id sid tid token lemma pos fb494744g4q 1 1 randomized randomized ADJ fb494744g4q 1 2 longitudinal longitudinal ADJ fb494744g4q 1 3 designs design NOUN fb494744g4q 1 4 are be AUX fb494744g4q 1 5 commonly commonly ADV fb494744g4q 1 6 used use VERB fb494744g4q 1 7 in in ADP fb494744g4q 1 8 medical medical ADJ fb494744g4q 1 9 and and CCONJ fb494744g4q 1 10 psychological psychological ADJ fb494744g4q 1 11 studies study NOUN fb494744g4q 1 12 to to PART fb494744g4q 1 13 investigate investigate VERB fb494744g4q 1 14 the the DET fb494744g4q 1 15 treatment treatment NOUN fb494744g4q 1 16 effect effect NOUN fb494744g4q 1 17 of of ADP fb494744g4q 1 18 an an DET fb494744g4q 1 19 intervention intervention NOUN fb494744g4q 1 20 method method NOUN fb494744g4q 1 21 or or CCONJ fb494744g4q 1 22 an an DET fb494744g4q 1 23 experimental experimental ADJ fb494744g4q 1 24 drug drug NOUN fb494744g4q 1 25 . . PUNCT fb494744g4q 2 1 traditional traditional ADJ fb494744g4q 2 2 linear linear ADJ fb494744g4q 2 3 mixed mixed ADJ fb494744g4q 2 4 - - PUNCT fb494744g4q 2 5 effects effect NOUN fb494744g4q 2 6 models model NOUN fb494744g4q 2 7 for for ADP fb494744g4q 2 8 randomized randomized ADJ fb494744g4q 2 9 longitudinal longitudinal ADJ fb494744g4q 2 10 designs design NOUN fb494744g4q 2 11 are be AUX fb494744g4q 2 12 limited limit VERB fb494744g4q 2 13 to to ADP fb494744g4q 2 14 maximum maximum ADJ fb494744g4q 2 15 likelihood likelihood NOUN fb494744g4q 2 16 ( ( PUNCT fb494744g4q 2 17 ml ml PROPN fb494744g4q 2 18 ) ) PUNCT fb494744g4q 2 19 methods method NOUN fb494744g4q 2 20 which which PRON fb494744g4q 2 21 assume assume VERB fb494744g4q 2 22 data datum NOUN fb494744g4q 2 23 are be AUX fb494744g4q 2 24 missing miss VERB fb494744g4q 2 25 at at ADP fb494744g4q 2 26 random random ADJ fb494744g4q 2 27 ( ( PUNCT fb494744g4q 2 28 mar mar PROPN fb494744g4q 2 29 ) ) PUNCT fb494744g4q 2 30 . . PUNCT fb494744g4q 3 1 in in ADP fb494744g4q 3 2 practice practice NOUN fb494744g4q 3 3 , , PUNCT fb494744g4q 3 4 because because SCONJ fb494744g4q 3 5 longitudinal longitudinal ADJ fb494744g4q 3 6 data datum NOUN fb494744g4q 3 7 are be AUX fb494744g4q 3 8 likely likely ADJ fb494744g4q 3 9 to to PART fb494744g4q 3 10 be be AUX fb494744g4q 3 11 missing missing ADJ fb494744g4q 3 12 not not PART fb494744g4q 3 13 at at ADP fb494744g4q 3 14 random random ADJ fb494744g4q 3 15 ( ( PUNCT fb494744g4q 3 16 mnar mnar PROPN fb494744g4q 3 17 ) ) PUNCT fb494744g4q 3 18 , , PUNCT fb494744g4q 3 19 the the DET fb494744g4q 3 20 traditional traditional ADJ fb494744g4q 3 21 ml ml PROPN fb494744g4q 3 22 method method NOUN fb494744g4q 3 23 might might AUX fb494744g4q 3 24 lead lead VERB fb494744g4q 3 25 to to ADP fb494744g4q 3 26 severely severely ADV fb494744g4q 3 27 biased biased ADJ fb494744g4q 3 28 estimation estimation NOUN fb494744g4q 3 29 results result NOUN fb494744g4q 3 30 . . PUNCT fb494744g4q 4 1 in in ADP fb494744g4q 4 2 such such ADJ fb494744g4q 4 3 cases case NOUN fb494744g4q 4 4 , , PUNCT fb494744g4q 4 5 an an DET fb494744g4q 4 6 alternative alternative ADJ fb494744g4q 4 7 approach approach NOUN fb494744g4q 4 8 is be AUX fb494744g4q 4 9 to to PART fb494744g4q 4 10 utilize utilize VERB