5ÔÂ12ÈÕ14ÈÕ£¬ÓÉÖйúÏÖ³¡Í³¼ÆÑо¿»áÒò¹ûÍƶϷֻáÖ÷Àí£¬×ðÁú¿Ê±³Ð°ìµÄ¡°µÚÒ»´ÎÒò¹ûÍƶÏѧÊõÄê»á¡±ÔÚ¸·³É·УÇø¾ÙÐС£Ñ§Ð£µ³Î¯³£Î¯¡¢¸±Ð£³¤×óÃô³öϯ²¢Ö´ǡ£½ÌÎñËæ´¦³¤²ÜÏÔ±ø¡¢ÊýѧÓëͳ¼ÆѧԺµ³Î¯Êé¼ÇÖÜÑ޽ܡ¢Ôº³¤ÁõÑÞ骺ÍÀ´×Ô±±¾©´óѧ¡¢Ç廪´óѧ¡¢¸´µ©´óѧ¡¢ÖйúÈËÃñ´óѧ¡¢ÄÏ¿ª´óѧ¡¢Ïã¸ÛÖÐÎÄ´óѧµÈ40ÓàËù¸ßУµÄר¼ÒѧÕßÒÔ¼°Òò¹ûÍƶϷֻáÀíʳÉÔ±¡¢»áÔ±ÏÈÉúͬÑâÃǹ²¼Æ170ÓàÈ˼ÓÈë±¾´Î¾Û»á£¬±¾´ÎѧÊõÄê»áÓÉÊýѧÓëͳ¼ÆѧԺÍõѧÀö½ÌÊÚÖ÷³Ö¡£
×óÃô´ú±í×ðÁú¿Ê±ÏòÝ°ÁÙ×ðÁú¿Ê±µÄÁÐλר¼ÒѧÕßÌåÏÖ×îÈÈÁҵĽӴý¡£ËûÖ¸³ö£¬Ñ§Ð£ÏòÀ´ÖØÊÓͳ¼Æѧ¿ÆµÄÉú³¤£¬Ã÷È·ÁËͳ¼Æѧ¿ÆÔÚ¡°Ë«Ò»Á÷¡±½¨ÉèÖÐÖ÷ÒªµÄ»ù´¡Ö§³Öְ룬Íƶ¯½¨Éè¾ßÓб±¹¤ÉÌÌØÉ«µÄͳ¼Æѧ¿Æ¡£×óÃôÖ¸³ö£¬Òò¹ûÍƶϷֻὫÒÀÍÐ×ðÁú¿Ê±ÊýѧÓëͳ¼ÆѧԺ£¬¼¯¾ÛÇàÄêѧÕß¹¥¼á¿ËÄÑ£¬¿ªÕ¹Ïà¹ØѧÊõÔ˶¯½»Á÷Óë×êÑУ¬Îª½¨³É¹ú¼ÊÒ»Á÷ͳ¼Æѧ¿ÆÆ𾢶·Õù£¬×£Ô¸·Ö»áÔ½°ìÔ½ºÃ¡£
ÖйúÏÖ³¡Í³¼ÆÑо¿»á¸±Àíʳ¤Öйú¿ÆѧԺʯ¼á½ÌÊÚÖ´ǣ¬Ëû´ú±í×Ü»á¶ÔÒò¹ûÍƶÏѧÊõ·Ö»áµÄ½¨ÉèÌåÏÖÈÈÁÒµÄ×£ºØ£¬Ö¸³ö¾ÙÐÐÕâ´ÎÒò¹ûÍƶÏѧÊõÄê»áÊÇÆÈÇеġ¢ÐëÒªµÄ¡£Òò¹ûÍƶϷֻáÀíʳ¤×ðÁú¿Ê±¹¢Ö±½ÌÊÚÖ´ǣ¬Ëû´ú±íÒò¹ûÍƶϷֻá¶Ô³öϯ´Ë´Î¾Û»áµÄÏòµ¼¡¢¼Î±öºÍѧÊõÙÉÐÐÌåÏÖÈÈÁҵĽӴý£¬ËûÖ¸³öËæ×ÅÈ˹¤ÖÇÄܵÄÉú³¤£¬Òò¹ûÍƶϵÄÑо¿ÏÔµÃÓÈΪÆÈÇУ¬´Ë´Î¾Û»áÓÐÖúÓÚÒò¹ûÍƶÏÔÚº£ÄڵĽøÒ»²½Éú³¤¡£
