{"id":4029,"date":"2025-06-17T19:57:55","date_gmt":"2025-06-17T10:57:55","guid":{"rendered":"https:\/\/best-biostatistics.com\/toukei-er\/?p=4029"},"modified":"2025-07-05T13:00:22","modified_gmt":"2025-07-05T04:00:22","slug":"repeated-measures-data-analysis-group-time-comparison-using-em-means","status":"publish","type":"post","link":"https:\/\/best-biostatistics.com\/toukei-er\/entry\/repeated-measures-data-analysis-group-time-comparison-using-em-means\/","title":{"rendered":"\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u306e\u89e3\u6790\uff1aEM\u5e73\u5747\u3092\u7528\u3044\u305f\u7fa4\u9593\u30fb\u6642\u70b9\u9593\u6bd4\u8f03"},"content":{"rendered":"\n<p>\u81e8\u5e8a\u7814\u7a76\u3067\u306f\u3001\u540c\u4e00\u306e\u5bfe\u8c61\u8005\u306b\u5bfe\u3057\u3066\u8907\u6570\u56de\u6e2c\u5b9a\u3092\u884c\u3046<strong>\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30b6\u30a4\u30f3<\/strong>\u304c\u983b\u7e41\u306b\u7528\u3044\u3089\u308c\u308b\u3002\u3053\u306e\u3088\u3046\u306a\u30c7\u30fc\u30bf\u306f\u3001\u6642\u9593\u7d4c\u904e\u306b\u4f34\u3046\u5909\u5316\u3084\u4ecb\u5165\u52b9\u679c\u3092\u8a55\u4fa1\u3059\u308b\u4e0a\u3067\u975e\u5e38\u306b\u6709\u7528\u3060\u304c\u3001\u4e00\u65b9\u3067\u3001\u8907\u96d1\u306a\u76f8\u95a2\u69cb\u9020\u3092\u30e2\u30c7\u30eb\u5316\u3059\u308b\u3068\u3044\u3046\u8ab2\u984c\u304c\u3042\u308b\u3002\u3053\u306e\u8907\u96d1\u306a\u76f8\u95a2\u69cb\u9020\u3092\u9069\u5207\u306b\u30e2\u30c7\u30eb\u5316\u3057\u3001\u7fa4\u9593\u304a\u3088\u3073\u6642\u70b9\u9593\u3067\u306e\u6bd4\u8f03\u3092\u884c\u3046\u305f\u3081\u306b\u3001<strong>\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (LMM)<\/strong> \u3084<strong>\u4e00\u822c\u5316\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (GLMM)<\/strong> \u304c\u7528\u3044\u3089\u308c\u308b\u3002\u672c\u7a3f\u3067\u306f\u3001\u3053\u308c\u3089\u306e\u6df7\u5408\u30e2\u30c7\u30eb\u3092\u7528\u3044\u3066\u89e3\u6790\u3057\u305f\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u306b\u304a\u3051\u308b\u7fa4\u9593\u6bd4\u8f03\u304a\u3088\u3073\u6642\u70b9\u9593\u6bd4\u8f03\u3092\u3001<strong>\u63a8\u5b9a\u5468\u8fba\u5e73\u5747 (Estimated Marginal Means: EM\u5e73\u5747)<\/strong> \u3092\u6d3b\u7528\u3057\u3066\u884c\u3046\u65b9\u6cd5\u306b\u3064\u3044\u3066\u89e3\u8aac\u3059\u308b\u3002\u7279\u306b\u3001\u6642\u70b9\u9593\u6bd4\u8f03\u306b\u304a\u3051\u308b<strong>\u591a\u91cd\u6bd4\u8f03\u8abf\u6574<\/strong>\u306e\u8003\u3048\u65b9\u306b\u3082\u7126\u70b9\u3092\u5f53\u3066\u3001R\u3092\u7528\u3044\u305f\u5177\u4f53\u7684\u306a\u8a08\u7b97\u65b9\u6cd5\u3068\u3068\u3082\u306b\u8a73\u8ff0\u3059\u308b\u3002<\/p>\n\n\n\n<!--more-->\n\n\n\n<h2 class=\"wp-block-heading\">EM \u5e73\u5747\u3092\u7528\u3044\u305f\u7fa4\u9593\u6bd4\u8f03\u30fb\u6642\u70b9\u9593\u6bd4\u8f03\u306e\u6982\u8981<\/h2>\n\n\n\n<p>EM\u5e73\u5747\u306f\u3001\u30e2\u30c7\u30eb\u306b\u3088\u3063\u3066\u4e88\u6e2c\u3055\u308c\u308b\u5404\u8981\u56e0\u6c34\u6e96\u306b\u304a\u3051\u308b\u5fdc\u7b54\u306e\u5e73\u5747\u5024\u3092\u63a8\u5b9a\u3059\u308b\u3082\u306e\u3067\u3042\u308b\u3002\u5171\u5909\u91cf\u306e\u5f71\u97ff\u3092\u8abf\u6574\u3057\u305f\u4e0a\u3067\u3001\u95a2\u5fc3\u306e\u3042\u308b\u8981\u56e0\uff08\u4f8b\u3048\u3070\u3001\u7fa4\u3084\u6642\u70b9\uff09\u306e\u52b9\u679c\u3092\u7d14\u7c8b\u306b\u8a55\u4fa1\u3067\u304d\u308b\u305f\u3081\u3001\u7fa4\u9593\u3084\u6642\u70b9\u9593\u3067\u306e\u6bd4\u8f03\u306b\u975e\u5e38\u306b\u9069\u3057\u3066\u3044\u308b\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>\u7fa4\u9593\u6bd4\u8f03:<\/strong> \u5404\u6642\u70b9\u306b\u304a\u3051\u308b\u7fa4\u9593\u306e\u5dee\u3001\u307e\u305f\u306f\u5168\u6642\u70b9\u3092\u901a\u3057\u3066\u306e\u7fa4\u9593\u306e\u5e73\u5747\u7684\u306a\u5dee\u3092\u8a55\u4fa1\u3059\u308b\u3002EM\u5e73\u5747\u3092\u7528\u3044\u308b\u3053\u3068\u3067\u3001\u4ed6\u306e\u5171\u5909\u91cf\u306e\u5f71\u97ff\u3092\u8abf\u6574\u3057\u305f\u4e0a\u3067\u306e\u7fa4\u9593\u306e\u6bd4\u8f03\u304c\u53ef\u80fd\u306b\u306a\u308b\u3002<\/li>\n\n\n\n<li><strong>\u6642\u70b9\u9593\u6bd4\u8f03:<\/strong> \u5404\u7fa4\u5185\u3067\u306e\u6642\u70b9\u9593\u306e\u5909\u5316\u3001\u307e\u305f\u306f\u5168\u7fa4\u3092\u901a\u3057\u3066\u306e\u6642\u70b9\u9593\u306e\u5e73\u5747\u7684\u306a\u5909\u5316\u3092\u8a55\u4fa1\u3059\u308b\u3002\u7279\u306b\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u3067\u306f\u3001\u6cbb\u7642\u4ecb\u5165\u306b\u3088\u308b\u6642\u9593\u7684\u306a\u52b9\u679c\u3092\u8a55\u4fa1\u3059\u308b\u4e0a\u3067\u91cd\u8981\u306a\u6bd4\u8f03\u3068\u306a\u308b\u3002<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">\u81e8\u5e8a\u7814\u7a76\u306e\u5177\u4f53\u4f8b<\/h2>\n\n\n\n<p>\u3053\u3053\u3067\u306f\u3001\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u304c\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb\u3084\u4e00\u822c\u5316\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb\u3067\u89e3\u6790\u3055\u308c\u308b\u5177\u4f53\u7684\u306a\u81e8\u5e8a\u7814\u7a76\u4f8b\u3092\u3001\u30c7\u30fc\u30bf\u306e\u7a2e\u985e\u5225\u306b\u7d39\u4ecb\u3057\u3066\u307f\u308b\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u9023\u7d9a\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>\u4f8b\uff1a\u9ad8\u8840\u5727\u60a3\u8005\u306b\u304a\u3051\u308b\u964d\u5727\u5264\u306e\u52b9\u679c<\/strong> <br>\u65b0\u3057\u3044\u964d\u5727\u5264A\u3068\u65e2\u5b58\u306e\u964d\u5727\u5264B\u306e\u52b9\u679c\u3092\u6bd4\u8f03\u3059\u308b\u305f\u3081\u306b\u3001\u9ad8\u8840\u5727\u60a3\u8005\u3092\u30e9\u30f3\u30c0\u30e0\u306b2\u7fa4\u306b\u5272\u308a\u4ed8\u3051\u3001\u670d\u85ac\u958b\u59cb\u524d\u30011\u30f6\u6708\u5f8c\u30013\u30f6\u6708\u5f8c\u30016\u30f6\u6708\u5f8c\u306b\u53ce\u7e2e\u671f\u8840\u5727\u3092\u6e2c\u5b9a\u3059\u308b\u3002\u76ee\u7684\u306f\u3001\u5404\u6642\u70b9\u3067\u306e\u7fa4\u9593\u5dee\u3001\u304a\u3088\u3073\u5404\u7fa4\u5185\u3067\u306e\u6642\u9593\u7d4c\u904e\u306b\u3088\u308b\u53ce\u7e2e\u671f\u8840\u5727\u306e\u5909\u5316\u3092\u8a55\u4fa1\u3059\u308b\u3053\u3068\u3067\u3042\u308b\u3002\u3053\u306e\u30c7\u30fc\u30bf\u306f\u9023\u7d9a\u5024\u3067\u3042\u308a\u3001\u6b63\u898f\u5206\u5e03\u3092\u4eee\u5b9a\u3067\u304d\u308b\u305f\u3081\u3001<strong>\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb<\/strong>\u304c\u9069\u7528\u3055\u308c\u308b\