3._CRD Experiment_NN1 2_CRD 
In_PRP the_AT0 previous_AJ0 experiment_NN1 ,_PUN we_PNP found_VVD no_AT0 evidence_NN1 of_PRF any_DT0 behavioral_AJ0 differences_NN2 between_PRP the_AT0 Refresh_NN1 and_CJC Remove_VVB instruction_NN1 groups_NN2 ._SENT Thus_AV0 ,_PUN Experiment_NN1 1_CRD offered_VVD no_AT0 support_NN1 for_PRP ,_PUN and_CJC in_PRP fact_NN1 some_DT0 support_NN1 against_PRP ,_PUN a_AT0 functional_AJ0 distinction_NN1 between_PRP the_AT0 refresh_NN1 and_CJC remove_VVB processes_NN2 ,_PUN although_CJS that_DT0 task_NN1 design_NN1 did_VDD not_XX0 allow_VVI us_PNP to_TO0 look_VVI for_PRP certain_AJ0 potential_AJ0 patterns_NN2 of_PRF differences_NN2 ,_PUN such_DT0 as_PRP differential_AJ0 effects_NN2 on_PRP irrelevant_AJ0 items_NN2 during_PRP the_AT0 main_AJ0 task_NN1 ._SENT Specifically_AV0 ,_PUN we_PNP did_VDD not_XX0 ever_AV0 probe_VVI irrelevant_AJ0 items_NN2 during_PRP the_AT0 main_AJ0 task_NN1 ,_PUN so_CJS we_PNP could_VM0 not_XX0 observe_VVI whether_CJS there_EX0 was_VBD an_AT0 RT_NP0 benefit_NN1 for_PRP relevant_AJ0 over_PRP irrelevant_AJ0 items_NN2 during_PRP the_AT0 initial_AJ0 task_NN1 ,_PUN or_CJC whether_CJS the_AT0 magnitude_NN1 of_PRF such_DT0 a_AT0 benefit_NN1 might_VM0 differ_VVI between_PRP instruction_NN1 groups_NN2 ._SENT Thus_AV0 ,_PUN Experiment_NN1 2_CRD was_VBD conducted_VVN in_PRP the_AT0 same_DT0 manner_NN1 as_PRP Experiment_NN1 1_CRD ,_PUN except_CJS for_PRP the_AT0 inclusion_NN1 of_PRF invalid_AJ0 probes_NN2 (_PUL i.e._AV0 ,_PUN occasional_AJ0 probes_NN2 of_PRF irrelevant_AJ0 words_NN2 )_PUR in_PRP the_AT0 LDT_NP0 (_PUL see_VVB Figure_NN1 1a_CRD )_PUR ._SENT Adding_VVG these_DT0 invalid_AJ0 probes_NN2 allowed_VVD us_PNP to_TO0 examine_VVI the_AT0 fate_NN1 of_PRF the_AT0 irrelevant_AJ0 words_NN2 during_PRP the_AT0 main_AJ0 task_NN1 and_CJC further_AV0 illuminate_VVB any_DT0 potential_AJ0 differences_NN2 between_PRP instruction_NN1 groups_NN2 ._SENT We_PNP predicted_VVD faster_AV0 RTs_NP0 to_PRP relevant_AJ0 than_CJS irrelevant_AJ0 items_NN2 in_PRP both_DT0 instruction_NN1 groups_NN2 ,_PUN but_CJC that_CJT any_DT0 difference_NN1 between_PRP the_AT0 refresh_NN1 and_CJC removal_NN1 processes_NN2 would_VM0 manifest_VVI as_PRP an_AT0 interaction_NN1 ._SENT If_CJS there_EX0 were_VBD such_DT0 a_AT0 difference_NN1 between_PRP the_AT0 refresh_NN1 and_CJC removal_NN1 processes_NN2 ,_PUN presumably_AV0 RTs_NP0 to_PRP irrelevant_AJ0 items_NN2 in_PRP the_AT0 Remove_VVB instruction_NN1 group_NN1 would_VM0 be_VBI particularly_AV0 slowed_VVN (_PUL versus_PRP their_DPS simply_AV0 not_XX0 being_VBG refreshed_VVN in_PRP the_AT0 Refresh_NN1 instruction_NN1 group_NN1 )_PUR ,_PUN which_DTQ would_VM0 drive_VVI the_AT0 hypothesized_AJ0 interaction_NN1 ._SENT 
3.1_CRD ._SENT Method_NN1 
3.1.1_CRD ._SENT Participants_NN2 
