Hello everyone!
This is the Yinaoyun scientific research circle, I am sister Miaojun~
Recently, I received a doubt from a fan who was conducting EEG experiments: Hello, I am also doing visual working memory here, and I use the lateralized change awareness paradigm. I would like to ask what is needed if I want to obtain stable CDA waves. How many trials?
This is a very interesting problem, and it is also a common problem that many partners may encounter when conducting EEG experiments. Because I remember when I first started doing EEG experiments, I asked the friends around me, how many trials do I need to do EEG experiments? No less than 60 trials seems to be a "gold standard" that everyone blurts out, but with the development of technology, it seems that this "gold standard" is constantly being questioned and challenged.
How to obtain a stable and effective EEG component in EEG experiments has always been the direction of our continuous thinking and exploration, and we also hope that some thinking and accumulation in this area will be beneficial. We also had some exchanges through some platforms.
Closer to home, regarding the doubts of fans, William XQ Ngiam et al. (2021) published an article titled "Estimating the statistical power to detect set-size effects in contralateral delay activity" in Psychophysiology (as shown in the figure above). The issue was discussed in detail.
It was found that the required number of trials (number of trials) and sample size (subjects) depend largely on the size of the effect size (effect size). In the memory item processing stage, when the number of subjects is 25, it only needs 30-50 effective trials to achieve about 80% statistical power for whether there is a CDA component.
Then, if it is necessary to compare the difference of CDA between different loads, when the number of subjects is 25, about 400 effective trials are needed to achieve a statistical test power of about 80%.
Therefore, when determining the number of trials, we need to comprehensively consider the influence of factors such as effect size, sample size, and statistical power. How many trials are needed for researchers to evaluate their experiments?
William XQ Ngiam and others developed a CDA Power Calculator online calculation panel (as shown in the figure below).
The access link is:
https://williamngiam.shinyapps.io/CDAPower/.
Researchers can adjust the corresponding parameters to estimate according to their actual needs.
In addition to the CDA component, some scholars have made some exploratory analyzes and attempts on the number of trials required in the ERP component.
Such as: ERN (reference material [3]), LPP (reference material [4]), N400 (reference material [5]) and so on. In addition, Jensen, KM, & MacDonald, JA (2022, reference [6]) is based on the ERP CORE reference [7] database for seven components including ERN, LRP, MMN, N170, N2pc, N400, and P3 (as follows As shown in the figure), a more detailed exploration of the number of trials was carried out.
See reference [6] for more details. If there is time later, we will conduct a more detailed interpretation.
We also look forward to some small partners interpreting relevant literature, and look forward to your submission of relevant interpretation manuscripts to help more small partners to better carry out EEG experiments.
In addition, in order to make it easier for friends to better refer to the content of this issue of tweets, we have compiled relevant documents and uploaded them to Baidu cloud disk.
Get the link (extraction code: mrbe)
https://pan.baidu.com/s/1daQk3jBtTd5N0wzAXQ_nWw?pwd=mrbe
If the link fails, please send an email to: [email protected] to get it, thank you!
That’s all for today’s sharing, see you next time~
references:
[1] Ngiam, W. X., Adam, K. C., Quirk, C., Vogel, E. K., & Awh, E. (2021).Estimating the statistical power to detect set‐size effects in contralateral delay activity. Psychophysiology, 58(5), e13791.
[2] CDA Power Calculator.
https://williamngiam.shinyapps.io/CDAPower/.
[3] Boudewyn, M. A., Luck, S. J., Farrens, J. L., & Kappenman, E. S. (2018). How many trials does it take to get a significant ERP effect? It depends. Psychophysiology, 55(6), e13049.
[4] Gibney, K. D., Kypriotakis, G., Cinciripini, P. M., Robinson, J. D., Minnix, J. A., & Versace, F. (2020). Estimating statistical power for event‐related potential studies using the late positive potential. Psychophysiology, 57(2), e13482.
[5] Šoškić, A., Jovanović, V., Styles, S. J., Kappenman, E. S., & Ković, V. (2022). How to do better N400 studies: reproducibility, consistency and adherence to research standards in the existing literature. Neuropsychology Review, 32(3), 577-600.
[6]Jensen, K. M., & MacDonald, J. A. (2022). Towards thoughtful planning of ERP studies: How participants, trials, and effect magnitude interact to influence statistical power across seven ERP components. Psychophysiology, e14245.
[7] Kappenman, E. S., Farrens, J. L., Zhang, W., Stewart, A. X., & Luck, S. J. (2021). ERP CORE: An open resource for human event-related potential research. NeuroImage, 225, 117465.
Author丨Nian Jingqing
Reprint | Wandering heart ball
Typography | Fat T
Proofreading | Sister Blue Eucalyptus Miaojun