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u4e8c\u9805\u30a4\u30d9\u30f3\u30c8\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>\u4f8b\uff1a\u65b0\u898f\u6297\u4e0d\u5b89\u85ac\u306e\u6cbb\u7642\u53cd\u5fdc\u7387<\/strong> <br>\u4e0d\u5b89\u75c7\u60a3\u8005\u3092\u5bfe\u8c61\u306b\u3001<strong>\u65b0\u898f\u6297\u4e0d\u5b89\u85ac<\/strong>\u3068<strong>\u5f93\u6765\u85ac<\/strong>\u306e\u52b9\u679c\u3092\u6bd4\u8f03\u3059\u308b\u3002\u6cbb\u7642\u958b\u59cb\u5f8c1\u9031\u9593\u30012\u9031\u9593\u30014\u9031\u9593\u30018\u9031\u9593\u5f8c\u306b\u3001\u5404\u60a3\u8005\u304c\u300c\u6cbb\u7642\u306b\u53cd\u5fdc\u3057\u305f\uff08\u4e0d\u5b89\u75c7\u72b6\u304c50%\u4ee5\u4e0a\u6539\u5584\uff09\u300d\u304b\u300c\u53cd\u5fdc\u3057\u306a\u304b\u3063\u305f\u300d\u304b\u3092\u8a55\u4fa1\u3059\u308b\u3002<strong>\u76ee\u7684\u306f\u3001\u5404\u6642\u70b9\u306b\u304a\u3051\u308b\u6cbb\u7642\u53cd\u5fdc\u7387\u306e\u7fa4\u9593\u5dee\u3001\u304a\u3088\u3073\u5404\u7fa4\u5185\u3067\u306e\u6cbb\u7642\u53cd\u5fdc\u7387\u306e\u6642\u9593\u7684\u306a\u5909\u5316\u3092\u8a55\u4fa1\u3059\u308b\u3053\u3068\u3067\u3042\u308b\u3002<\/strong>\u5fdc\u7b54\u5909\u6570\u304c\u4e8c\u9805\uff08\u53cd\u5fdc\/\u975e\u53cd\u5fdc\uff09\u3067\u3042\u308b\u305f\u3081\u3001<strong>\u4e00\u822c\u5316\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (\u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u6df7\u5408\u30e2\u30c7\u30eb)<\/strong> \u304c\u9069\u7528\u3055\u308c\u308b\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u9806\u5e8f\u30ab\u30c6\u30b4\u30ea\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>\u4f8b\uff1a\u5909\u5f62\u6027\u819d\u95a2\u7bc0\u75c7\u60a3\u8005\u306e\u75bc\u75db\u75c7\u72b6\u6539\u5584\u8a55\u4fa1<\/strong> <br>\u5909\u5f62\u6027\u819d\u95a2\u7bc0\u75c7\u60a3\u8005\u306b\u5bfe\u3057\u3001\u6a19\u6e96\u7684\u306a\u85ac\u5264A\u3068\u65b0\u3057\u3044\u85ac\u5264B\u3092\u6bd4\u8f03\u3059\u308b\u3002\u6cbb\u7642\u958b\u59cb\u524d\uff081\u304b\u6708\u524d\u3068\u6bd4\u8f03\uff09\u30011\u304b\u6708\u5f8c\u30012\u30f6\u6708\u5f8c\u306b\u3001\u75bc\u75db\u75c7\u72b6\u306e\u6539\u5584\u306e\u7a0b\u5ea6\u3092\u300c\u5168\u304f\u6539\u5584\u306a\u3057\u300d\u300c\u5c11\u3057\u6539\u5584\u300d\u300c\u304b\u306a\u308a\u6539\u5584\u300d\u300c\u5b8c\u5168\u306b\u6539\u5584\u300d\u306e4\u6bb5\u968e\u3067\u8a55\u4fa1\u3059\u308b\u3002\u76ee\u7684\u306f\u3001\u5404\u6642\u70b9\u3067\u306e\u75bc\u75db\u75c7\u72b6\u6539\u5584\u8a55\u4fa1\u306e\u7fa4\u9593\u5dee\u3001\u304a\u3088\u3073\u5404\u7fa4\u5185\u3067\u306e\u75bc\u75db\u75c7\u72b6\u6539\u5584\u306e\u6642\u9593\u7684\u306a\u5909\u5316\u3092\u8a55\u4fa1\u3059\u308b\u3053\u3068\u3067\u3042\u308b\u3002\u5fdc\u7b54\u5909\u6570\u304c\u9806\u5e8f\u30ab\u30c6\u30b4\u30ea\u3067\u3042\u308b\u305f\u3081\u3001<strong>\u4e00\u822c\u5316\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (\u9806\u5e8f\u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u6df7\u5408\u30e2\u30c7\u30eb\u3001\u3088\u308a\u5177\u4f53\u7684\u306b\u306f\u7d2f\u7a4d\u30ed\u30b8\u30c3\u30c8\u6df7\u5408\u30e2\u30c7\u30eb)<\/strong> \u304c\u9069\u7528\u3055\u308c\u308b\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u30ab\u30a6\u30f3\u30c8\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>\u4f8b\uff1a\u5598\u606f\u60a3\u8005\u306e\u591c\u9593\u899a\u9192\u56de\u6570<\/strong> <br>\u5598\u606f\u60a3\u8005\u3092\u5bfe\u8c61\u306b\u3001\u65b0\u3057\u3044\u5438\u5165\u30b9\u30c6\u30ed\u30a4\u30c9\u85ac\u3068\u30d7\u30e9\u30bb\u30dc\u306e\u52b9\u679c\u3092\u6bd4\u8f03\u3059\u308b\u3002\u6cbb\u7642\u958b\u59cb\u5f8c1\u30f6\u6708\u3054\u3068\u306b\u3001\u5404\u60a3\u8005\u306e\u591c\u9593\u306b\u304a\u3051\u308b\u5598\u606f\u306b\u3088\u308b\u899a\u9192\u56de\u6570\u3092\u8a18\u9332\u3057\u30011\u30f6\u6708\u5f8c\u30013\u30f6\u6708\u5f8c\u30016\u30f6\u6708\u5f8c\u306b\u305d\u306e\u6765\u9662\u524d\u4e00\u30f6\u6708\u306e\u5408\u8a08\u56de\u6570\u3092\u8a55\u4fa1\u3059\u308b\u3002\u76ee\u7684\u306f\u3001\u5404\u6642\u70b9\u3067\u306e\u591c\u9593\u899a\u9192\u56de\u6570\u306e\u7fa4\u9593\u5dee\u3001\u304a\u3088\u3073\u5404\u7fa4\u5185\u3067\u306e\u591c\u9593\u899a\u9192\u56de\u6570\u306e\u6642\u9593\u7684\u306a\u5909\u5316\u3092\u8a55\u4fa1\u3059\u308b\u3053\u3068\u3067\u3042\u308b\u3002\u5fdc\u7b54\u5909\u6570\u304c\u30ab\u30a6\u30f3\u30c8\u30c7\u30fc\u30bf\u3067\u3042\u308a\u3001\u30dd\u30a2\u30bd\u30f3\u5206\u5e03\u306a\u3069\u3092\u4eee\u5b9a\u3067\u304d\u308b\u305f\u3081\u3001<strong>\u4e00\u822c\u5316\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (\u30dd\u30a2\u30bd\u30f3\u6df7\u5408\u30e2\u30c7\u30eb\u307e\u305f\u306f\u8ca0\u306e\u4e8c\u9805\u6df7\u5408\u30e2\u30c7\u30eb)<\/strong> \u304c\u9069\u7528\u3055\u308c\u308b\u3002<\/p>\n\n\n\n<div id=\"biost-2777266248\" class=\"biost- biost-entity-placement\"><p style=\"text-align: center;\"><span style=\"font-size: 20px;\"><strong><a href=\"https:\/\/best-biostatistics.com\/kmhl\">\uff1e\uff1e\u3082\u3046\u7d71\u8a08\u3067\u60a9\u3080\u306e\u306f\u7d42\u308f\u308a\u306b\u3057\u307e\u305b\u3093\u304b\uff1f\u00a0<\/a><\/strong><\/span><\/p>\r\n<a href=\"https:\/\/best-biostatistics.com\/kmhl\"><img class=\"aligncenter wp-image-2794 size-full\" src=\"https:\/\/best-biostatistics.com\/wp\/wp-content\/uploads\/2023\/11\/bn_r_03.png\" alt=\"\" width=\"500\" height=\"327\" \/><\/a>\r\n<p style=\"text-align: center;\"><span style=\"color: #ff0000; font-size: 20px;\"><strong><span class=\"marker2\">\u21911\u4e07\u4eba\u4ee5\u4e0a\u306e\u533b\u7642\u5f93\u4e8b\u8005\u304c\u8cfc\u8aad\u4e2d<\/span><\/strong><\/span><\/p><\/div><h2 class=\"wp-block-heading\">R\u3067\u306e\u8a08\u7b97\u65b9\u6cd5<\/h2>\n\n\n\n<p>R\u3067\u306e\u6df7\u5408\u30e2\u30c7\u30eb\u306e\u8a08\u7b97\u306b\u306f\u3001<code>lme4<\/code> \u30d1\u30c3\u30b1\u30fc\u30b8\uff08LMM, GLMM\uff09\u3001<code>nlme<\/code> \u30d1\u30c3\u30b1\u30fc\u30b8\uff08LMM\uff09\u3001<code>emmeans<\/code> \u30d1\u30c3\u30b1\u30fc\u30b8\uff08EM\u5e73\u5747\u306e\u8a08\u7b97\u3068\u591a\u91cd\u6bd4\u8f03\u8abf\u6574\uff09\u306a\u3069\u304c\u3088\u304f\u7528\u3044\u3089\u308c\u308b\u3002<\/p>\n\n\n\n<p>\u4ee5\u4e0b\u306b\u3001\u67b6\u7a7a\u306e\u30c7\u30fc\u30bf\u3092\u7528\u3044\u305f\u5404\u30e2\u30c7\u30eb\u306eR\u30b3\u30fc\u30c9\u4f8b\u3092\u793a\u3059\u3002<\/p>\n\n\n\n<p>\u307e\u305a\u3001\u30d1\u30c3\u30b1\u30fc\u30b8\u306e\u6e96\u5099\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code># \u5fc5\u8981\u306a\u30d1\u30c3\u30b1\u30fc\u30b8\u306e\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb (\u521d\u56de\u306e\u307f)\n# install.packages(\"lme4\")\n# install.packages(\"ordinal\")\n# install.packages(\"emmeans\")\n# install.packages(\"MASS\") # \u8ca0\u306e\u4e8c\u9805\u56de\u5e30\u7528\n# install.packages(\"dplyr\") # \u30c7\u30fc\u30bf\u64cd\u4f5c\u7528\n\nlibrary(lme4)\nlibrary(ordinal)\nlibrary(emmeans)\nlibrary(MASS) # for negative binomial GLMM\nlibrary(dplyr) # for data manipulation (optional, but useful)<\/code><\/pre>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u9023\u7d9a\u30c7\u30fc\u30bf (\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb)<\/h3>\n\n\n\n<p><strong>R \u8a08\u7b97\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code># \u30c0\u30df\u30fc\u30c7\u30fc\u30bf\u306e\u4f5c\u6210\nset.seed(123)\ndata_lmm &lt;- expand.grid(subject = 1:50, time = c(0, 1, 3, 6))\ndata_lmm$group &lt;- rep(rep(c(\"A\", \"B\"), each = 25), times = 4)\ndata_lmm$blood_pressure &lt;- rnorm(200, mean = 140, sd = 10) # \u57fa\u6e96\u5024\ndata_lmm$blood_pressure&#91;data_lmm$group == \"A\" &amp; data_lmm$time == 1] &lt;- data_lmm$blood_pressure&#91;data_lmm$group == \"A\" &amp; data_lmm$time == 1] - 5\ndata_lmm$blood_pressure&#91;data_lmm$group == \"A\" &amp; data_lmm$time == 3] &lt;- data_lmm$blood_pressure&#91;data_lmm$group == \"A\" &amp; data_lmm$time == 3] - 10\ndata_lmm$blood_pressure&#91;data_lmm$group == \"A\" &amp; data_lmm$time == 6] &lt;- data_lmm$blood_pressure&#91;data_lmm$group == \"A\" &amp; data_lmm$time == 6] - 15\ndata_lmm$blood_pressure&#91;data_lmm$group == \"B\" &amp; data_lmm$time == 1] &lt;- data_lmm$blood_pressure&#91;data_lmm$group == \"B\" &amp; data_lmm$time == 1] - 2\ndata_lmm$blood_pressure&#91;data_lmm$group == \"B\" &amp; data_lmm$time == 3] &lt;- data_lmm$blood_pressure&#91;data_lmm$group == \"B\" &amp; data_lmm$time == 3] - 4\ndata_lmm$blood_pressure&#91;data_lmm$group == \"B\" &amp; data_lmm$time == 6] &lt;- data_lmm$blood_pressure&#91;data_lmm$group == \"B\" &amp; data_lmm$time == 6] - 6\ndata_lmm$blood_pressure &lt;- round(data_lmm$blood_pressure)\ndata_lmm$subject &lt;- as.factor(data_lmm$subject)\ndata_lmm$time &lt;- as.factor(data_lmm$time)\n\nhead(data_lmm)\n\n# \u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb\u306e\u69cb\u7bc9\n# subject\u3054\u3068\u306e\u5207\u7247\u3068time\u306e\u52b9\u679c\u306b\u30e9\u30f3\u30c0\u30e0\u52b9\u679c\u3092\u4eee\u5b9a\nmodel_lmm &lt;- lmer(blood_pressure ~ group * time + (1 | subject), data = data_lmm)\n\n\n# \u4f8b: \u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (model_lmm) \u3092\u4f7f\u7528\n# \u5404\u6642\u70b9\u3067\u306e\u7fa4\u9593\u6bd4\u8f03\nemmeans_lmm_time_group &lt;- emmeans(model_lmm, ~ group | time)\nemmeans_lmm_time_group2 &lt;- emmeans(model_lmm, specs = \"group\", by = \"time\") # ~ \u3068 | \u306f\u305d\u308c\u305e\u308c specs, by \u3067\u66f8\u3044\u3066\u3082\u826f\u3044\nprint(emmeans_lmm_time_group)\nprint(emmeans_lmm_time_group2)\n\n# pair = TRUE \u3067\u5404\u6642\u70b9\u5185\u3067\u306e\u7fa4\u9593\u6bd4\u8f03\u3092\u884c\u3046\ncontrasts_lmm_time_group &lt;- pairs(emmeans_lmm_time_group)\nsummary(contrasts_lmm_time_group)\n\n# \u4f8b: \u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb (model_lmm) \u3092\u4f7f\u7528\n# \u5404\u7fa4\u5185\u3067\u306e\u6642\u70b9\u9593\u6bd4\u8f03\nemmeans_lmm_group_time &lt;- emmeans(model_lmm, ~ time | group)\n# pair = TRUE \u3067\u5404\u7fa4\u5185\u3067\u306e\u6642\u70b9\u9593\u6bd4\u8f03\u3092\u884c\u3046\ncontrasts_lmm_group_time &lt;- pairs(emmeans_lmm_group_time, adjust = \"none\", reverse = TRUE) # \u591a\u91cd\u6bd4\u8f03\u8abf\u6574\u3092\u9069\u7528\nsummary(contrasts_lmm_group_time)<\/code><\/pre>\n\n\n\n<p><strong>\u5b9f\u884c\u7d50\u679c\uff1a<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>&gt; summary(contrasts_lmm_time_group)\ntime = 0:\n contrast estimate   SE  df t.ratio p.value\n A - B       -1.36 2.66 192  -0.511  0.6101\n\ntime = 1:\n contrast estimate   SE  df t.ratio p.value\n A - B       -5.60 2.66 192  -2.103  0.0367\n\ntime = 3:\n contrast estimate   SE  df t.ratio p.value\n A - B       -6.32 2.66 192  -2.374  0.0186\n\ntime = 6:\n contrast estimate   SE  df t.ratio p.value\n A - B       -5.40 2.66 192  -2.028  0.0439\n\nDegrees-of-freedom method: kenward-roger <\/code><\/pre>\n\n\n\n<p>\u6642\u70b9\u3054\u3068\u306b\u6bd4\u8f03\u3057\u305f\u7d50\u679c\u3067\u3042\u308b\u3002\u30d9\u30fc\u30b9\u30e9\u30a4\u30f3\uff08time = 0\uff09\u306f\u7570\u306a\u308b\u3068\u306f\u8a00\u3048\u306a\u3044\u304c\u3001\u305d\u306e\u5f8c time = 1, 3, 6 \u3067\u306f 5 \u304b\u3089 6 \u7a0b\u5ea6\u306e\u5dee\u3067\u3001\u6709\u610f\u306b\u7570\u306a\u3063\u3066\u3044\u308b\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>&gt; summary(contrasts_lmm_group_time)\ngroup = A:\n contrast      estimate   SE  df t.ratio p.value\n time1 - time0    -4.48 2.66 144  -1.683  0.0946\n time3 - time0   -12.44 2.66 144  -4.672  &lt;.0001\n time3 - time1    -7.96 2.66 144  -2.990  0.0033\n time6 - time0   -12.48 2.66 144  -4.687  &lt;.0001\n time6 - time1    -8.00 2.66 144  -3.005  0.0031\n time6 - time3    -0.04 2.66 144  -0.015  0.9880\n\ngroup = B:\n contrast      estimate   SE  df t.ratio p.value\n time1 - time0    -0.24 2.66 144  -0.090  0.9283\n time3 - time0    -7.48 2.66 144  -2.809  0.0057\n time3 - time1    -7.24 2.66 144  -2.719  0.0073\n time6 - time0    -8.44 2.66 144  -3.170  0.0019\n time6 - time1    -8.20 2.66 144  -3.080  0.0025\n time6 - time3    -0.96 2.66 144  -0.361  0.7190\n\nDegrees-of-freedom method: kenward-roger <\/code><\/pre>\n\n\n\n<p>\u3053\u3061\u3089\u306f\u3001\u7fa4\u5225\u306e\u6642\u70b9\u9593\u6bd4\u8f03\u7d50\u679c\u3067\u3042\u308b\u3002group A \u3082 B \u3082 time0 \u307e\u305f\u306f time1 \u3068 time3 \u307e\u305f\u306f time6 \u3068\u306e\u6bd4\u8f03\u3067\u6709\u610f\u306b\u306a\u3063\u3066\u3044\u308b\u3053\u3068\u304b\u3089\u3001\u305d\u308c\u3089\u306e\u6642\u70b9\u9593\u304c\u7570\u306a\u308b\u3053\u3068\u304c\u793a\u5506\u3055\u308c\u3066\u3044\u308b\u3002estimate \u304b\u3089\u5224\u65ad\u3057\u3066 group A \u306e\u307b\u3046\u304c\u5927\u304d\u306a\u5909\u5316\u3067\u3042\u308b\u3053\u3068\u304c\u3046\u304b\u304c\u3048\u308b\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u4e8c\u9805\u30a4\u30d9\u30f3\u30c8\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>R \u8a08\u7b97\u4f8b<\/strong>\uff1a<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code># \u30c0\u30df\u30fc\u30c7\u30fc\u30bf\u306e\u4f5c\u6210\ndata_binomial &lt;- expand.grid(subject = 1:50, time = c(1, 2, 4, 8))\ndata_binomial &lt;- data_binomial&#91;order(data_binomial$subject), ]\ndata_binomial$group &lt;- rep(c(\"NewDrug\", \"Control\"), each = 100)\n# \u6cbb\u7642\u53cd\u5fdc\u7387\u3092 group \u3068 time \u306b\u5fdc\u3058\u3066\u8a2d\u5b9a\ndata_binomial$response_prob &lt;- ifelse(data_binomial$group == \"NewDrug\",\n    0.4 + data_binomial$time * 0.05,\n    0.2 + data_binomial$time * 0.02\n)\nset.seed(123)\ndata_binomial$response &lt;- rbinom(nrow(data_binomial), 1, data_binomial$response_prob)\ndata_binomial$subject &lt;- as.factor(data_binomial$subject)\ndata_binomial$time &lt;- as.factor(data_binomial$time)\n\n# \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u6df7\u5408\u30e2\u30c7\u30eb\u306e\u69cb\u7bc9\nmodel_binomial &lt;- glmer(response ~ group * time + (1 | subject), data = data_binomial, family = binomial(link = \"logit\"))\n\n\n# \u4f8b: \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u6df7\u5408\u30e2\u30c7\u30eb (model_binomial) \u3092\u4f7f\u7528\n# \u5404\u6642\u70b9\u3067\u306e\u7fa4\u9593\u6bd4\u8f03 (\u30aa\u30c3\u30ba\u6bd4\u3067\u8868\u793a)\nemmeans_binomial_time_group &lt;- emmeans(model_binomial, ~ group | time, type = \"response\") # type=\"response\" \u3067\u78ba\u7387\u30b9\u30b1\u30fc\u30eb\u306b\u5909\u63db\ncontrasts_binomial_time_group &lt;- pairs(emmeans_binomial_time_group, reverse=TRUE)\nsummary(contrasts_binomial_time_group)\n\n# \u4f8b: \u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u6df7\u5408\u30e2\u30c7\u30eb (model_binomial) \u3092\u4f7f\u7528\n# \u5404\u7fa4\u5185\u3067\u306e\u6642\u70b9\u9593\u6bd4\u8f03 (\u30aa\u30c3\u30ba\u6bd4\u3067\u8868\u793a)\nemmeans_binomial_group_time &lt;- emmeans(model_binomial, ~ time | group, type = \"response\")\ncontrasts_binomial_group_time &lt;- pairs(emmeans_binomial_group_time, adjust = \"bonferroni\")\nsummary(contrasts_binomial_group_time)<\/code><\/pre>\n\n\n\n<p><strong>\u5b9f\u884c\u7d50\u679c<\/strong>\uff1a<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>> summary(contrasts_binomial_time_group)\ntime = 1:\n contrast          odds.ratio    SE  df null z.ratio p.value\n NewDrug \/ Control       4.11  2.60 Inf    1   2.237  0.0253\n\ntime = 2:\n contrast          odds.ratio    SE  df null z.ratio p.value\n NewDrug \/ Control       2.13  1.34 Inf    1   1.200  0.2300\n\ntime = 4:\n contrast          odds.ratio    SE  df null z.ratio p.value\n NewDrug \/ Control       5.60  3.56 Inf    1   2.708  0.0068\n\ntime = 8:\n contrast          odds.ratio    SE  df null z.ratio p.value\n NewDrug \/ Control      16.12 12.40 Inf    1   3.601  0.0003\n\nTests are performed on the log odds ratio scale <\/code><\/pre>\n\n\n\n<p>\u6642\u70b9\u3054\u3068\u306e\u7fa4\u9593\u6bd4\u8f03\u3067\u3001NewDrug \u3068 Control \u3092 time = 1, 2, 4, 8 \u3067\u6bd4\u8f03\u3057\u3066\u3044\u308b\u3002\u305d\u306e\u7d50\u679c\u3001time = 2 \u4ee5\u5916\u3067\u306f\u3001\u7d71\u8a08\u5b66\u7684\u6709\u610f\u5dee\u3042\u308a\u3068\u306e\u7d50\u679c\u3067\u3042\u3063\u305f\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>> summary(contrasts_binomial_group_time)\ngroup = Control:\n contrast      odds.ratio    SE  df null z.ratio p.value\n time2 \/ time1      1.000 0.666 Inf    1   0.000  0.9999\n time4 \/ time1      1.497 0.957 Inf    1   0.632  0.5275\n time4 \/ time2      1.497 0.957 Inf    1   0.632  0.5274\n time8 \/ time1      0.429 0.333 Inf    1  -1.090  0.2756\n time8 \/ time2      0.429 0.333 Inf    1  -1.090  0.2757\n time8 \/ time4      0.286 0.216 Inf    1  -1.656  0.0976\n\ngroup = NewDrug:\n contrast      odds.ratio    SE  df null z.ratio p.value\n time2 \/ time1      0.519 0.301 Inf    1  -1.133  0.2574\n time4 \/ time1      2.040 1.240 Inf    1   1.176  0.2396\n time4 \/ time2      3.933 2.420 Inf    1   2.227  0.0260\n time8 \/ time1      1.682 0.998 Inf    1   0.876  0.3810\n time8 \/ time2      3.243 1.950 Inf    1   1.956  0.0505\n time8 \/ time4      0.824 0.513 Inf    1  -0.310  0.7563\n\nTests are performed on the log odds ratio scale <\/code><\/pre>\n\n\n\n<p>\u3053\u3061\u3089\u306f\u3001\u7fa4\u3054\u3068\u306e\u6642\u70b9\u9593\u6bd4\u8f03\u306e\u7d50\u679c\u3067\u3042\u308b\u3002NewDrug \u3067\u3001time = 4 \u3068 time = 2 \u306e\u6bd4\u8f03\u304c\u6709\u610f\u5dee\u3042\u308a\u3067\u3042\u3063\u305f\u3002\u660e\u3089\u304b\u306a\u90e8\u5206\u306f\u305d\u308c\u3060\u3051\u3067\u3042\u308a\u30012\u7fa4\u306e\u6642\u70b9\u9593\u6bd4\u8f03\u306b\u660e\u78ba\u306a\u50be\u5411\u306e\u9055\u3044\u306f\u898b\u3089\u308c\u306a\u304b\u3063\u305f\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u9806\u5e8f\u30ab\u30c6\u30b4\u30ea\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>R \u8a08\u7b97\u4f8b\uff1a<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code># \u4f8b\u3068\u3057\u3066\u3001\u67b6\u7a7a\u306e\u30c7\u30fc\u30bf\u3092\u751f\u6210\nset.seed(456)\nn_patients &lt;- 60 \nn_visits &lt;- 3\n\npatient_id &lt;- rep(1:n_patients, each = n_visits)\ntreatment &lt;- sample(c(\"DrugA\", \"DrugB\"), n_patients, replace = TRUE) %>% rep(each = n_visits)\nvisit &lt;- rep(c(\"Baseline\", \"Post_Treatment_1\", \"Post_Treatment_2\"), times = n_patients)\n\n# \u60a3\u8005\u3054\u3068\u306e\u30e9\u30f3\u30c0\u30e0\u306a\u52b9\u679c\uff08\u5207\u7247\uff09\u3092\u751f\u6210\nrandom_effect_patient &lt;- rnorm(n_patients, mean = 0, sd = 1.2) %>% rep(each = n_visits)\n\n# \u75c7\u72b6\u6539\u5584\u5ea6\u306e\u771f\u306e\u78ba\u7387\u7684\u306a\u50be\u5411\u3092\u8a2d\u5b9a\nlinear_predictor &lt;- case_when(\n    treatment == \"DrugA\" &amp; visit == \"Baseline\" ~ 0.2,\n    treatment == \"DrugA\" &amp; visit == \"Post_Treatment_1\" ~ 0.8,\n    treatment == \"DrugA\" &amp; visit == \"Post_Treatment_2\" ~ 1.2,\n    treatment == \"DrugB\" &amp; visit == \"Baseline\" ~ 0.1,\n    treatment == \"DrugB\" &amp; visit == \"Post_Treatment_1\" ~ 0.4,\n    treatment == \"DrugB\" &amp; visit == \"Post_Treatment_2\" ~ 0.7,\n    TRUE ~ NA_real_\n)\n\n# \u30e9\u30f3\u30c0\u30e0\u52b9\u679c\u3092\u7dda\u5f62\u4e88\u6e2c\u5b50\u306b\u52a0\u3048\u308b\nlinear_predictor_with_re &lt;- linear_predictor + random_effect_patient\n\n# \u7d2f\u7a4d\u78ba\u7387\u304b\u3089\u75c7\u72b6\u6539\u5584\u5ea6\u3092\u751f\u6210\nthresholds &lt;- c(-1.5, 0.5, 2.0) \n\n# \u75c7\u72b6\u30ab\u30c6\u30b4\u30ea\u3092\u5272\u308a\u5f53\u3066\u308b\u95a2\u6570\nget_ordinal_category &lt;- function(lp_val, thresholds) {\n    prob_cum1 &lt;- plogis(thresholds&#91;1] - lp_val) # P(Y &lt;= 1)\n    prob_cum2 &lt;- plogis(thresholds&#91;2] - lp_val) # P(Y &lt;= 2)\n    prob_cum3 &lt;- plogis(thresholds&#91;3] - lp_val) # P(Y &lt;= 3)\n\n    u &lt;- runif(1) # \u4e00\u69d8\u4e71\u6570\u3092\u751f\u6210\n\n    if (u &lt; prob_cum1) {\n        return(1) # \u5168\u304f\u6539\u5584\u306a\u3057\n    } else if (u &lt; prob_cum2) {\n        return(2) # \u5c11\u3057\u6539\u5584\n    } else if (u &lt; prob_cum3) {\n        return(3) # \u304b\u306a\u308a\u6539\u5584\n    } else {\n        return(4) # \u5b8c\u5168\u306b\u6539\u5584\n    }\n}\n\n# \u5404\u884c\u306b\u3064\u3044\u3066\u75c7\u72b6\u6539\u5584\u5ea6\u3092\u751f\u6210\nsymptom_improvement &lt;- mapply(get_ordinal_category, linear_predictor_with_re, MoreArgs = list(thresholds))\n\n# \u30c7\u30fc\u30bf\u30d5\u30ec\u30fc\u30e0\u306e\u4f5c\u6210\ndf_ordinal_glmm &lt;- data.frame(\n    patient_id = factor(patient_id),\n    treatment = factor(treatment),\n    visit = factor(visit, levels = c(\"Baseline\", \"Post_Treatment_1\", \"Post_Treatment_2\")), # \u9806\u5e8f\u3092\u8003\u616e\n    symptom_improvement = ordered(symptom_improvement,\n        levels = 1:4\n    )\n)\n\n# \u30c7\u30fc\u30bf\u306e\u78ba\u8a8d\nhead(df_ordinal_glmm)\nstr(df_ordinal_glmm)\n\n# --- clmm \u30e2\u30c7\u30eb\u306e\u69cb\u7bc9 ---\n# formula: \u5fdc\u7b54\u5909\u6570 ~ \u56fa\u5b9a\u52b9\u679c + (\u30e9\u30f3\u30c0\u30e0\u52b9\u679c | \u30b0\u30eb\u30fc\u30d4\u30f3\u30b0\u5909\u6570)\n# `symptom_improvement` \u3092\u5fdc\u7b54\u5909\u6570\u306b\u3001`treatment` \u3068 `visit` \u306e\u4ea4\u4e92\u4f5c\u7528\u3092\u56fa\u5b9a\u52b9\u679c\u306b\u6307\u5b9a\n# `(1 | patient_id)` \u306f\u3001\u5404\u88ab\u9a13\u8005\u3054\u3068\u306e\u30e9\u30f3\u30c0\u30e0\u5207\u7247\u3092\u610f\u5473\u3057\u3001\u500b\u4eba\u5185\u306e\u53cd\u5fa9\u6e2c\u5b9a\u306b\u3088\u308b\u76f8\u95a2\u3092\u8003\u616e\u3057\u307e\u3059\u3002\nmodel_clmm &lt;- clmm(symptom_improvement ~ treatment * visit + (1 | patient_id), data = df_ordinal_glmm)\n\n# \u30e2\u30c7\u30eb\u306e\u30b5\u30de\u30ea\u30fc\nsummary(model_clmm)\n\n# --- EM\u5e73\u5747\u306e\u8a08\u7b97\u3068\u7fa4\u9593\u30fb\u6642\u70b9\u9593\u6bd4\u8f03 ---\n\n# 1. \u6642\u70b9\u5225\u7fa4\u9593\u6bd4\u8f03 (Between-Group Comparisons at Each Time Point)\n# \u5404\u6642\u70b9\u306b\u304a\u3044\u3066\u3001\"DrugA\" \u3068 \"DrugB\" \u306e\u75c7\u72b6\u6539\u5584\u5ea6\u3092\u6bd4\u8f03\n\n# EM\u5e73\u5747\u306e\u7b97\u51fa (\u75bc\u75db\u30ec\u30d9\u30eb\u306e\u5404\u30ab\u30c6\u30b4\u30ea\u306e\u78ba\u7387\u3068\u3057\u3066\u8868\u793a\u3055\u308c\u307e\u3059)\n# type = \"response\" \u3067\u30ab\u30c6\u30b4\u30ea\u78ba\u7387\u3092\u51fa\u529b\nemmeans_clmm_time_group &lt;- emmeans(model_clmm, ~ treatment | visit, type = \"response\")\nprint(emmeans_clmm_time_group)\n\n# \u5404\u6642\u70b9\u3067\u306e\u7fa4\u9593\u6bd4\u8f03\n# pairs() \u3067\u6bd4\u8f03\u3092\u6307\u5b9a\u3002adjust \u3067\u591a\u91cd\u6bd4\u8f03\u8abf\u6574\u65b9\u6cd5\u3092\u6307\u5b9a\u3002\n# \u6ce8\u610f: \u9806\u5e8f\u30ed\u30b8\u30b9\u30c6\u30a3\u30c3\u30af\u56de\u5e30\u306eEMmeans\u306e\u51fa\u529b\u306f\u5404\u30ab\u30c6\u30b4\u30ea\u306e\u7d2f\u7a4d\u78ba\u7387\u307e\u305f\u306f\u500b\u3005\u306e\u78ba\u7387\u306b\u306a\u308b\u305f\u3081\u3001\n# \u6bd4\u8f03\u306f\u30aa\u30c3\u30ba\u6bd4\u3067\u306f\u306a\u304f\u3001\u7dda\u5f62\u4e88\u6e2c\u5b50\u30b9\u30b1\u30fc\u30eb\u3067\u306e\u5dee\u5206\u3001\u307e\u305f\u306f\u78ba\u7387\u30b9\u30b1\u30fc\u30eb\u3067\u306e\u5dee\u5206\u3068\u306a\u308b\u3053\u3068\u304c\u591a\u3044\u3067\u3059\u3002\n# \u3053\u3053\u3067\u306f\u78ba\u7387\u30b9\u30b1\u30fc\u30eb\u3067\u3001\u5404\u30ab\u30c6\u30b4\u30ea\u306e\u78ba\u7387\u3092\u76f4\u63a5\u6bd4\u8f03\u3057\u307e\u3059\u3002\n# \u3042\u308b\u3044\u306f\u3001\u7dda\u5f62\u4e88\u6e2c\u5b50\u30b9\u30b1\u30fc\u30eb\u3067\u6bd4\u8f03\u3057\u3001\u305d\u306e\u7d50\u679c\u304b\u3089\u30aa\u30c3\u30ba\u6bd4\u3092\u89e3\u91c8\u3059\u308b\u3053\u3068\u3082\u53ef\u80fd\u3067\u3059\u3002\n\n# \u7dda\u5f62\u4e88\u6e2c\u5b50\u30b9\u30b1\u30fc\u30eb\u3067\u306e\u7fa4\u9593\u6bd4\u8f03 (\u30aa\u30c3\u30ba\u6bd4\u306f\u89e3\u91c8\u304c\u96e3\u3057\u3044\u304c\u3001\u52b9\u679c\u306e\u65b9\u5411\u6027\u306f\u308f\u304b\u308b)\n# \u3053\u308c\u306f\u3001\u4f8b\u3048\u3070\u300c\u4e2d\u7b49\u5ea6\u4ee5\u4e0a\u306b\u306a\u308b\u7d2f\u7a4d\u30aa\u30c3\u30ba\u300d\u306a\u3069\u306e\u6bd4\u8f03\u306b\u306a\u308b\u305f\u3081\u3001\u6ce8\u610f\u304c\u5fc5\u8981\u3067\u3059\u3002\ncontrasts_clmm_time_group_linear &lt;- pairs(emmeans(model_clmm, ~ treatment | visit), reverse = TRUE)\nsummary(contrasts_clmm_time_group_linear)\n\n\n\n# 2. \u7fa4\u5225\u6642\u70b9\u9593\u6bd4\u8f03 (Within-Group Comparisons Across Time Points)\n# \u5404\u6cbb\u7642\u7fa4\u5185\u306b\u304a\u3044\u3066\u3001\u6642\u9593\u7d4c\u904e\u306b\u3088\u308b\u75bc\u75db\u75c7\u72b6\u6539\u5584\u30ec\u30d9\u30eb\u306e\u5909\u5316\u3092\u6bd4\u8f03\n\n# EM\u5e73\u5747\u306e\u7b97\u51fa\nemmeans_clmm_group_time &lt;- emmeans(model_clmm, ~ visit | treatment, type = \"response\")\nprint(emmeans_clmm_group_time)\n\n# \u5404\u7fa4\u5185\u3067\u306e\u6642\u70b9\u9593\u6bd4\u8f03\ncontrasts_clmm_group_time_linear &lt;- pairs(emmeans(model_clmm, ~ visit | treatment), reverse = TRUE, adjust = \"none\")\nsummary(contrasts_clmm_group_time_linear)\n<\/code><\/pre>\n\n\n\n<p><strong>\u5b9f\u884c\u7d50\u679c\uff1a<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>> summary(contrasts_clmm_time_group_linear)\nvisit = Baseline:\n contrast      estimate    SE  df z.ratio p.value\n DrugB - DrugA   -0.238 0.508 Inf  -0.468  0.6401\n\nvisit = Post_Treatment_1:\n contrast      estimate    SE  df z.ratio p.value\n DrugB - DrugA    0.313 0.512 Inf   0.612  0.5407\n\nvisit = Post_Treatment_2:\n contrast      estimate    SE  df z.ratio p.value\n DrugB - DrugA   -0.141 0.519 Inf  -0.272  0.7858<\/code><\/pre>\n\n\n\n<p>\u6642\u70b9\u3054\u3068\u306e Drug B \u3068 Drug A\u306e\u5dee\u3067\u3001\u3044\u305a\u308c\u306e\u6642\u70b9\u3082\u6709\u610f\u5dee\u306a\u3057\u3067\u3042\u3063\u305f\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>> summary(contrasts_clmm_group_time_linear)\ntreatment = DrugA:\n contrast                            estimate    SE  df z.ratio p.value\n Post_Treatment_1 - Baseline            0.132 0.476 Inf   0.278  0.7813\n Post_Treatment_2 - Baseline            0.350 0.486 Inf   0.719  0.4720\n Post_Treatment_2 - Post_Treatment_1    0.218 0.476 Inf   0.457  0.6476\n\ntreatment = DrugB:\n contrast                            estimate    SE  df z.ratio p.value\n Post_Treatment_1 - Baseline            0.683 0.473 Inf   1.444  0.1486\n Post_Treatment_2 - Baseline            0.446 0.471 Inf   0.948  0.3432\n Post_Treatment_2 - Post_Treatment_1   -0.236 0.483 Inf  -0.490  0.6245\n<\/code><\/pre>\n\n\n\n<p>Drug\u306e\u7fa4\u3054\u3068\u306e\u6642\u70b9\u9593\u6bd4\u8f03\u3067\u3001\u3069\u3061\u3089\u306e\u7fa4\u3082\u3069\u306e\u6642\u70b9\u9593\u3082\u6709\u610f\u5dee\u306a\u3057\u3067\u3042\u3063\u305f\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u30ab\u30a6\u30f3\u30c8\u30c7\u30fc\u30bf<\/h3>\n\n\n\n<p><strong>R \u8a08\u7b97\u4f8b\uff1a<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code># \u30c0\u30df\u30fc\u30c7\u30fc\u30bf\u306e\u4f5c\u6210\nset.seed(123)\ndata_count &lt;- expand.grid(subject = 1:50, time = c(1, 3, 6))\ndata_count &lt;- data_count&#91;order(data_count$subject), ]\ndata_count$group &lt;- rep(c(\"NewDrug\", \"Control\"), each = 75)\ndata_count$time &lt;- as.numeric(data_count$time)\n# \u591c\u9593\u899a\u9192\u56de\u6570\u3092 group \u3068 time \u306b\u5fdc\u3058\u3066\u8a2d\u5b9a (lambda)\ndata_count$lambda &lt;- ifelse(data_count$group == \"NewDrug\",\n    3 - data_count$time * 0.1,\n    5 - data_count$time * 0.05\n)\ndata_count$awakening_count &lt;- rpois(nrow(data_count), data_count$lambda)\ndata_count$subject &lt;- as.factor(data_count$subject)\ndata_count$time &lt;- as.factor(data_count$time)\n\n# \u30dd\u30a2\u30bd\u30f3\u6df7\u5408\u30e2\u30c7\u30eb\u306e\u69cb\u7bc9\nmodel_poisson &lt;- glmer(awakening_count ~ group * time + (1 | subject), data = data_count, family = poisson(link = \"log\"))\n\n# \u904e\u5206\u6563\u304c\u3042\u308b\u5834\u5408\u3001\u8ca0\u306e\u4e8c\u9805\u6df7\u5408\u30e2\u30c7\u30eb\u3092\u4f7f\u7528 (MASS::glm.nb \u3092\u62e1\u5f35\u3057\u305f glmmTMB \u306a\u3069\u304c\u5fc5\u8981\u3060\u304c\u3001\u3053\u3053\u3067\u306f glmer \u306e\u4f8b\u3068\u3057\u3066\u30dd\u30a2\u30bd\u30f3\u3092\u4f7f\u7528)\n# library(glmmTMB)\n# model_negbin &lt;- glmmTMB(awakening_count ~ group * time + (1 | subject), data = data_count, family = nbinom2(link = \"log\"))\n\n\n# \u30e2\u30c7\u30eb\u306e\u30b5\u30de\u30ea\u30fc\nsummary(model_poisson)\n\n# --- EM\u5e73\u5747\u306e\u8a08\u7b97\u3068\u7fa4\u9593\u30fb\u6642\u70b9\u9593\u6bd4\u8f03 ---\n\n#### 1. \u6642\u70b9\u5225\u7fa4\u9593\u6bd4\u8f03 (Between-Group Comparisons at Each Time Point)\n# \u5404\u6e2c\u5b9a\u6642\u70b9\u306b\u304a\u3044\u3066\u3001\"NewDrug\" \u3068 \"Control\" \u306e\u5e73\u5747\u899a\u9192\u56de\u6570\u3092\u6bd4\u8f03\n\n# EM\u5e73\u5747\u306e\u7b97\u51fa (type = \"response\" \u3067\u5143\u306e\u30ab\u30a6\u30f3\u30c8\u30b9\u30b1\u30fc\u30eb\u306b\u5909\u63db)\n# \u30c7\u30d5\u30a9\u30eb\u30c8\u3067\u306f\u7dda\u5f62\u4e88\u6e2c\u5b50\u30b9\u30b1\u30fc\u30eb (log-odds) \u3067\u8a08\u7b97\u3055\u308c\u308b\u305f\u3081\u3001\n# \u899a\u9192\u56de\u6570\u305d\u306e\u3082\u306e\u306e\u6bd4\u8f03\u306b\u306f type=\"response\" \u3092\u6307\u5b9a\u3057\u307e\u3059\u3002\nemmeans_poisson_time_group &lt;- emmeans(model_poisson, ~ group | time, type = \"response\")\nprint(emmeans_poisson_time_group)\n\n# \u5404\u6642\u70b9\u3067\u306e\u7fa4\u9593\u6bd4\u8f03\n# pairs() \u3067\u30da\u30a2\u30ef\u30a4\u30ba\u6bd4\u8f03\u3092\u6307\u5b9a\u3002\n# \u30dd\u30a2\u30bd\u30f3\u30e2\u30c7\u30eb\u306a\u306e\u3067\u3001\u6bd4\u8f03\u7d50\u679c\u306f\u300c\u6bd4\u7387\u300d\u307e\u305f\u306f\u300c\u6bd4\u7387\u306e\u5bfe\u6570\uff08\u30ed\u30b0\u30aa\u30c3\u30ba\uff09\u300d\u3068\u3057\u3066\u89e3\u91c8\u3055\u308c\u307e\u3059\u3002\n# type = \"response\" \u3067\u6bd4\u7387\u3068\u3057\u3066\u7d50\u679c\u304c\u51fa\u529b\u3055\u308c\u307e\u3059 (Risk Ratio \u307e\u305f\u306f Rate Ratio)\u3002\ncontrasts_poisson_time_group &lt;- pairs(emmeans_poisson_time_group, reverse = TRUE)\nsummary(contrasts_poisson_time_group, infer = TRUE) # infer=TRUE\u3067\u63a8\u5b9a\u5024\u3068\u4fe1\u983c\u533a\u9593\u3082\u8868\u793a\n\n\n\n#### 2. \u7fa4\u5225\u6642\u70b9\u9593\u6bd4\u8f03 (Within-Group Comparisons Across Time Points)\n# \u5404\u6cbb\u7642\u7fa4\u5185\u306b\u304a\u3044\u3066\u3001\u6642\u9593\u7d4c\u904e\u306b\u3088\u308b\u5e73\u5747\u899a\u9192\u56de\u6570\u306e\u5909\u5316\u3092\u6bd4\u8f03\n\n# EM\u5e73\u5747\u306e\u7b97\u51fa\nemmeans_poisson_group_time &lt;- emmeans(model_poisson, ~ time | group, type = \"response\")\nprint(emmeans_poisson_group_time)\n\n# \u5404\u7fa4\u5185\u3067\u306e\u6642\u70b9\u9593\u6bd4\u8f03\ncontrasts_poisson_group_time &lt;- pairs(emmeans_poisson_group_time, adjust = \"none\", reverse = TRUE)\nsummary(contrasts_poisson_group_time, infer = TRUE)\n<\/code><\/pre>\n\n\n\n<p><strong>\u5b9f\u884c\u7d50\u679c\uff1a<\/strong><\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>> summary(contrasts_poisson_time_group, infer = TRUE) # infer=TRUE\u3067\u63a8\u5b9a\u5024\u3068\u4fe1\u983c\u533a\u9593\u3082\u8868\u793a\ntime = 1:\n contrast          ratio     SE  df asymp.LCL asymp.UCL null z.ratio p.value\n NewDrug \/ Control 0.594 0.0844 Inf     0.450     0.785    1  -3.667  0.0002\n\ntime = 3:\n contrast          ratio     SE  df asymp.LCL asymp.UCL null z.ratio p.value\n NewDrug \/ Control 0.694 0.1040 Inf     0.517     0.932    1  -2.426  0.0153\n\ntime = 6:\n contrast          ratio     SE  df asymp.LCL asymp.UCL null z.ratio p.value\n NewDrug \/ Control 0.400 0.0669 Inf     0.288     0.555    1  -5.476  &lt;.0001\n\nConfidence level used: 0.95 \nIntervals are back-transformed from the log scale \nTests are performed on the log scale \n> <\/code><\/pre>\n\n\n\n<p>\u307e\u305a\u3001\u6642\u70b9\u5225\u7fa4\u9593\u6bd4\u8f03\u306e\u7d50\u679c\u3092\u793a\u3059\u3002<\/p>\n\n\n\n<p>time = 1, 3, 6 \u306e\u3044\u305a\u308c\u306b\u304a\u3044\u3066\u3082\u3001NewDrug \u306e\u307b\u3046\u304c Control \u306b\u6bd4\u3079\u3066\u4f4e\u3044\uff08ratio \u304c 1 \u672a\u6e80\u3067\u3042\u308b\uff09\u3053\u3068\u304c\u308f\u304b\u308a\u3001\u6709\u610f\u5dee\u3042\u308a\u3067\u3042\u3063\u305f\u3002<\/p>\n\n\n\n<pre class=\"wp-block-code\"><code>> summary(contrasts_poisson_group_time, infer = TRUE)\ngroup = Control:\n contrast      ratio    SE  df asymp.LCL asymp.UCL null z.ratio p.value\n time3 \/ time1 0.812 0.105 Inf     0.630     1.047    1  -1.607  0.1079\n time6 \/ time1 0.940 0.117 Inf     0.736     1.200    1  -0.498  0.6185\n time6 \/ time3 1.157 0.152 Inf     0.895     1.497    1   1.113  0.2658\n\ngroup = NewDrug:\n contrast      ratio    SE  df asymp.LCL asymp.UCL null z.ratio p.value\n time3 \/ time1 0.949 0.153 Inf     0.692     1.302    1  -0.322  0.7472\n time6 \/ time1 0.633 0.114 Inf     0.444     0.902    1  -2.531  0.0114\n time6 \/ time3 0.667 0.122 Inf     0.466     0.953    1  -2.221  0.0264\n\nConfidence level used: 0.95 \nIntervals are back-transformed from the log scale \nTests are performed on the log scale \n> <\/code><\/pre>\n\n\n\n<p>\u7fa4\u5225\u306e\u6642\u70b9\u6bd4\u8f03\u3067\u306f\u3001Control \u7fa4\u3067\u306f\u6642\u70b9\u9593\u306b\u9055\u3044\u304c\u306a\u3044\u306e\u306b\u5bfe\u3057\u3001NewDrug \u7fa4\u3067\u306f\u3001time 6 vs. time 3 \u53ca\u3073 time 6 vs. time 1 \u306b\u304a\u3044\u3066\u3001\u6709\u610f\u5dee\u3042\u308a\u3068\u3044\u3046\u7d50\u679c\u3067\u3042\u3063\u305f\u3002<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u53cd\u5fa9\u6e2c\u5b9a\u6642\u70b9\u9593\u6bd4\u8f03\u306b\u304a\u3051\u308b\u691c\u5b9a\u306e\u591a\u91cd\u6027\u306b\u3064\u3044\u3066<\/h2>\n\n\n\n<p>\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u306b\u304a\u3044\u3066\u3001\u8907\u6570\u306e\u6642\u70b9\u9593\u3067\u6bd4\u8f03\u3092\u884c\u3046\u5834\u5408\uff08\u4f8b\uff1a\u57fa\u6e96\u6642\u70b9\u30681\u30f6\u6708\u5f8c\u3001\u57fa\u6e96\u6642\u70b9\u30683\u30f6\u6708\u5f8c\u30011\u30f6\u6708\u5f8c\u30683\u30f6\u6708\u5f8c\u306a\u3069\uff09\u3001<strong>\u591a\u91cd\u6bd4\u8f03\u306e\u554f\u984c<\/strong>\u304c\u751f\u3058\u308b\u3002\u591a\u91cd\u6bd4\u8f03\u3068\u306f\u3001\u591a\u6570\u306e\u7d71\u8a08\u7684\u691c\u5b9a\u3092\u540c\u6642\u306b\u884c\u3046\u3053\u3068\u3067\u3001\u672c\u6765\u306f\u6709\u610f\u3067\u306f\u306a\u3044\u5dee\u3092\u8aa4\u3063\u3066\u300c\u6709\u610f\u3067\u3042\u308b\u300d\u3068\u5224\u65ad\u3057\u3066\u3057\u307e\u3046\uff08<strong>\u7b2c\u4e00\u7a2e\u306e\u904e\u8aa4<\/strong>\uff09\u78ba\u7387\u304c\u5897\u52a0\u3059\u308b\u73fe\u8c61\u3067\u3042\u308b\u3002<\/p>\n\n\n\n<p>\u4f8b\u3048\u3070\u3001\u6709\u610f\u6c34\u6e96 \u03b1=0.05 \u30671\u56de\u306e\u691c\u5b9a\u3092\u884c\u3046\u5834\u5408\u3001\u8aa4\u3063\u3066\u7b2c\u4e00\u7a2e\u306e\u904e\u8aa4\u3092\u72af\u3059\u78ba\u7387\u306f5%\u3067\u3042\u308b\u3002\u3057\u304b\u3057\u3001\u3053\u308c\u304c10\u56de\u306e\u691c\u5b9a\u306b\u306a\u308b\u3068\u3001\u5c11\u306a\u304f\u3068\u30821\u56de\u7b2c\u4e00\u7a2e\u306e\u904e\u8aa4\u3092\u72af\u3059\u78ba\u7387\u306f 1\u2212(1\u22120.05)10\u22480.40 (\u7d0440%) \u306b\u307e\u3067\u4e0a\u6607\u3057\u3066\u3057\u307e\u3046\u3002<\/p>\n\n\n\n<p>\u3057\u305f\u304c\u3063\u3066\u3001\u53cd\u5fa9\u6e2c\u5b9a\u306b\u304a\u3051\u308b\u6642\u70b9\u9593\u6bd4\u8f03\u3067\u306f\u3001\u691c\u5b9a\u306e\u591a\u91cd\u6027\u306b\u3088\u3063\u3066\u3001\u3053\u306e\u7b2c\u4e00\u7a2e\u306e\u904e\u8aa4\u306e\u78ba\u7387\u306e\u30a4\u30f3\u30d5\u30ec\u30fc\u30b7\u30e7\u30f3\u304c\u8d77\u304d\u308b\u3053\u3068\u3092\u8a8d\u8b58\u3059\u308b\u5fc5\u8981\u304c\u3042\u308b\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u591a\u91cd\u6bd4\u8f03\u8abf\u6574\u306e\u65b9\u6cd5<\/h3>\n\n\n\n<p>\u591a\u91cd\u6bd4\u8f03\u8abf\u6574\u306e\u76ee\u7684\u306f\u3001\u30d5\u30a1\u30df\u30ea\u30fc\u30ef\u30a4\u30ba\u30a8\u30e9\u30fc\u7387 (Family-Wise Error Rate: FWER) \u307e\u305f\u306f\u507d\u767a\u898b\u7387 (False Discovery Rate: FDR) \u3092\u5236\u5fa1\u3059\u308b\u3053\u3068\u306b\u3042\u308b\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>FWER\u5236\u5fa1:<\/strong> \u5c11\u306a\u304f\u3068\u30821\u3064\u306e\u7b2c\u4e00\u7a2e\u306e\u904e\u8aa4\u3092\u72af\u3059\u78ba\u7387\u3092 \u03b1 \u4ee5\u4e0b\u306b\u6291\u3048\u308b\u3053\u3068\u3092\u76ee\u6307\u3059\u3002\u3088\u308a\u53b3\u5bc6\u306a\u65b9\u6cd5\u3067\u3042\u308a\u3001\u30dc\u30f3\u30d5\u30a7\u30ed\u30fc\u30cb\u88dc\u6b63 (Bonferroni correction) \u3084\u30db\u30eb\u30e0\u88dc\u6b63 (Holm correction) \u306a\u3069\u304c\u3042\u308b\u3002\n<ul class=\"wp-block-list\">\n<li><strong>\u30dc\u30f3\u30d5\u30a7\u30ed\u30fc\u30cb\u88dc\u6b63:<\/strong> \u5404\u691c\u5b9a\u306e\u6709\u610f\u6c34\u6e96\u3092 k \uff08\u6bd4\u8f03\u306e\u7dcf\u6570\uff09\u3067\u5272\u308b\uff08\u03b1adjusted\u200b=\u03b1\/k\uff09\u3002\u6700\u3082\u4fdd\u5b88\u7684\u306a\u65b9\u6cd5\u3067\u3042\u308a\u3001\u691c\u51fa\u529b\u304c\u4f4e\u4e0b\u3057\u3084\u3059\u3044\u3068\u3044\u3046\u6b20\u70b9\u304c\u3042\u308b\u3002<\/li>\n\n\n\n<li><strong>\u30db\u30eb\u30e0\u88dc\u6b63:<\/strong> \u30dc\u30f3\u30d5\u30a7\u30ed\u30fc\u30cb\u88dc\u6b63\u3088\u308a\u3082\u691c\u51fa\u529b\u304c\u9ad8\u304f\u3001\u5e83\u304f\u63a8\u5968\u3055\u308c\u3066\u3044\u308b\u3002P\u5024\u3092\u5c0f\u3055\u3044\u9806\u306b\u4e26\u3079\u3001\u9806\u6b21\u8abf\u6574\u3092\u884c\u3046\u3002<\/li>\n<\/ul>\n<\/li>\n\n\n\n<li><strong>FDR\u5236\u5fa1:<\/strong> \u8aa4\u3063\u3066\u68c4\u5374\u3055\u308c\u305f\u5e30\u7121\u4eee\u8aac\u306e\u3046\u3061\u3001\u5b9f\u969b\u306b\u6b63\u3057\u3044\u5e30\u7121\u4eee\u8aac\u306e\u5272\u5408\u3092\u5236\u5fa1\u3059\u308b\u3002\u30d9\u30f3\u30b8\u30e3\u30df\u30cb\u30fb\u30db\u30c3\u30af\u30d0\u30fc\u30b0\u6cd5 (Benjamini-Hochberg