The Effect Of Switching Function On Adult’s Strategy Utilization In Computational Estimation: Evidence From Behavioral And ERP Study

Cognitive strategy is one of the indexes which reflect human’s cognitiveflexibility. And the adaptation of strategy, that one can choose the best strategy frommany candidates which he/she has already held to solve a particular problem, chieflyindicates the developmental level of people’s strategy utilization. The SCADS arguedthat switch plays an important role during the strategy selection. Task switching, asthe basic and general ability of human beings, helps people immediately deal with anew task after finishing the previous one in a complicate situation. In general, theperformance of the new task is impaired by the process of switch, which may costmore response time and/or even a lower accuracy rate and the switch cost is regardedas the sign of the switch processing. It has been confirmed to a certain amount ofdomains that switching has a widespread effect on people’s cognitive processing.Although a large number of studies on switching involving simple mathematical tasks(e.g. magnitude comparison, parity judgment), there are few researches revealing theeffect of switching on strategy utilization of arithmetic computation and itsmechanism.In the present study, we combined the modified choice/no choice method withthe operand recognition paradigm to explore how did the switching and cognitiveprocesses interact with each other and how did the switching processing affectindividuals’ strategy utilization (including strategy selection and strategy executionaspects) in two-digit arithmetic estimation tasks with the help of behavioral cognitivemeasures and event-related brain potentials (ERPs) technique. In Study1(includingExperiment1and2), the participants were asked to solve multiplication estimationtask first and then recognize the operands which they have operated before. InExperiment1, we explored the relationship between the problem type and switchingby the operand recognition paradigm (ORP). In Experiment2, we investigated whether the recognition performance was influenced in different strategy utilizationsituations. In Study2(including experiment3and4), we examined the effect ofswitching on strategy utilization directly. The participants had to perform two tasks inevery trail. First, they were asked to solve one of the comparison tasks (choose thelarger number or the deeper color one from two numbers as the first operand), andthen finish the estimation task. In Experiment3, we were interested in whether theaspects estimation strategy utilization was various in different switch situations. InExperiment4, we further scoped the brain mechanism in three switch conditions bythe technique of ERPs.Following are the major results:(1) The task type, problem type and the order of operands independently impactthe cost of both response time and correct rate. What’s more, the type of problemsimply induced the cost of response time.(2) Type of problem had a significant effect on switch cost, while the influencewas regulated by strategy utilization conditions. Different problems induce differentrecognition costs in the strategy selection condition.(3) Comparing with the color task, the magnitude task induced more estimationcosts in the strategy execution condition, which didn’t take place in the strategyselection condition.(4) Compared with the repeat condition, the switch conditions causedsignificant costs of initial strategy selection, and the costs from the magnitude taskwere more than that from the color task. The data from ERP showed that, N100peakelicited by baseline condition was larger than switch conditions both in prefrontal areaand central area., while there was no difference between the two switchconditions.P220latency of color condition was longer than that of repeat condition inthe central area, and the latency of P220was longer in prefrontal and central area thanthat in parietal area in the condition of both repeat and magnitude. The color conditiontriggered more positive P300than that of repeat condition in the central area, while the repeat condition triggered more positive P300in the parietal area.To sum up, the following conclusions were drawn:(1) Problem type trigered specific swith cost.The response time of recognition ofthe second operand was influenced by problem type.(2) The accuracy of recognition was influenced by the problem type, while theresponse time of recognition was varied by the effect of strategy utilizationconditions.(3) Different switch conditions caused unusual costs of estimation strategyexecution and initial strategy selection.(4) The effect of switching on initial strategy selection was varied over timeaccording the results of N100, P220and P300.