Sixty-nine_CRD undergraduate_NN1 students_NN2 (_PUL N_ZZ0 =_NN1 54_CRD female_NN1 )_PUR ,_PUN between_PRP the_AT0 ages_NN2 of_PRF 18_CRD and_CJC 26_CRD (_PUL M_CRD =_NP0 19.74_CRD ,_PUN SD_NP0 =_NP0 1.66_CRD )_PUR ,_PUN from_PRP the_AT0 University_NN1 of_PRF Nebraska-Lincoln_NP0 participated_VVD in_PRP a_AT0 one-hour_AJ0 experiment_NN1 for_PRP course_NN1 credit_NN1 ._SENT All_DT0 participants_NN2 provided_CJS informed_AJ0 consent_NN1 prior_AJ0 to_PRP the_AT0 study_NN1 ,_PUN and_CJC procedures_NN2 were_VBD approved_VVN by_PRP the_AT0 Institutional_AJ0 Review_NN1 Board_NN1 at_PRP the_AT0 University_NN1 of_PRF Nebraska-_AJ0 Lincoln_NP0 ._SENT In_PRP Experiment_NN1 2_CRD ,_PUN each_DT0 participant_NN1 was_VBD randomly_AV0 assigned_VVN to_PRP either_AV0 the_AT0 Refresh_NN1 or_CJC Remove_VVB instruction_NN1 group_NN1 (_PUL N_NP0 =_NP0 30_CRD and_CJC 39_CRD ,_PUN respectively_AV0 )_PUR ._SENT Participants_NN2 that_CJT had_VHD already_AV0 participated_VVN in_PRP Experiment_NN1 1_CRD were_VBD ineligible_AJ0 to_TO0 take_VVI part_NN1 in_PRP Experiment_NN1 2._CRD As_CJS in_PRP Experiment_NN1 1_CRD ,_PUN participants_NN2 were_VBD not_XX0 informed_VVN that_CJT there_EX0 were_VBD different_AJ0 instruction_NN1 groups_NN2 until_CJS after_PRP the_AT0 experiment_NN1 had_VHD ended_VVN ;_PUN each_DT0 participant_NN1 only_AV0 received_VVN instructions_NN2 for_PRP his_DPS or_CJC her_DPS own_DT0 group_NN1 (_PUL Refresh_VVB or_CJC Remove_VVB )_PUR ._SENT An_AT0 additional_AJ0 18_CRD participants_NN2 were_VBD excluded_VVN from_PRP the_AT0 analyses_NN2 ,_PUN four_CRD due_AJ0 to_PRP poor_AJ0 compliance_NN1 with_PRP task_NN1 instructions_NN2 (_PUL 2_CRD Refresh_NN1 group_NN1 ,_PUN 2_CRD Remove_VVB group_NN1 )_PUR and_CJC 14_CRD due_AJ0 to_PRP LDT_NP0 accuracy_NN1 below_PRP 75_CRD %_NN0 correct_AJ0 (_PUL 5_CRD Refresh_NN1 group_NN1 ,_PUN 9_CRD Remove_VVB group_NN1 )_PUR in_PRP any_DT0 of_PRF the_AT0 three_CRD probe_NN1 type_NN1 conditions_NN2 ._SENT Exclusions_NN2 based_VVN on_PRP accuracy_NN1 ,_PUN with_PRP the_AT0 exception_NN1 of_PRF one_CRD participant_NN1 ,_PUN were_VBD due_AJ0 to_PRP accuracy_NN1 specifically_AV0 on_PRP invalid_AJ0 trials_NN2 ._SENT Accuracies_NN2 for_PRP remaining_AJ0 participants_NN2 were_VBD near_PRP ceiling_NN1 (_PUL see_VVB 3.2_CRD ._SENT Results_NN2 )_PUR ._SENT 
3.1.2_CRD ._SENT Procedure_NN1 and_CJC Tasks_NN2 
Experiment_NN1 1_CRD was_VBD fully_AV0 counterbalanced_VVN for_PRP all_DT0 of_PRF the_AT0 key_AJ0 trial_NN1 variables_NN2 in_PRP order_NN1 to_TO0 control_VVI for_PRP potential_NN1 confounds_VVZ ,_PUN rather_AV0 than_CJS relying_VVG on_PRP randomization_NN1 to_TO0 produce_VVI equivalence_NN1 between_PRP conditions_NN2 ._SENT Given_VVN the_AT0 introduction_NN1 of_PRF invalid_AJ0 trials_NN2 and_CJC the_AT0 resulting_AJ0 increase_NN1 in_PRP the_AT0 number_NN1 of_PRF permutations_NN2 of_PRF stimuli_NN2 ,_PUN the_AT0 roles_NN2 and_CJC positions_NN2 of_PRF