procedure) \u306a\u3069\u304c\u3042\u308b\u3002\u63a2\u7d22\u7684\u306a\u7814\u7a76\u3084\u3001\u591a\u304f\u306e\u6bd4\u8f03\u3092\u884c\u3046\u5834\u5408\u306b\u9069\u3057\u3066\u304a\u308a\u3001FWER\u5236\u5fa1\u3088\u308a\u3082\u691c\u51fa\u529b\u304c\u9ad8\u304f\u306a\u308b\u50be\u5411\u304c\u3042\u308b\u3002<\/li>\n<\/ul>\n\n\n\n<p>\u3069\u306e\u8abf\u6574\u65b9\u6cd5\u3092\u9078\u629e\u3059\u308b\u304b\u306f\u3001\u7814\u7a76\u306e\u76ee\u7684\u3001\u691c\u5b9a\u306e\u6570\u3001\u305d\u3057\u3066\u7b2c\u4e00\u7a2e\u306e\u904e\u8aa4\u3068\u7b2c\u4e8c\u7a2e\u306e\u904e\u8aa4\u306e\u3069\u3061\u3089\u3092\u3088\u308a\u91cd\u8996\u3059\u308b\u304b\u306b\u3088\u3063\u3066\u6c7a\u5b9a\u3055\u308c\u308b\u3002\u81e8\u5e8a\u7814\u7a76\u3067\u306f\u3001\u614e\u91cd\u306a\u7d50\u8ad6\u304c\u6c42\u3081\u3089\u308c\u308b\u305f\u3081\u3001\u30dc\u30f3\u30d5\u30a7\u30ed\u30fc\u30cb\u88dc\u6b63\u3084\u30db\u30eb\u30e0\u88dc\u6b63\u304c\u3088\u304f\u7528\u3044\u3089\u308c\u308b\u3002<code>emmeans<\/code> \u30d1\u30c3\u30b1\u30fc\u30b8\u3067\u306f\u3001<code>adjust<\/code> \u5f15\u6570\u3067\u69d8\u3005\u306a\u8abf\u6574\u65b9\u6cd5\u3092\u6307\u5b9a\u3067\u304d\u308b\uff08\u4f8b: <code>\"bonferroni\"<\/code>, <code>\"holm\"<\/code>, <code>\"fdr\"<\/code>, <code>\"tukey\"<\/code> \u306a\u3069\uff09\u3002<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">\u591a\u91cd\u8abf\u6574\u306b\u95a2\u3059\u308b\u6ce8\u610f\u70b9<\/h3>\n\n\n\n<p>\u591a\u91cd\u8abf\u6574\u306f\u3001\u7814\u7a76\u306e\u4fe1\u983c\u6027\u3092\u9ad8\u3081\u308b\u4e0a\u3067\u91cd\u8981\u306a\u7d71\u8a08\u7684\u624b\u6cd5\u3067\u3042\u308b\u3002\u3057\u304b\u3057\u3001\u305d\u306e\u5b9f\u65bd\u306b\u306f\u9069\u5207\u306a\u30bf\u30a4\u30df\u30f3\u30b0\u304c\u3042\u308b\u3002<\/p>\n\n\n\n<p>\u591a\u91cd\u8abf\u6574\u306f\u3001\u7814\u7a76\u30c7\u30b6\u30a4\u30f3\u306e\u6bb5\u968e\u3001\u7279\u306b<strong>\u30b5\u30f3\u30d7\u30eb\u30b5\u30a4\u30ba\u3092\u8a08\u7b97\u3059\u308b\u969b\u306b\u8003\u616e\u3059\u3079\u304d\u4e8b\u9805<\/strong>\u3067\u3042\u308b\u3002\u591a\u304f\u306e\u81e8\u5e8a\u7814\u7a76\u3067\u306f\u3001\u5b9f\u969b\u306e\u8a3a\u7642\u3067\u5f97\u3089\u308c\u305f\u75c7\u4f8b\u306b\u57fa\u3065\u3044\u3066\u63a2\u7d22\u7684\u306b\u89e3\u6790\u304c\u884c\u308f\u308c\u308b\u304c\u3001\u305d\u306e\u969b\u3001\u30b5\u30f3\u30d7\u30eb\u30b5\u30a4\u30ba\u304c\u5341\u5206\u3067\u306a\u304f\u3001\u7d71\u8a08\u7684\u306a\u691c\u51fa\u529b\u304c\u4e0d\u8db3\u3057\u3066\u3044\u308b\u5834\u5408\u304c\u5c11\u306a\u304f\u306a\u3044\u3002<\/p>\n\n\n\n<p>\u4e8b\u524d\u306b\u8abf\u6574\u65b9\u6cd5\u3092\u898f\u5b9a\u305b\u305a\u306b\u4e8b\u5f8c\u7684\u306b\u591a\u91cd\u6bd4\u8f03\u8abf\u6574\u3092\u53b3\u5bc6\u306b\u884c\u3063\u3066\u3082\u3001\u305d\u306e\u6709\u52b9\u6027\u306f\u9650\u5b9a\u7684\u3060\u3002<\/p>\n\n\n\n<p>\u3057\u305f\u304c\u3063\u3066\u3001\u89b3\u5bdf\u7814\u7a76\u3067\u63a2\u7d22\u7684\u306a\u89e3\u6790\u3092\u884c\u3046\u5834\u5408\u306a\u3069\u3001\u591a\u91cd\u8abf\u6574\u3092\u5b9f\u65bd\u3057\u306a\u3044\u9078\u629e\u3092\u3059\u308b\u969b\u306b\u306f\u3001\u300c<strong>\u89b3\u5bdf\u578b\u7814\u7a76\u3067\u3042\u308a\u3001\u63a2\u7d22\u7684\u306a\u89e3\u6790\u3067\u3042\u308b\u305f\u3081\u3001\u591a\u91cd\u6bd4\u8f03\u8abf\u6574\u306f\u884c\u308f\u306a\u304b\u3063\u305f<\/strong>\u300d\u3068\u660e\u78ba\u306b\u8a18\u8f09\u3059\u308b\u3053\u3068\u3092\u304a\u3059\u3059\u3081\u3059\u308b\u3002\u3053\u308c\u306f\u3001\u7814\u7a76\u306e\u900f\u660e\u6027\u3092\u4fdd\u3061\u3001\u7d50\u679c\u306e\u89e3\u91c8\u3092\u3088\u308a\u660e\u78ba\u306b\u3059\u308b\u4e0a\u3067\u5f79\u7acb\u3064\u3002<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u307e\u3068\u3081<\/h2>\n\n\n\n<p>\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb\u3084\u4e00\u822c\u5316\u7dda\u5f62\u6df7\u5408\u30e2\u30c7\u30eb\u306f\u3001\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u306e\u8907\u96d1\u306a\u69cb\u9020\u3092\u9069\u5207\u306b\u8003\u616e\u3057\u3001\u7fa4\u9593\u304a\u3088\u3073\u6642\u70b9\u9593\u3067\u306e\u52b9\u679c\u3092\u8a55\u4fa1\u3059\u308b\u305f\u3081\u306e\u5f37\u529b\u306a\u30c4\u30fc\u30eb\u3060\u3002EM\u5e73\u5747\u306f\u3001\u3053\u308c\u3089\u306e\u30e2\u30c7\u30eb\u304b\u3089\u5171\u5909\u91cf\u3092\u8abf\u6574\u3057\u305f\u4e0a\u3067\u306e\u7fa4\u3084\u6642\u70b9\u306e\u5e73\u5747\u5024\u3092\u63a8\u5b9a\u3057\u3001\u6bd4\u8f03\u3092\u884c\u3046\u306e\u306b\u975e\u5e38\u306b\u6709\u7528\u3067\u3042\u308b\u3002<code>emmeans<\/code> \u30d1\u30c3\u30b1\u30fc\u30b8\u3092\u6d3b\u7528\u3059\u308b\u3053\u3068\u3067\u3001\u3053\u308c\u3089\u306e\u8907\u96d1\u306a\u5206\u6790\u3092R\u4e0a\u3067\u7c21\u6f54\u304b\u3064\u6b63\u78ba\u306b\u5b9f\u884c\u3059\u308b\u3053\u3068\u304c\u3067\u304d\u308b\u3002\u672c\u7a3f\u3067\u8ff0\u3079\u305f\u8003\u3048\u65b9\u3068R\u30b3\u30fc\u30c9\u4f8b\u304c\u3001\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30fc\u30bf\u306e\u89e3\u6790\u306b\u304a\u3051\u308b\u7406\u89e3\u3068\u5b9f\u8df5\u306e\u4e00\u52a9\u3068\u306a\u308c\u3070\u5e78\u3044\u3067\u3042\u308b\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u81e8\u5e8a\u7814\u7a76\u3067\u306f\u3001\u540c\u4e00\u306e\u5bfe\u8c61\u8005\u306b\u5bfe\u3057\u3066\u8907\u6570\u56de\u6e2c\u5b9a\u3092\u884c\u3046\u53cd\u5fa9\u6e2c\u5b9a\u30c7\u30b6\u30a4\u30f3\u304c\u983b\u7e41\u306b\u7528\u3044\u3089\u308c\u308b\u3002\u3053\u306e\u3088\u3046\u306a\u30c7\u30fc\u30bf\u306f\u3001\u6642\u9593\u7d4c\u904e\u306b\u4f34\u3046\u5909\u5316\u3084\u4ecb\u5165\u52b9\u679c\u3092\u8a55\u4fa1\u3059\u308b\u4e0a\u3067\u975e\u5e38\u306b\u6709\u7528\u3060\u304c\u3001\u4e00\u65b9\u3067\u3001\u8907\u96d1\u306a\u76f8\u95a2\u69cb\u9020\u3092\u30e2\u30c7\u30eb\u5316\u3059\u308b\u3068\u3044\u3046\u8ab2\u984c\u304c\u3042\u308b\u3002\u3053 [&hellip;]<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"swell_btn_cv_data":"","_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[12,77,148,59],"tags":[],"class_list":["post-4029","post","type-post","status-publish","format-standard","hentry","category-em","category-77","category-148","category-59"],"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts\/4029","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/comments?post=4029"}],"version-history":[{"count":9,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts\/4029\/revisions"}],"predecessor-version":[{"id":4110,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/posts\/4029\/revisions\/4110"}],"wp:attachment":[{"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/media?parent=4029"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/categories?post=4029"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/best-biostatistics.com\/toukei-er\/wp-json\/wp\/v2\/tags?post=4029"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}