fb494744g4q 4 11 pattern pattern NOUN fb494744g4q 4 12 - - PUNCT fb494744g4q 4 13 mixture mixture NOUN fb494744g4q 4 14 models model NOUN fb494744g4q 4 15 . . PUNCT fb494744g4q 5 1 in in ADP fb494744g4q 5 2 this this DET fb494744g4q 5 3 dissertation dissertation NOUN fb494744g4q 5 4 , , PUNCT fb494744g4q 5 5 monte monte PROPN fb494744g4q 5 6 carlo carlo PROPN fb494744g4q 5 7 simulation simulation PROPN fb494744g4q 5 8 studies study NOUN fb494744g4q 5 9 are be AUX fb494744g4q 5 10 undertaken undertake VERB fb494744g4q 5 11 to to PART fb494744g4q 5 12 compare compare VERB fb494744g4q 5 13 the the DET fb494744g4q 5 14 traditional traditional ADJ fb494744g4q 5 15 ml ml PROPN fb494744g4q 5 16 method method NOUN fb494744g4q 5 17 and and CCONJ fb494744g4q 5 18 two two NUM fb494744g4q 5 19 different different ADJ fb494744g4q 5 20 approaches approach NOUN fb494744g4q 5 21 of of ADP fb494744g4q 5 22 pattern pattern NOUN fb494744g4q 5 23 - - PUNCT fb494744g4q 5 24 mixture mixture NOUN fb494744g4q 5 25 models model NOUN fb494744g4q 5 26 ( ( PUNCT fb494744g4q 5 27 i.e. i.e. X fb494744g4q 5 28 , , PUNCT fb494744g4q 5 29 the the DET fb494744g4q 5 30 d d NOUN fb494744g4q 5 31 - - PUNCT fb494744g4q 5 32 a a DET fb494744g4q 5 33 method method NOUN fb494744g4q 5 34 and and CCONJ fb494744g4q 5 35 the the DET fb494744g4q 5 36 a a DET fb494744g4q 5 37 - - PUNCT fb494744g4q 5 38 d d NOUN fb494744g4q 5 39 method method NOUN fb494744g4q 5 40 ) ) PUNCT fb494744g4q 5 41 across across ADP fb494744g4q 5 42 ( ( PUNCT fb494744g4q 5 43 1 1 NUM fb494744g4q 5 44 ) ) PUNCT fb494744g4q 5 45 different different ADJ fb494744g4q 5 46 variations variation NOUN fb494744g4q 5 47 of of ADP fb494744g4q 5 48 mixed mixed ADJ fb494744g4q 5 49 - - PUNCT fb494744g4q 5 50 effects effect NOUN fb494744g4q 5 51 models model NOUN fb494744g4q 5 52 ( ( PUNCT fb494744g4q 5 53 i.e. i.e. X fb494744g4q 5 54 , , PUNCT fb494744g4q 5 55 with with ADP fb494744g4q 5 56 or or CCONJ fb494744g4q 5 57 without without ADP fb494744g4q 5 58 adjustments adjustment NOUN fb494744g4q 5 59 ) ) PUNCT fb494744g4q 5 60 , , PUNCT fb494744g4q 5 61 ( ( PUNCT fb494744g4q 5 62 2 2 X fb494744g4q 5 63 ) ) PUNCT fb494744g4q 5 64 different different ADJ fb494744g4q 5 65 missing missing ADJ fb494744g4q 5 66 mechanisms mechanism NOUN fb494744g4q 5 67 ( ( PUNCT fb494744g4q 5 68 i.e. i.e. X fb494744g4q 5 69 , , PUNCT fb494744g4q 5 70 mar mar PROPN fb494744g4q 5 71 , , PUNCT fb494744g4q 5 72 random random ADJ fb494744g4q 5 73 - - PUNCT fb494744g4q 5 74 coefficient coefficient NOUN fb494744g4q 5 75 - - PUNCT fb494744g4q 5 76 dependent dependent ADJ fb494744g4q 5 77 mnar mnar ADJ fb494744g4q 5 78 or or CCONJ fb494744g4q 5 79 outcome outcome NOUN fb494744g4q 5 80 - - PUNCT fb494744g4q 5 81 dependent dependent ADJ fb494744g4q 5 82 mnar mnar PROPN fb494744g4q 5 83 ) ) PUNCT fb494744g4q 5 84 , , PUNCT fb494744g4q 5 85 and and CCONJ fb494744g4q 5 86 ( ( PUNCT fb494744g4q 5 87 3 3 X fb494744g4q 5 88 ) ) PUNCT fb494744g4q 5 89 different different ADJ fb494744g4q 5 90 types type NOUN fb494744g4q 5 91 of of ADP fb494744g4q 5 92 group group NOUN fb494744g4q 5 93 - - PUNCT fb494744g4q 5 94 based base VERB fb494744g4q 5 95 missing miss VERB