specific_AJ0 words_NN2 within_PRP Experiment_NN1 2_CRD were_VBD simply_AV0 randomized_VVN rather_AV0 than_CJS counterbalanced_VVN ._SENT The_AT0 same_DT0 constraints_NN2 on_PRP trial_NN1 presentation_NN1 remained_VVD ;_PUN no_AV0 more_AV0 than_CJS three_CRD consecutive_AJ0 trials_NN2 having_VHG the_AT0 same_DT0 cue_NN1 position_NN1 or_CJC probe_NN1 type_NN1 (_PUL valid_AJ0 ,_PUN invalid_AJ0 ,_PUN or_CJC non-word_NN1 )_PUR ._SENT Participant_NN1 instructions_NN2 were_VBD changed_VVN slightly_AV0 to_TO0 accommodate_VVI the_AT0 invalid_AJ0 trials_NN2 ._SENT Other_AJ0 researchers_NN2 (_PUL Lewis-Peacock_NP0 et_NP0 al._NP0 ,_PUN 2018_CRD )_PUR have_VHB noted_VVN that_CJT the_AT0 presence_NN1 of_PRF invalid_AJ0 probes_NN2 may_VM0 complicate_VVI the_AT0 study_NN1 of_PRF WM_NP0 removal_NN1 ,_PUN as_CJS invalid_AJ0 trials_NN2 might_VM0 negate_VVI any_DT0 incentive_NN1 to_TO0 remove_VVI ._SENT In_PRP order_NN1 to_TO0 address_VVI the_AT0 presence_NN1 of_PRF invalid_AJ0 probes_NN2 and_CJC incentivize_VVB participants_NN2 to_TO0 comply_VVI with_PRP removal_NN1 cues_NN2 ,_PUN we_PNP acknowledged_VVD that_CJT we_PNP would_VM0 rarely_AV0 probe_VVI the_AT0 irrelevant_AJ0 words_NN2 ;_PUN however_AV0 ,_PUN since_CJS we_PNP were_VBD measuring_VVG reaction_NN1 time_NN1 ,_PUN we_PNP informed_VVD participants_NN2 that_CJT they_PNP would_VM0 still_AV0 be_VBI faster_AV0 overall_AJ0 if_CJS they_PNP followed_VVD the_AT0 cue_NN1 instructions_NN2 ._SENT Thus_AV0 ,_PUN as_CJS it_PNP was_VBD framed_VVN to_PRP participants_NN2 ,_PUN in_PRP service_NN1 of_PRF the_AT0 overall_AJ0 goal_NN1 of_PRF low_AJ0 average_AJ0 reaction_NN1 times_NN2 ,_PUN it_PNP was_VBD most_AV0 effective_AJ0 to_TO0 remove_VVI as_CJS cued_VVN ,_PUN given_VVN the_AT0 higher_AJC percentage_NN1 of_PRF valid_AJ0 and_CJC nonword_AJ0 probe_NN1 trials_NN2 ._SENT In_PRP any_DT0 case_NN1 ,_PUN if_CJS the_AT0 existence_NN1 of_PRF invalid_AJ0 trials_NN2 indeed_AV0 de-incentivized_AJ0 participants_NN2 from_PRP complying_VVG with_PRP removal_NN1 cues_NN2 ,_PUN that_DT0 should_VM0 become_VVI apparent_AJ0 in_PRP the_AT0 data_NN0 as_PRP a_AT0 reduction_NN1 in_PRP LTM_AJ0 differences_NN2 between_PRP relevant_AJ0 and_CJC irrelevant_AJ0 items_NN2 ._SENT As_CJS we_PNP did_VDD ultimately_AV0 find_VVI a_AT0 preserved_AJ0 LTM_NN1 benefit_NN1 for_PRP relevant_AJ0 items_NN2 (_PUL see_VVB Results_NN2 )_PUR ,_PUN in_PRP this_DT0 case_NN1 we_PNP do_VDB not_XX0 find_VVI reason_NN1 to_TO0 suspect_VVI that_CJT the_AT0 presence_NN1 of_PRF invalid_AJ0 trials_NN2 presented_VVD an_AT0 issue_NN1 for_PRP participants_NN2 '_POS compliance_NN1 with_PRP the_AT0 instructions_NN2 ._SENT 