³öϯ´ó»áµÄÌØÑû¼Î±öÓÐÏã¸ÛÖÐÎÄ´óѧ¡¢ÉîÛÚ´óÊý¾ÝÑо¿Ôº°¬´ºÈÙ½ÌÊÚ¡¢¹ã¶«¹¤Òµ´óѧ²ÌÈð³õ½ÌÊÚ¡¢ºþ±±Ê¦·¶´óѧ¸±Ð£³¤À²Ó½ÌÊÚ¡¢Ê׶¼Ê¦·¶´óѧ×Þ¹ú»ª½ÌÊÚ¡¢¸´µ©´óѧÂÞÇ¿½ÌÊÚÒÔ¼°±±¾©´óѧ¹«¹²ÎÀÉúѧԺÖÜÏþ»ª½ÌÊڵȡ£
ÔÚ´ó»áÖ÷»á³¡ÖУ¬±±¾©´óѧ¹«¹²ÎÀÉúѧԺÖÜÏþ»ª½ÌÊÚ×÷Ö÷Ìⱨ¸æ¡¶Inference for treatment effects on many derived random variables¡·¡¢¸´µ©´óѧÂÞÇ¿½ÌÊÚ×÷Ö÷Ìⱨ¸æ¡¶ÄÔÖØ´óϵͳµÄÒò¹û½¨Ä£ÀíÂÛÓëµ÷¿ØÒªÁì¡·¡¢¹ã¶«¹¤Òµ´óѧ²ÌÈð³õ½ÌÊÚ×÷Ö÷Ìⱨ¸æ¡¶Òò¹û¹Øϵ·¢Ã÷Èô¸ÉÎÊÌâ̽ÌÖ¡·£¬ÏÖ³¡×¨¼ÒѧÕß×êÑÐÆø·ÕÈÈÁÒ¡£
±¾´Î¾Û»áÉèÓÐÈý¸ö·ÖÂÛ̳£¬¹²°Ë¸öÖ÷Ìⱨ¸æ£¬Î§ÈÆ¡°Òò¹ûÍƶϡ±¹²¾ÙÐÐÁË24¸ö¾«²ÊµÄѧÊõ±¨¸æ¡£¾Û»áÔ¼ÇëÁ˺þ±±Ê¦·¶´óѧÀ²Ó½ÌÊںͱ±¾©´óѧ¼Ö½ðÖùÑо¿Ô±µ£µ±·Ö»á³¡Ò»µÄÖ÷³Ö£¬·Ö»á³¡¶þÓÉÖйúÈËÃñ´óѧͳ¼ÆѧԺ¸±Ôº³¤Òü½¨öκͱ±¾©´óѧºÎÑ󲨽ÌÊÚÖ÷³Ö£¬·Ö»á³¡ÈýÓÉɽ¶«´óѧÀîºé¿½ÌÊÚºÍÉϺ£½»Í¨´óѧÁõÁÖÖúÀí½ÌÊÚÖ÷³Ö¡£
ÖйúÏÖ³¡Í³¼ÆÑо¿»áÒò¹ûÍƶϷֻáµÚÒ»´ÎѧÊõÄê»áµÄÕÙ¿ª£¬ÊÇͳ¼Æѧѧ¿ÆÉú³¤µÄÐèÒª£¬ÔÚ¾Û»áÖУ¬Ìýµ½ÁÐλר¼ÒµÄ¾«²Ê·ÖÏí£¬ÅäºÏ̽ÌÖÒò¹ûÍƶÏδÀ´µÄÉú³¤Ç÷ÊÆ£¬ÐÅÍÐ×êÑн«¶ÔÌìÏ¿Ƽ¼ºÍÉç»áÇ°½ø±¬·¢Æð¾¢µÄÓ°Ï졣δÀ´£¬Òò¹ûÍƶϷֻὫÒÔ×ðÁú¿Ê±ÎªÒÀÍе¥Î»£¬¼¯ÖÐÌìÏÂÓÅÒìºÍÇàÄêѧÕß¹¥¼á¿ËÄÑ£¬Ôö½øÏà¹ØѧÊõÔ˶¯½»Á÷Ó뿪չ£¬Îª½¨³É¹ú¼ÊÒ»Á÷ͳ¼Æѧ¿Æ¶ø²»Ð¸Æ𾢶·Õù£¡