determined in different ways. The actual standard deviation of the participant results after elimination of outliers is the most used way.<\/p>\n<p>***<\/p>\n<p>Note 1: Literature reference in the video has incorrect number of year. The correct reference is as stated here: A. Menditto, et al <em>Accred. Qual. Assur<\/em>. <strong>2007<\/strong>, 12, 45<br>Note 2: For example, it is not unusual to have participant values differing from other submitted values by 1000 times. This can happen if mistake is made with units, e.g. a result that should have unit \u03bcg kg<sup>-1<\/sup> is presented with the unit mg kg<sup>-1<\/sup>.<\/p>\n\n\n<div class=\"wp-block-group attached-files-group is-layout-constrained wp-block-group-is-layout-constrained\">\n<div class=\"wp-block-file\"><a href=\"https:\/\/sisu.ut.ee\/wp-content\/uploads\/sites\/130\/7_accuracy_and_measurement_uncertainty.pdf\" target=\"_blank\" rel=\"noreferrer noopener\">7_accuracy_and_measurement_uncertainty.pdf<\/a><\/div>\n<\/div>\n","protected":false},"excerpt":{"rendered":"<p>To view third-party content, please accept cookies. Change consent Accuracy\u00a0\u2013 relations to different concepts http:\/\/www.uttv.ee\/naita?id=23668 https:\/\/www.youtube.com\/watch?v=iunClGAivzo\u00a0 Measurement result accuracy indicates its closeness to the true value [ref 6].\u00a0Accuracy differs from trueness: accuracy can be used to characterize an individual result, &#8230;<\/p>\n","protected":false},"author":60,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"_acf_changed":false,"inline_featured_image":false,"footnotes":""},"class_list":["post-51","page","type-page","status-publish","hentry"],"acf":[],"_links":{"self":[{"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/pages\/51","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/users\/60"}],"replies":[{"embeddable":true,"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/comments?post=51"}],"version-history":[{"count":8,"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/pages\/51\/revisions"}],"predecessor-version":[{"id":1554,"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/pages\/51\/revisions\/1554"}],"wp:attachment":[{"href":"https:\/\/sisu.ut.ee\/lcms_method_validation\/wp-json\/wp\/v2\/media?parent=51"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}