In_PRP the_AT0 first_ORD experiment_NN1 ,_PUN LDT_NP0 word_NN1 probes_NN2 had_VHD an_AT0 equal_AJ0 probability_NN1 of_PRF being_VBG the_AT0 relevant_AJ0 word_NN1 or_CJC a_AT0 non-word_NN1 ;_PUN the_AT0 irrelevant_AJ0 words_NN2 were_VBD never_AV0 tested_VVN ._SENT The_AT0 probability_NN1 of_PRF encountering_VVG each_DT0 probe_NN1 type_NN1 in_PRP the_AT0 second_ORD experiment_NN1 was_VBD 40_CRD %_NN0 non-word_NN1 ,_PUN 40_CRD %_NN0 valid_AJ0 ,_PUN 20_CRD %_NN0 invalid_AJ0 ._SENT While_CJS this_DT0 would_VM0 change_VVI the_AT0 overall_AJ0 ratio_NN1 of_PRF word_NN1 to_PRP non-word_AJ0 probes_NN2 from_PRP equal_AJ0 (_PUL as_CJS in_PRP Experiment_NN1 1_CRD )_PUR to_PRP 60_CRD %_NN0 word_NN1 ,_PUN we_PNP felt_VVD that_CJT this_DT0 change_NN1 would_VM0 be_VBI unnoticeable_AJ0 to_PRP participants_NN2 ._SENT This_DT0 afforded_VVD us_PNP the_AT0 opportunity_NN1 to_TO0 test_VVI a_AT0 sufficient_AJ0 number_NN1 of_PRF invalid_AJ0 trials_NN2 while_CJS keeping_VVG the_AT0 proportion_NN1 of_PRF invalid_NN1 to_PRP valid_AJ0 trials_NN2 low_AJ0 enough_AV0 that_CJT invalid_AJ0 probes_NN2 would_VM0 not_XX0 influence_VVI participants_NN2 '_POS strategy_NN1 ._SENT Total_AJ0 trials_NN2 were_VBD decreased_VVN from_PRP 108_CRD (_PUL Experiment_NN1 1_CRD )_PUR to_PRP 90_CRD for_PRP Experiment_NN1 2._CRD Accordingly_AV0 ,_PUN the_AT0 LTM_AJ0 task_NN1 in_PRP Experiment_NN1 2_CRD consisted_VVD of_PRF all_DT0 270_CRD words_NN2 encountered_VVN in_PRP the_AT0 main_AJ0 task_NN1 and_CJC 180_CRD foils_NN2 ,_PUN keeping_VVG the_AT0 ratio_NN1 of_PRF previously_AV0 seen_VVN words_NN2 to_PRP foils_NN2 the_AT0 same_DT0 as_CJS in_PRP Experiment_NN1 1._CRD 
3.2_CRD ._SENT Results_NN2 
Response_NN1 time_NN1 to_PRP the_AT0 main_AJ0 task_NN1 LDT_NP0 probe_NN1 was_VBD tested_VVN with_PRP a_AT0 2_CRD (_PUL instruction_NN1 group_NN1 :_PUN Refresh_VVB or_CJC Remove_VVB )_PUR ×_PRP 3_CRD (_PUL probe_VVB type_NN1 :_PUN valid_AJ0 ,_PUN invalid_AJ0 ,_PUN non-word_AJ0 )_PUR mixed_AJ0 ANOVA_NN1 (_PUL see_VVB Figure_NN1 3a_CRD )_PUR ._SENT RTs_NP0 shorter_AJC than_CJS 200ms_NN2 and_CJC /_PUN or_CJC more_AV0 than_CJS three_CRD standard_AJ0 deviations_NN2 away_AV0 from_PRP the_AT0 participant_NN1 's_POS mean_AJ0 RT_NN1 were_VBD excluded_VVN from_PRP the_AT0 analysis_NN1 ,_PUN as_CJS in_PRP Experiment_NN1 1._CRD RTs_NP0 corresponding_AJ0 to_PRP incorrect_AJ0 LDT_NP0 responses_NN2 were_VBD also_AV0 excluded_VVN ._SENT In_PRP total_NN1 ,_PUN 6.02_CRD %_NN0 of_PRF all_DT0 trials_NN2 from_PRP analyzed_AJ0 participants_NN2 were_VBD excluded_VVN from_PRP analysis_NN1 ._SENT As_CJS noted_VVN above_AV0 ,_PUN participants_NN2 with_PRP <_NP0 75_CRD %_NN0 accuracy_NN1 in_PRP one_CRD or_CJC more_DT0 probe_NN1 type_NN1 conditions_NN2 were_VBD excluded_VVN from_PRP analysis_NN1 ;_PUN accuracy_NN1 was_VBD high_AJ0 in_PRP the_AT0 remaining_AJ0 participants_NN2 (_PUL 95.53_CRD %_NN0 on_PRP average_NN1 )_PUR ._SENT There_EX0 was_VBD a_AT0 main_AJ0 effect_NN1 of_PRF probe_NN1 type_NN1 ,_PUN F_NP0 (_PUL 1.828_CRD ,_PUN 122.465_CRD )_PUR =_NP0 102.583_CRD ,_PUN p_ZZ0 <_NP0 .001_CRD ,_PUN Greenhouse-Geisser_NP0 corrected_VVD ,_PUN with_PRP faster_AJC responses_NN2 to_PRP valid_AJ0 probes_NN2 (_PUL M_CRD =_NP0 548.379_CRD ,_PUN SD_NP0 =_NP0 116.661_CRD )_PUR than_CJS to_PRP non-word_AJ0 probes_NN2 (_PUL M_CRD =_NP0 663.781_CRD ,_PUN SD_NP0 =_NP0 135.021_CRD )_PUR ,_PUN p_ZZ0 <_NP0 .001_CRD ,_PUN and_CJC invalid_AJ0 probes_NN2 (_PUL M_CRD =_NP0 679.576_CRD ,_PUN SD_NP0 =_NP0 124.733_CRD )_PUR ,_PUN p_ZZ0 <_NP0 .001_CRD ._SENT Responses_NN2 to_PRP non-word_AJ0 probes_NN2 were_VBD not_XX0 significantly_AV0 different_AJ0 than_CJS to_PRP invalid_AJ0 probes_NN2 ,_PUN p_ZZ0 =_NP0 .300_CRD ._SENT The_AT0 main_AJ0 effect_NN1 of_PRF instruction_NN1 group_NN1 was_VBD not_XX0 significant_AJ0 ,_PUN F_NP0 (_PUL 1_CRD ,_PUN 67_CRD )_PUR =_NP0 3.727_CRD ,_PUN p_ZZ0 =_NP0 .058_CRD (_PUL Remove_VVB group_NN1 :_PUN M_CRD =_NP0 660.763_CRD ,_PUN SD_NP0 =_NP0 111.773_CRD ;_PUN Refresh_VVB group_NN1 :_PUN M_CRD =_NP0 607.359_CRD ,_PUN SD_NP0 =_NP0 115.525_CRD )_PUR ._SENT There_EX0 was_VBD no_AT0 significant_AJ0 probe_NN1 type_NN1 by_PRP instruction_NN1 group_NN1 interaction_NN1 ,_PUN F_NP0 (_PUL 1.829_CRD ,_PUN 122.465_CRD )_PUR =_NP0 0.821_CRD ,_PUN p_ZZ0 =_NP0 .433_CRD ._SENT 
LTM_AJ0 confidence_NN1 ratings_NN2 were_VBD tested_VVN with_PRP a_AT0 2_CRD (_PUL instruction_NN1 group_NN1 :_PUN Refresh_VVB or_CJC Remove_VVB )_PUR ×_PRP 8_CRD (_PUL word_VVB role_NN1 )_PUR mixed_AJ0 ANOVA_NN1 (_PUL see_VVB Figure_NN1 3b_CRD )_PUR ._SENT Word_NN1 roles_NN2 were_VBD the_AT0 same_DT0 as_CJS in_PRP Experiment_NN1 1_CRD ,_PUN with_PRP the_AT0 addition_NN1 of_PRF three_CRD new_AJ0 roles_NN2 corresponding_VVG to_PRP the_AT0 words_NN2 that_CJT had_VHD been_VBN seen_VVN on_PRP invalid_AJ0 probe_NN1 trials_NN2 (_PUL wherein_AVQ the_AT0 probe_NN1 was_VBD one_CRD of_PRF the_AT0 irrelevant_AJ0 words_NN2 ;_PUN thus_AV0 ,_PUN the_AT0 new_AJ0 word_NN1 roles_NN2 for_PRP invalid_AJ0 probe_NN1 trials_NN2 were_VBD the_AT0 relevant_AJ0 word_NN1 ,_PUN the_AT0 irrelevant_AJ0 word_NN1 that_CJT was_VBD probed_VVN ,_PUN and_CJC the_AT0 irrelevant_AJ0 word_NN1 that_CJT was_VBD not_XX0 probed_VVN )_PUR ._SENT LTM_AJ0 responses_NN2 faster_AV0 than_CJS 200ms_NN2 were_VBD deemed_VVN to_TO0 be_VBI accidental_AJ0 keypresses_NN2 ,_PUN resulting_VVG in_PRP 0.03_CRD %_NN0 of_PRF total_AJ0 responses_NN2 excluded_VVN from_PRP analysis_NN1 ._SENT 
For_PRP the_AT0 LTM_AJ0 confidence_NN1 ratings_NN2 ,_PUN there_EX0 was_VBD a_AT0 main_AJ0 effect_NN1 of_PRF word_NN1 role_NN1 ,_PUN F_NP0 (_PUL 4.430_CRD ,_PUN 296.799_CRD )_PUR =_NP0 118.537_CRD ,_PUN p_ZZ0 <_NP0 .001_CRD ,_PUN Greenhouse-Geisser_NP0 corrected_VVD ._SENT There_EX0 was_VBD no_AT0 main_AJ0 effect_NN1 of_PRF instruction_NN1 group_NN1 (_PUL Refresh_VVB versus_PRP Remove_VVB )_PUR ,_PUN F_ZZ0 (_PUL 1_CRD ,_PUN 67_CRD )_PUR =_NP0 1.454_CRD ,_PUN p_ZZ0 =_NP0 .232_CRD ,_PUN and_CJC