fb494744g4q 5 96 probabilities probability NOUN fb494744g4q 5 97 . . PUNCT fb494744g4q 6 1 analytical analytical ADJ fb494744g4q 6 2 derivations derivation NOUN fb494744g4q 6 3 are be AUX fb494744g4q 6 4 also also ADV fb494744g4q 6 5 provided provide VERB fb494744g4q 6 6 to to PART fb494744g4q 6 7 explain explain VERB fb494744g4q 6 8 the the DET fb494744g4q 6 9 source source NOUN fb494744g4q 6 10 of of ADP fb494744g4q 6 11 bias bias NOUN fb494744g4q 6 12 across across ADP fb494744g4q 6 13 different different ADJ fb494744g4q 6 14 estimation estimation NOUN fb494744g4q 6 15 methods method NOUN fb494744g4q 6 16 . . PUNCT fb494744g4q 7 1 results result NOUN fb494744g4q 7 2 suggest suggest VERB fb494744g4q 7 3 that that SCONJ fb494744g4q 7 4 the the DET fb494744g4q 7 5 traditional traditional ADJ fb494744g4q 7 6 ml ml PROPN fb494744g4q 7 7 method method NOUN fb494744g4q 7 8 is be AUX fb494744g4q 7 9 well well ADV fb494744g4q 7 10 suited suited ADJ fb494744g4q 7 11 for for ADP fb494744g4q 7 12 mar mar PROPN fb494744g4q 7 13 data datum NOUN fb494744g4q 7 14 whereas whereas SCONJ fb494744g4q 7 15 the the DET fb494744g4q 7 16 proposed propose VERB fb494744g4q 7 17 pm pm NOUN fb494744g4q 7 18 - - PUNCT fb494744g4q 7 19 ad ad NOUN fb494744g4q 7 20 model model NOUN fb494744g4q 7 21 has have VERB fb494744g4q 7 22 the the DET fb494744g4q 7 23 best good ADJ fb494744g4q 7 24 overall overall ADJ fb494744g4q 7 25 performance performance NOUN fb494744g4q 7 26 for for ADP fb494744g4q 7 27 mnar mnar ADJ fb494744g4q 7 28 data datum NOUN fb494744g4q 7 29 . . PUNCT fb494744g4q 8 1 omitting omit VERB fb494744g4q 8 2 the the DET fb494744g4q 8 3 group group NOUN fb494744g4q 8 4 membership membership NOUN fb494744g4q 8 5 predictor predictor NOUN fb494744g4q 8 6 for for ADP fb494744g4q 8 7 the the DET fb494744g4q 8 8 intercept intercept NOUN fb494744g4q 8 9 or or CCONJ fb494744g4q 8 10 including include VERB fb494744g4q 8 11 the the DET fb494744g4q 8 12 baseline baseline NOUN fb494744g4q 8 13 score score NOUN fb494744g4q 8 14 as as ADP fb494744g4q 8 15 a a DET fb494744g4q 8 16 covariate covariate NOUN fb494744g4q 8 17 can can AUX fb494744g4q 8 18 lead lead VERB fb494744g4q 8 19 to to ADP fb494744g4q 8 20 larger large ADJ fb494744g4q 8 21 power power NOUN fb494744g4q 8 22 and and CCONJ fb494744g4q 8 23 smaller small ADJ fb494744g4q 8 24 bias bias NOUN fb494744g4q 8 25 in in ADP fb494744g4q 8 26 treatment treatment NOUN fb494744g4q 8 27 effect effect NOUN fb494744g4q 8 28 estimates estimate VERB fb494744g4q 8 29 when when SCONJ fb494744g4q 8 30 the the DET fb494744g4q 8 31 ml ml PROPN fb494744g4q 8 32 method method NOUN fb494744g4q 8 33 is be AUX fb494744g4q 8 34 used use VERB fb494744g4q 8 35 . . PUNCT fb494744g4q 9 1 applications application NOUN fb494744g4q 9 2 of of ADP fb494744g4q 9 3 different different ADJ fb494744g4q 9 4 estimation estimation NOUN fb494744g4q 9 5 methods method NOUN fb494744g4q 9 6 are be AUX fb494744g4q 9 7 also also ADV fb494744g4q 9 8 illustrated illustrate VERB fb494744g4q 9 9 using use VERB fb494744g4q 9 10 a a DET fb494744g4q 9 11 real real ADJ fb494744g4q 9 12 data data NOUN fb494744g4q 9 13 example example NOUN fb494744g4q 9 14 . . PUNCT