Salient Cues Enhance Strategy Monitoring Of Prospective Memory Retrieval

ObjectiveProspective memory(PM),here we explored event base prospective memory only,which is different from retrospective memory,means memory for the execution of planned events.It is always followed by a delayed intention to be fulfilled.The prospective interference effect means the decrease of behavioral performance of ongoing tasks when there are prospective memory tasks inserted into them.Smith explained there should be cognitive resources deployed by participants in the preparatory attentional processes to fulfill prospective memory tasks,and proposed the preparatory attentional processes and memory processes theory(PAM).McDaniel and Einstein proposed,however,that there should be automatic retrieval process in prospective memory retrieval in case of PM salient cues,highly associated target-intended action pairings.Chen et al.proved prospective interference effect in both behavior performance and psychophysiological data.They discovered significant difference in bilateral frontal areas from 200 ms to 300 ms.But they didn't classify salience of PM cues, and there are 20%PM cues in their experiment in total.Finally,in their experiment, the reactions for PM task were more complicate than ongoing task.Those two factors may blur the proof of automatic retrieval process.In this experiment,we intended to explore whether automatic retrieve process exist or not using stimulus pattern with both occurrence of low probability and saliency of different degree as PM cues and taking a simple button-pressing as PM response.MethodsTwenty college students(10 males and 10 females) aged between 20 and 28 years were recruited.All participants reported corrected to normal visual acuity, right-handed.The experiment was conducted in accordance with the Declaration of Helsinki and all subjects received￥40.00 for their participation.The experiment takes 50 minutes averagely for each participant.Every stimulus pattern was consist of a 4×4 gray grids array,subtending to a 5.60×5.60°(9.80×9.80 cm) visual angle.Each pattern had four different units (square) in other colors.Each unit subtended to a 1.15×1.15°(2.00×2.00 cm) visual angle at a 100 cm distance,and placed on a black background.The four squares were in yellow,purple,blue or green and differed from each other in ongoing tasks.The four grids of high salient cues were in identical color.The low salient cues had two squares in identical color only.Those cues were used in PM blocks and accounts for 10%of all trials.There were 4 stimuli at least between each two PM cue.Participants were instructed to judge whether every two successive patterns were identical in ongoing 1-back tasks.Both PM blocks included three types of trials and responses[i.e.,match 40%(press A key),non-match 60%(press B key),and prospective cues 10%(press C key)].There are no prospective cues and responses in control block.Participants were asked to response as quickly and accurately.There were three blocks[the control block,the high salient PM cue block(HS) and the low salient PM cue block(LS)].The control block(200 trials) was presented first,in which the cue patterns(10%trials) were irrelevant.The PM blocks(600 trials each,include 10%PM cues) were counterbalanced between subjects.Each pattern presented 500 ms,followed by black blank screen for 1500 ms. Reactions were valid between 150 ms and 2000 ms.The response keys were counterbalanced between subjects.There were 20 sec for participants to have a rest every 2 min interval within blocks and 3 min break between blocks.EEG(bandpass 0.5-100 Hz) was recorded using an ERP system developed in our lab,which could record behavior and EEG data.It was sampled from 19 electrodes mounted in a elastic cap according to the international 10-20 system referenced to both earlobes(Fig.2).Frontal midline site(Fz) was grounding.ERP epochs were extracted off-line and included 100 ms of pre-stimulus activity and of 1000 ms post-stimulus activity for ongoing tasks and PM tasks.Baseline was from -100 ms to 0 ms.Artifact above 70μV was eliminated manually.Data from two participants were excluded because excessive blink or over lowering correct rate.One way ANOVA and paired t test were performed for ongoing trials in three blocks and PM trials in two PM blocks respectively.The response time and correct rates were analyzed using SPSS 13.0 based on Greenhouse-Geissor correction.ERP data was analyzed using self-developed statistical software.We performed One way ANOVA for ongoing trials based on lower-bound epsilon and paired t test for PM trials.Statistical Parametric Mapping F values[SPM(F)]and t values[SPM(t)]were gained from interpolation calculated by each channel's F values or t values respectively.Significant level was 0.05.SPM(t) was corrected based on Bonferroni correction.ResultsThere was no significant difference among three blocks on ongoing tasks on both response time[F(1.52,25.82)=0.482,P=0.571]and correct rates[F(1,50, 25.57)=2.456,P=0.118].Response time was faster for high salient cue trials than for low salient cue trials[t(17)=-4.839,P=0.000＜0.001].Correct rates were higher for the former also[t(17)=4.611,P=0.000＜0.001].ERP from ongoing tasks and PM tasks 18 participants showed that:ERP waveforms from 160-240 ms(electrodes F7,F3,Fz,F4,F8,C3,Cz and C4) showed increased,but equivalent,activity in both the HS block and LS block relative to control block.Although it was no significant difference on electrode FP1.ERP waveforms from 360-440 ms(electrodes FP1 and FP2) showed increased,but equivalent,activity in both the control block and LS block relative to HS block.SPM(F) and SPM(t) of ongoing tasks in three blocks show average values in 80 ms(upper row:160-240 ms;lower row:360-440 ms).SPM(F) showed significant difference on bilateral frontal areas and right frontal pole in 160-240 ms and 360-440 ms respectively.SPM(t) showed significant difference between control block and both PM blocks on bilateral frontal areas in 160-240 ms at the level 0.05.SPM(t) showed significant difference between control block and high salient cue PM block in 360—440 ms at the level 0.05.Although there were no differences on all mentioned areas above all in SPM(t) by Bonferroni correction.It did showed the trends of being difference.Therefore we can still draw the conclusions that the two PM blocks were difference from the control block and the HS block was difference from the control block.SPM(t) between two PM blocks showed the amplitude of HS block was higher than LS block on frontal areas in 160-320 ms.Besides,amplitude of HS block was lower than LS block over occipital-parietal region in 160-320 ms,this region showed negative activity.And HS block showed higher positive amplitude than LS block over frontal and occipital-parietal regions in 360-620 ms.HS block also showed lower negative amplitude than LS block over frontal,central,parietal and occipital regions from 500 ms to 1000 ms.ConclusionNo significant prospective interference effect was found on behavior performance among three ongoing groups.So it seems PM retrieval is automatic and don't need to deploy cognitive resources.The significant differences between two PM groups on both correction rate and reaction times indicate cue salience improve behavior performance of PM tasks.We repeated N300 and prospective positivity.On the other way,we found that ERP waveforms from 160-240 ms on frontal areas showed increased,but equivalent, activity in both the HS group and LS group relative to control group.The difference was reliable.So our results were consistent with PAM theory but not multiprocess theory.On the other hand,ERP waveforms from 360-440 ms on right frontal pole showed increased,but equivalent,activity in both the control group and LS group relative to HS group.Salient cue does not lead to automatic retrieval.Contradictory,it works as vigilance that leads to more cognitive resources are deployed by the tasks. Moreover,the monitoring for task implementation does not always accompanied by behavior performance and conscious perception.