no_AT0 group_NN1 by_PRP word_NN1 role_NN1 interaction_NN1 ,_PUN F_NP0 (_PUL 4.430_CRD ,_PUN 296.799_CRD )_PUR =_NP0 0.538_CRD ,_PUN p_ZZ0 =_NP0 .726_CRD ,_PUN Greenhouse-Geisser_NP0 corrected_VVD ._SENT A_AT0 planned_AJ0 comparison_NN1 testing_VVG the_AT0 difference_NN1 in_PRP confidence_NN1 ratings_NN2 between_PRP relevant_AJ0 (_PUL M_CRD =_NP0 2.651_CRD ,_PUN SD_NP0 =_NP0 0.420_CRD )_PUR and_CJC irrelevant_AJ0 (_PUL M_CRD =_NP0 2.448_CRD ,_PUN SD_NP0 =_NP0 0.365_CRD )_PUR words_NN2 from_PRP nonword-probe_AJ0 trials_NN2 was_VBD significant_AJ0 ,_PUN t_ZZ0 (_PUL 68_CRD )_PUR =_NP0 6.271_CRD ,_PUN p_ZZ0 <_NP0 .001_CRD ,_PUN Cohen_NP0 's_POS d_ZZ0 =_NN1 0.755_CRD ._SENT This_DT0 replicated_VVD the_AT0 pattern_NN1 seen_VVN in_PRP Experiment_NN1 1_CRD ,_PUN and_CJC again_AV0 was_VBD consistent_AJ0 with_PRP both_DT0 traditional_AJ0 directed_VVD forgetting_VVG and_CJC traditional_AJ0 refreshing_AJ0 LTM_NN1 effects_NN2 ._SENT No_AT0 attenuation_NN1 of_PRF the_AT0 effects_NN2 was_VBD observed_VVN in_PRP this_DT0 experiment_NN1 ,_PUN suggesting_VVG that_CJT the_AT0 inclusion_NN1 of_PRF invalid_AJ0 probe_NN1 trials_NN2 in_PRP the_AT0 main_AJ0 task_NN1 did_VDD not_XX0 generally_AV0 cause_VVI non-compliance_NN1 with_PRP the_AT0 remove_NN1 cue_NN1 instructions_NN2 ._SENT The_AT0 addition_NN1 of_PRF invalid_AJ0 probe_NN1 trials_NN2 in_PRP Experiment_NN1 2_CRD allowed_VVD a_AT0 second_ORD test_NN1 of_PRF the_AT0 relevant_AJ0 versus_PRP irrelevant_AJ0 words_NN2 that_CJT were_VBD not_XX0 subsequently_AV0 probed_VVN ,_PUN which_DTQ revealed_VVD that_DT0 relevant_AJ0 ,_PUN unprobed_AJ0 words_NN2 from_PRP invalid-probe_NN1 trials_NN2 (_PUL M_CRD =_NP0 2.682_CRD ,_PUN SD_NP0 =_NP0 0.485_CRD )_PUR were_VBD rated_VVN higher_AJC than_CJS irrelevant_AJ0 ,_PUN unprobed_AJ0 words_NN2 from_PRP invalid-probe_NN1 trials_NN2 (_PUL M_CRD =_NP0 2.496_CRD ,_PUN SD_NP0 =_NP0 0.467_CRD )_PUR ,_PUN t_ZZ0 (_PUL 68_CRD )_PUR =_NP0 3.959_CRD ,_PUN p_ZZ0 =_NP0 .005_CRD ,_PUN Cohen_NP0 's_POS d_ZZ0 =_NN1 0.477_CRD ,_PUN thus_AV0 paralleling_VVG the_AT0 effects_NN2 seen_VVN on_PRP nonword-probe_AJ0 trials_NN2 ._SENT 
As_CJS in_PRP the_AT0 previous_AJ0 experiment_NN1 ,_PUN further_AJC interpretation_NN1 of_PRF the_AT0 results_NN2 was_VBD explored_VVN with_PRP Bayesian_AJ0 analyses_NN2 ._SENT BANOVAs_NN2 of_PRF main_AJ0 task_NN1 RT_NP0 and_CJC LTM_NP0 both_AV0 had_VHD broadly_AV0 similar_AJ0 patterns_NN2 of_PRF results_NN2 to_PRP Experiment_NN1 1._CRD For_PRP main_AJ0 task_NN1 RT_NN1 (_PUL see_VVB Table_NN1 3_CRD )_PUR ,_PUN the_AT0 best_AJS model_NN1 did_VDD include_VVI both_DT0 a_AT0 main_AJ0 effect_NN1 of_PRF probe_NN1 type_NN1 (_PUL valid_AJ0 ,_PUN invalid_AJ0 ,_PUN nonword_AJ0 )_PUR and_CJC a_AT0 main_AJ0 effect_NN1 of_PRF instruction_NN1 group_NN1 (_PUL BF10_CRD =_NP0 4.422_CRD ×_PRP 1024_CRD 507_CRD ;_PUN extreme_AJ0 evidence_NN1 