Research On Timing Accuracy Of Visual Stimuli In Online Experiments

Background With the gradual development of online experiments,online experiments are gradually recognized in various fields.However,the design tools of local experimental environment(PSYCHTOOLBOX,VISION EGG,e-prime,etc)are very limited for non-computer professionals,Such as MATLAB programming,PSYCHTOOLBOX installation,and experimental environment configuration.All this adds up to a lot of unnecessary learning costs.Local experiments are also subject to site constraints.Local experiments do not allow cross-regional research,For example,experimental analysis of different RACES.Online experiments can easily solve these problems.Online experiments can highly automate the experimental steps,making the cost of experiments more affordable.Now there are a lot of online experimental tools and platforms,some of which can conduct online questionnaire survey,some of which can design simple experimental paradigm.These experimental kits can achieve basic experimental functions but cannot present high-timing visual stimuli.The timing accuracy of visual stimuli in online experiments has not been sufficiently measured.The exact timing of visual stimuli in online experiments is unclear.This adds potential uncertainty to the results,even many online experiments fail to replicate the results of local experiments.There are three methods to realize the timing design of visual stimulation in online experiment.It is not clear whether the implementation of the three technologies has different effects on the timing accuracy.Objective This paper studies the timing accuracy of online visual stimuli by external measurement.The timing accuracy of online visual stimulation was measured and compared with the local visual stimulation tools.Determine the optimal technology scheme to achieve high precision timing sequence.Method Using SETINTERVAL,CSS3,RAF technology scheme to render dynamic raster and scintillation visual stimulation in WINDOWS,MAC OS,LINUX(UBUNTU)system CHROME,FIREFOX,EDGE,IE,SAFARI browser.External measures were used to measure the duration of visual stimuli.Obtain the timing accuracy of visual stimulation.PSYCHTOOLBOX,a local experimental visual stimulation kit,was used to generate the same raster and scintillating visual stimuli,and to measure their timing accuracy.The temporal accuracy of visual stimuli was compared between online and local experiments.Determine the optimal technique scheme of visual stimulation in online experiment.The timing accuracy of JSPSYCH was measured using the same method and compared with the results of local experiments.Potential defects in timing accuracy found in the JSPSYCH toolkit.Results 1)SETINTERVAL and CSS3 have low timing accuracy in the order of 10ms-100 ms.2)the timing accuracy of RAF is close to that of PSYCHTOOLBOX,reaching the order of sub-millisecond in WINDOWS,MAC,real-time priority and CHROME.3)RAF timing accuracy is greatly affected by browser,browser priority and operating system.The higher the priority,the better the timing accuracy.In the same condition,the timing accuracy of CHROME browser is generally higher than or equal to that of FIREFOX,SAFARI,IE and EDGE.RAF behaves similarly in Windows and MAC OS systems,but frame loss occurs in LINUX systems.4)in the case of real-time priority,RAF is not affected by button interference.5)g-sync technology has little influence on timing accuracy,and cannot improve the timing accuracy of SETINTERVAL and CSS3.6)the timing accuracy of JSPSYCH toolbox is consistent with SETINTERVAL,and the timing accuracy is on the order of 100 MS.7)RAF'shigh-precision timing sequence relies on high-performance stimulus rendering methodssuch as CAVAS.Conclusion The timing accuracy of online experiments varies widely,ranging from sub-milliseconds to hundreds of milliseconds.Timing accuracy depends on the technical scheme of the stimulus implementation.RAF's technical scheme can obtain the timing accuracy of sub-millisecond and meet the timing requirement of high-precision online experiment.