favoring_VVG this_DT0 model_NN1 over_PRP the_AT0 null_AJ0 )_PUR ,_PUN although_CJS this_DT0 model_NN1 's_POS likelihood_NN1 was_VBD only_AV0 slightly_AV0 higher_AJC than_CJS the_AT0 model_NN1 including_PRP only_AV0 a_AT0 main_AJ0 effect_NN1 of_PRF probe_NN1 type_NN1 (_PUL BF_NP0 =_NP0 1.455_CRD )_PUR ._SENT This_DT0 is_VBZ consistent_AJ0 with_PRP the_AT0 non-significant_AJ0 trend_NN1 for_PRP a_AT0 main_AJ0 effect_NN1 of_PRF instruction_NN1 group_NN1 (_PUL p_ZZ0 =_NN1 .058_CRD )_PUR we_PNP saw_VVD using_VVG NHST_NP0 ;_PUN thus_AV0 ,_PUN we_PNP cannot_VVB rule_VVB out_AVP the_AT0 possibility_NN1 of_PRF a_AT0 main_AJ0 effect_NN1 of_PRF group_NN1 in_PRP Experiment_NN1 2_CRD ,_PUN but_CJC the_AT0 evidence_NN1 is_VBZ not_XX0 conclusive_AJ0 either_DT0 way_NN1 ._SENT However_AV0 ,_PUN the_AT0 model_NN1 including_PRP an_AT0 interaction_NN1 of_PRF instruction_NN1 group_NN1 and_CJC probe_NN1 type_NN1 (_PUL as_CJS well_AV0 as_CJS both_DT0 main_AJ0 effects_NN2 )_PUR had_VHD lower_AJC likelihood_NN1 than_CJS either_AV0 the_AT0 model_NN1 including_PRP both_DT0 main_AJ0 effects_NN2 (_PUL BF_NP0 =_NP0 6.429_CRD )_PUR or_CJC the_AT0 model_NN1 including_PRP only_AV0 a_AT0 main_AJ0 effect_NN1 of_PRF probe_NN1 type_NN1 (_PUL BF_NP0 =_NP0 4.418_CRD )_PUR ._SENT Thus_AV0 ,_PUN the_AT0 magnitude_NN1 of_PRF the_AT0 evidence_NN1 against_PRP an_AT0 interaction_NN1 between_PRP probe_NN1 type_NN1 and_CJC instruction_NN1 group_NN1 was_VBD similar_AJ0 to_PRP that_DT0 found_VVN in_PRP Experiment_NN1 1._CRD 
An_AT0 LTM_AJ0 task_NN1 BANOVA_NN1 (_PUL see_VVB Table_NN1 4_CRD )_PUR found_VVD that_CJT the_AT0 best_AJS model_NN1 was_VBD a_AT0 main_AJ0 effect_NN1 model_NN1 of_PRF word_NN1 role_NN1 alone_AV0 (_PUL BF10_CRD =_NP0 5.668_CRD ×_PRP 1097_CRD 518_CRD ;_PUN overwhelming_AJ0 support_NN1 versus_PRP the_AT0 null_AJ0 model_NN1 )_PUR ,_PUN and_CJC including_PRP instruction_NN1 group_NN1 in_PRP the_AT0 model_NN1 penalized_VVD it_PNP by_PRP a_AT0 factor_NN1 of_PRF BF_NP0 =_NP0 1.730_CRD ._SENT Adding_VVG another_DT0 term_NN1 for_PRP an_AT0 interaction_NN1 of_PRF instruction_NN1 group_NN1 and_CJC word_NN1 role_NN1 further_AV0 lowered_VVD the_AT0 likelihood_NN1 of_PRF the_AT0 model_NN1 (_PUL BF_NP0 =_NP0 93.808_CRD )_PUR relative_AJ0 to_PRP the_AT0 model_NN1 including_PRP both_DT0 main_AJ0 effects_NN2 ._SENT Thus_AV0 ,_PUN the_AT0 simplest_AJS word-_AJ0 role-only_AJ0 model_NN1 was_VBD favored_VVN by_PRP a_AT0 factor_NN1 of_PRF 1.7_CRD over_PRP the_AT0 model_NN1 including_PRP an_AT0 effect_NN1 of_PRF group_NN1 ,_PUN and_CJC by_PRP a_AT0 factor_NN1 of_PRF 162.288_CRD (_PUL 1.730_CRD ×_PRP 93.808_CRD )_PUR over_PRP the_AT0 model_NN1 that_CJT also_AV0 included_VVD an_AT0 interaction_NN1 ._SENT Thus_AV0 ,_PUN the_AT0 Bayesian_AJ0 analyses_NN2 suggest_VVB that_CJT the_AT0 lack_NN1 of_PRF a_AT0 significant_AJ0 interaction_NN1 found_VVN with_PRP NHST_NP0 should_VM0 be_VBI interpreted_VVN as_CJS extremely_AV0 strong_AJ0 support_NN1 against_PRP the_AT0 existence_NN1 of_PRF such_DT0 an_AT0 interaction_NN1 ._SENT 
A_AT0 final_AJ0 pair_NN0 of_PRF mixed_AJ0 ANOVAs_NN2 (_PUL main_AJ0 task_NN1 RT_NN1 ,_PUN LTM_NP0 confidence_NN1 )_PUR were_VBD performed_VVN to_TO0 directly_AV0 compare_VVI the_AT0 results_NN2 of_PRF Experiments_NN2 1_CRD and_CJC 2._CRD For_PRP these_DT0 analyses_NN2 ,_PUN only_AV0 valid-probe_NN1 and_CJC nonword-probe_NN1 trials_NN2 were_VBD considered_VVN ,_PUN as_CJS there_EX0 were_VBD no_AT0 invalid-probe_NN1 trials_NN2 in_PRP Experiment_NN1 1._CRD For_PRP main_AJ0 task_NN1 RT_NN1 ,_PUN the_AT0 three-way_AJ0 interaction_NN1 between_PRP probe_NN1 type_NN1 ,_PUN Refresh_VVB /_PUN Remove_VVB instruction_NN1 group_NN1 ,_PUN and_CJC experiment_NN1 was_VBD non-significant_AJ0 ,_PUN F_NP0 (_PUL 1_CRD ,_PUN 137_CRD )_PUR =_NP0 0.320_CRD ,_PUN p_ZZ0 =_NP0 .572_CRD ._SENT For_PRP LTM_AJ0 ratings_NN2 ,_PUN the_AT0 three-_AJ0 way_NN1 interaction_NN1 of_PRF word_NN1 role_NN1 ,_PUN Refresh_VVB /_PUN Remove_VVB instruction_NN1 group_NN1 ,_PUN and_CJC experiment_NN1 was_VBD also_AV0 non-_AJ0 significant_AJ0 ,_PUN F_NP0 (_PUL 2.472_CRD ,_PUN 338.635_CRD )_PUR =_NP0 0.380_CRD ,_PUN p_ZZ0 =_NP0 .728_CRD ,_PUN Greenhouse-Geisser_NP0 corrected_VVD ._SENT Taken_VVN together_AV0 ,_PUN these_DT0 results_NN2 suggest_VVB that_CJT none_PNI of_PRF the_AT0 changes_NN2 between_PRP Experiments_NN2 1_CRD and_CJC 2_CRD substantively_AV0 affected_VVN the_AT0 results_NN2 for_PRP any_DT0 of_PRF the_AT0 conditions_NN2 shared_VVN between_PRP the_AT0 two_CRD experiments_NN2 ,_PUN and_CJC specifically_AV0 that_CJT the_AT0 inclusion_NN1 of_PRF invalid_AJ0 probes_NN2 in_PRP Experiment_NN1 2_CRD did_VDD not_XX0 alter_VVI the_AT0 way_NN1 in_PRP which_DTQ participants_NN2 approached_VVD the_AT0 task_NN1 ._SENT 
3.3_CRD ._SENT Discussion_NN1 
In_PRP the_AT0 main_AJ0 task_NN1 ,_PUN as_CJS in_PRP Experiment_NN1 1_CRD ,_PUN participants_NN2 were_VBD faster_AJC to_TO0 respond_VVI to_PRP valid_AJ0 probes_NN2 than_CJS to_PRP nonword_AJ0 probes_NN2 ._SENT Response_NN1 times_NN2 to_PRP invalid_AJ0 probes_NN2 were_VBD slower_AJC than_CJS to_PRP valid_AJ0 probes_NN2 ,_PUN reflecting_VVG the_AT0 expected_AJ0 directed_VVD forgetting_VVG effect_NN1 (_PUL Remove_VVB instruction_NN1 group_NN1 )_PUR or_CJC refresh_NN1 advantage_NN1 (_PUL Refresh_VVB instruction_NN1 group_NN1 )_PUR ._SENT Response_NN1 times_NN2 were_VBD slower_AV0 overall_AJ0 in_PRP Experiment_NN1 2_CRD than_CJS Experiment_NN1 1_CRD (_PUL main_AJ0 effect_NN1 of_PRF experiment_NN1 :_PUN p_ZZ0 <_NN1 .001_CRD )_PUR ,_PUN suggesting_VVG that_CJT the_AT0 inclusion_NN1 of_PRF invalid_AJ0 probes_NN2 added_VVD some_DT0 degree_NN1 of_PRF difficulty_NN1 or_CJC ambiguity_NN1 to_PRP the_AT0 task_NN1 overall_AJ0 ._SENT 
